Advances in Intelligent Systems and Interactive

Applications: Proceedings of the 2nd

International Conference on Intelligent and
Interactive Systems and Applications (IISA2017)
1st Edition Fatos Xhafa download
https://textbookfull.com/product/advances-in-intelligent-systems-
and-interactive-applications-proceedings-of-the-2nd-
international-conference-on-intelligent-and-interactive-systems-
and-applications-iisa2017-1st-edition-fatos-xhafa/

Download more ebook from https://textbookfull.com

We believe these products will be a great fit for you. Click
the link to download now, or visit textbookfull.com
to discover even more!

Emerging Trends in Intelligent and Interactive Systems

and Applications: Proceedings of the 5th International
Conference on Intelligent, Interactive Systems and
Applications (IISA2020) 1st Edition Madjid Tavana
https://textbookfull.com/product/emerging-trends-in-intelligent-
and-interactive-systems-and-applications-proceedings-of-the-5th-
international-conference-on-intelligent-interactive-systems-and-
applications-iisa2020-1st-edition-mad/

Smart Sensors Networks Communication Technologies and

Intelligent Applications A volume in Intelligent Data
Centric Systems Fatos Xhafa

https://textbookfull.com/product/smart-sensors-networks-
communication-technologies-and-intelligent-applications-a-volume-
in-intelligent-data-centric-systems-fatos-xhafa/

Intelligent Systems and Applications: Proceedings of

the 2020 Intelligent Systems Conference (IntelliSys)
Volume 2 Kohei Arai

https://textbookfull.com/product/intelligent-systems-and-
applications-proceedings-of-the-2020-intelligent-systems-
conference-intellisys-volume-2-kohei-arai/

Intelligent Systems and Applications: Proceedings of

the 2020 Intelligent Systems Conference (IntelliSys)
Volume 3 Kohei Arai

https://textbookfull.com/product/intelligent-systems-and-
applications-proceedings-of-the-2020-intelligent-systems-
conference-intellisys-volume-3-kohei-arai/
International Proceedings on Advances in Soft Computing
Intelligent Systems and Applications ASISA 2016 1st
Edition M. Sreenivasa Reddy

https://textbookfull.com/product/international-proceedings-on-
advances-in-soft-computing-intelligent-systems-and-applications-
asisa-2016-1st-edition-m-sreenivasa-reddy/

Intelligent Systems Design and Applications 19th

International Conference on Intelligent Systems Design
and Applications ISDA 2019 held December 3 5 2019 Ajith
Abraham
https://textbookfull.com/product/intelligent-systems-design-and-
applications-19th-international-conference-on-intelligent-
systems-design-and-applications-isda-2019-held-
december-3-5-2019-ajith-abraham/

Intelligent Systems Design and Applications 16th

International Conference on Intelligent Systems Design
and Applications ISDA 2016 held in Porto Systems and
Computing 557 Band 557 Ana Maria Madureira (Editor)
https://textbookfull.com/product/intelligent-systems-design-and-
applications-16th-international-conference-on-intelligent-
systems-design-and-applications-isda-2016-held-in-porto-systems-
and-computing-557-band-557-ana-maria-madureira/

Intelligent Interactive Multimedia Systems and Services

2016 1st Edition Giuseppe De Pietro

https://textbookfull.com/product/intelligent-interactive-
multimedia-systems-and-services-2016-1st-edition-giuseppe-de-
pietro/

Research and Development in Intelligent Systems XXX

Incorporating Applications and Innovations in
Intelligent Systems XXI Proceedings of AI 2013 The
Thirty third SGAI International Conference on
Innovative Techniques and Applications of Artificial
https://textbookfull.com/product/research-and-development-in-
intelligent-systems-xxx-incorporating-applications-and-
Intel 1st Edition Sebastian Peter
innovations-in-intelligent-systems-xxi-proceedings-of-
Advances in Intelligent Systems and Computing 686

Fatos Xhafa
Srikanta Patnaik
Albert Y. Zomaya Editors

Advances in
Intelligent Systems
and Interactive
Applications
Proceedings of the 2nd International
Conference on Intelligent and
Interactive Systems and Applications
(IISA2017)
Advances in Intelligent Systems and Computing

Volume 686

Series editor
Janusz Kacprzyk, Polish Academy of Sciences, Warsaw, Poland
e-mail: [email protected]
About this Series
The series “Advances in Intelligent Systems and Computing” contains publications on theory,
applications, and design methods of Intelligent Systems and Intelligent Computing. Virtually
all disciplines such as engineering, natural sciences, computer and information science, ICT,
economics, business, e-commerce, environment, healthcare, life science are covered. The list
of topics spans all the areas of modern intelligent systems and computing.
The publications within “Advances in Intelligent Systems and Computing” are primarily
textbooks and proceedings of important conferences, symposia and congresses. They cover
significant recent developments in the field, both of a foundational and applicable character.
An important characteristic feature of the series is the short publication time and world-wide
distribution. This permits a rapid and broad dissemination of research results.
Advisory Board
Chairman
Nikhil R. Pal, Indian Statistical Institute, Kolkata, India
e-mail: [email protected]
Members
Rafael Bello Perez, Universidad Central “Marta Abreu” de Las Villas, Santa Clara, Cuba
e-mail: [email protected]
Emilio S. Corchado, University of Salamanca, Salamanca, Spain
e-mail: [email protected]
Hani Hagras, University of Essex, Colchester, UK
e-mail: [email protected]
László T. Kóczy, Széchenyi István University, Győr, Hungary
e-mail: [email protected]
Vladik Kreinovich, University of Texas at El Paso, El Paso, USA
e-mail: [email protected]
Chin-Teng Lin, National Chiao Tung University, Hsinchu, Taiwan
e-mail: [email protected]
Jie Lu, University of Technology, Sydney, Australia
e-mail: [email protected]
Patricia Melin, Tijuana Institute of Technology, Tijuana, Mexico
e-mail: [email protected]
Nadia Nedjah, State University of Rio de Janeiro, Rio de Janeiro, Brazil
e-mail: [email protected]
Ngoc Thanh Nguyen, Wroclaw University of Technology, Wroclaw, Poland
e-mail: [email protected]
Jun Wang, The Chinese University of Hong Kong, Shatin, Hong Kong
e-mail: [email protected]

More information about this series at http://www.springer.com/series/11156

Fatos Xhafa Srikanta Patnaik
•

Albert Y. Zomaya
Editors

Advances in Intelligent
Systems and Interactive
Applications
Proceedings of the 2nd International Conference
on Intelligent and Interactive Systems
and Applications (IISA2017)

123
Editors
Fatos Xhafa Albert Y. Zomaya
Technical University of Catalonia School of Information Technologies
Barcelona University of Sydney
Spain Sydney, NSW
Australia
Srikanta Patnaik
Department of Computer Science
and Engineering, Faculty of Engineering
and Technology
SOA University
Bhubaneswar, Odisha
India

ISSN 2194-5357 ISSN 2194-5365 (electronic)

Advances in Intelligent Systems and Computing
ISBN 978-3-319-69095-7 ISBN 978-3-319-69096-4 (eBook)
https://doi.org/10.1007/978-3-319-69096-4
Library of Congress Control Number: 2017956074

© Springer International Publishing AG 2018

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

As we all know, Intelligent-interactive Systems (IIS) are systems that interact with
human beings, media, or virtual agents in intelligent computing environments. The
emergence of Big Data and Internet of Things opened new opportunities in both
academic and industrial research for successful design and development of
Intelligent-interactive Systems. These systems are usually embodied with human
capabilities such as to reason, learn, perceive, take decisions, make plans, and
perform actions in a wide range of applications such as smart navigation of auto-
mobiles so as to avoid traffic congestion, intelligent transportations, smart light,
virtual environments, mood-based music systems, and many more. The develop-
ment of these Intelligent-interactive Systems involves fabricating and controlling
machines to perform tasks usually need to be done by human beings.
Since research on Intelligent-interactive Systems deals with the pressing need of
designing innovative systems requiring new approaches to current computing
architectures and interactions between systems and their peripherals, it explicitly
demands knowledge acquisition in many cross-disciplinary subjects such as algo-
rithmic foundations, computational and mathematical modeling fundamentals,
artificial and cognitive intelligence foundations, human–computer interaction,
software architecture, and developmental concepts and logics of embedded systems.
Some of the significant underlying questions include the following: How huge
amounts of data are to be collected and how can they be further analyzed for
enhancing the interaction between intelligent systems and users, what are the
various issues and challenges faced during technical design and realization of
Intelligent-interactive Systems?
This volume explores how novel interactive systems can intelligently face var-
ious challenges and limitations previously encountered by human beings using
different machine learning algorithms along with analysis of recent trends.
However, designing and development of Intelligent-interactive Systems are quite
hard to implement due to variations in abilities of human being, preferences, and
limitations. The current researchers are focusing on two major areas: artificial
intelligence and human–computer interaction. Further, these two areas are classified
into application-specific sub-areas such as computer vision, speech recognition or

v
vi Preface

processing, Web-based systems, data visualization. Some of the widely adopted

research areas include information retrieval, recommender systems, intelligent
learning systems, natural language processing, multimodal interactions, ubiquitous
computing, personalized applications, and semantic-based applications.
Although many researchers are working on various issues of the above-
mentioned areas, still lots of challenges and problems arise since it is still in an
immature stage. One of the common issues is how both human and artificial
intelligence can be combined together to attain efficient results. Apart from these,
there are issues relating to users’ preferences and privacy protection, which have to
be addressed by various intelligent algorithms.
The volume further discusses current research issues and sheds light on future
research directions by including various recent and original research articles sys-
tematically addressing real-world problems, proposing novel interfaces such as
virtual environments, wearable interfaces, and novel interaction models such as
touch screens, speech-enabled systems, establishing new standards and metrics for
evaluations of model effectiveness and so on. The current research scenario of
Intelligent-interactive Systems involves understanding the mechanism lying behind
human interactions and how that can be embedded into various applications.
Various dimensions of the IIS includes intelligent user interfaces, context
awareness and adaptability, human–computer interaction, wearable technologies,
smart solutions for real-time problems, smart navigation, data-driven social ana-
lytics, mobile robotics, virtual communication environments, face and gesture
analysis, and crowdsourcing.
This volume contains 125 contributions from diverse areas of Intelligent-
interactive Systems (IIS), which has been categorized into seven sections, namely
(i) Autonomous Systems; (ii) Pattern Recognition and Vision Systems;
(iii) E-Enabled Systems. (iv) Mobile Computing and Intelligent Networking;
(v) Internet & Cloud Computing; (vi) Intelligent Systems and (vii) Various
Applications.
(I) Autonomous Systems: This is one of the established areas of interactive
intelligence system typically consisting of learning, reasoning, decision
making which support the system’s primary function. There are nine
(9) contributions consisting of various algorithms, models, and learning
techniques.
(II) Pattern Recognition and Vision Systems: This is one of the primary
functions of any Intelligent-interactive Systems. There are forty-five
(45) contributions comprised in this section covering the developments in
this area of deep learning to binocular stereovision to 3D vision.
(III) E-Enabled Systems: This is one of the essential areas of Intelligent-
interactive Systems, as many interactive systems are now designed through
Internet. It covers information navigation and retrieval, designing intelligent
learning environments, and model-based user interface design. There are
twelve (12) contributions covered in this section.
Preface vii

(IV) Mobile and Wireless Communication: This area is one of the leading areas
of IIS, which covers ubiquitous or mobile computing and networking. This
section covers twenty-three (23) contributions.
(V) Internet and Cloud Computing: It is one of the essential areas of IIS, which
caters to enhance communication between the system and users, in a way
which may not be closely related to the system’s main function. This is
commonly found in the areas of multimodal interaction, natural language
processing, embodied conversational agents, computer graphics, and acces-
sible computing. In this section, there are four (4) contributions consisting of
micro-blogging, user satisfaction modeling to the design and construction of
graphical cloud computing platform.
(VI) Intelligent Systems: This is the nervous system of IIS, and many researchers
are engaged in this area of research. This section contains 11 contributions.
(VII) Applications: Applications of IIS in various domains are covered in the last
section which consists of 21 contributions.

Acknowledgements

The contributions covered in this proceeding are the outcome of the contributions
from more than hundred researchers. We are thankful to the authors and paper
contributors of this volume.
We are thankful to the editor in chief of the Springer Book series on “Advances
in Intelligent Systems and Computing” Prof Janusz Kacprzyk for his support to
bring out the second volume of the conference, i.e., IISA-2017. It is noteworthy to
mention here that constant support from the editor in chief and the members of the
publishing house makes the conference fruitful for the second edition.
We would like to appreciate the encouragement and support of Dr. Thomas
Ditzinger, executive editor, and his Springer publishing team.
We are thankful to our friend Prof. Madjid Tavana, Professor and Lindback
Distinguished Chair of Business Analytics of La Salle University, USA, for his
keynote address. We are also thankful to the experts and reviewers who have
worked for this volume despite the veil of their anonymity.
We look forward to your valued contribution and supports to next editions of the
International Conference on Intelligent and Interactive Systems and Applications.
We are sure that the readers shall get immense benefit and knowledge from the
second volume of the area of Intelligent and Interactive Systems and Applications.

Fatos Xhafa
Srikanta Patnaik
Albert Y. Zomaya
Contents

Autonomous Systems
Prediction of Rolling Force Based on a Fusion of Extreme Learning
Machine and Self Learning Model of Rolling Force . . . . . . . . . . . . . . . . 3
AZiGuLi, Can Cui, Yonghong Xie, Shuang Ha, and Xiaochen Wang
User Behavior Profiles Establishment in Electric Power Industry . . . . . 12
Wei Song, Gang Liu, Di Luo, and Di Gao
The Optimal Pan for Baking Brownie . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Haiyan Li, Tangyu Wang, Xiaoyi Zhou, Guo Lei, Pengfei Yu,
Yaqun Huang, and Jun Wu
Intelligent Steering Control Based the Mathematical Motion Models
of Collision Avoidance for Fishing Vessel . . . . . . . . . . . . . . . . . . . . . . . . 27
Renqiang Wang, Yuelin Zhao, Keyin Miao, and Jianming Sun
Dynamical Analysis of Fractional-Order Hyper-chaotic System . . . . . . . 36
Junqing Feng and Guohong Liang
Evaluating the Performance of the Logistics Parks:
A State-of-the-Art Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Yingxiu Qi, Yan Sun, and Maoxiang Lang
A Tabu Search Algorithm for Loading Containers
on Double-Stack Cars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Zepeng Wang, Maoxiang Lang, Xuesong Zhou, Jay Przybyla,
and Yan Sun
Optimization Model Under Grouping Batch for Prefabricated
Components Production Cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Chun Guang Chang and Yu Zhang
Air Quality Evaluation System Based on Stacked Auto-Encoder . . . . . . 63
Yuxuan Zhuang, Liang Chen, and Xiaojie Guo

ix
x Contents

Pattern Recognition and Vision Systems

Urban-Rural Difference of the Coupling Between Social-Economic
Development and Landscape Pattern in Chengdu Plain . . . . . . . . . . . . . 71
Zhang Huabing, Lu Dapei, Zhen Yan, Han Shuang, and Chen Hongquan
A New Model and Algorithm for Clustering . . . . . . . . . . . . . . . . . . . . . 76
Guohong Liang, Ying Li, and Junqing Feng
Gibbs Phenomenon for Bi-orthogonal Wavelets . . . . . . . . . . . . . . . . . . . 82
Jie Zhou and Hongchan Zheng
a-Convergence Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Chun-Qiu Ji, Zhen-Guo Xu, and Yi Wang
Feature Selection Optimization Based on Atomic Set and Genetic
Algorithm in Software Product Line . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Zhijuan Zhan, Weilin Luo, Zonghao Guo, and Yumei Liu
Product Configuration based on Feature Model . . . . . . . . . . . . . . . . . . . 101
Zhijuan Zhan, Yunjiao Zhan, Mingyu Huang, and Yumei Liu
Research on Fast Browsing for Massive Image . . . . . . . . . . . . . . . . . . . 107
Fang Wang, Ying Peng, and Xiaoya Lu
RVM for Recognition of QRS Complexes in Electrocardiogram . . . . . . 113
Lu Bing, Xiaolei Han, and Wen Si
The Architecture of the RFID-Based Intelligent Parking System . . . . . . 119
Xiao Xiao and Qingquan Zou
A Review of Cognitive Psychology Applied in Robotics . . . . . . . . . . . . . 125
Huang Qin, Tao Yun, Wang Yujin, Wu Changlin, and Yu Lianqing
Robot Vision Navigation Based on Improved ORB Algorithm . . . . . . . . 134
Sun-Wen He, Zhang-Guo Wei, and Lu-Qiu Hong
An Improved Multi-label Relief Feature Selection Algorithm
for Unbalanced Datasets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Yonghong Xie, Daole Li, Dezheng Zhang, and Ha Shuang
Area Topic Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Hongchen Guo, Liang Zhang, and Zhiqiang Li
Predicting Popularity of Topic Based on Similarity Relation
and Co-occurrence Relation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Lu Deng, Qiang Liu, Jing Xu, Jiuming Huang, Bin Zhou, and Yan Jia
A New Bayesian-Based Method for Privacy-Preserving Data Mining . . . 171
Guang Li
Contents xi

Gesture Recognition Algorithm Based on Fingerprint Detection . . . . . . 178

Ge-Yan Ru, Zhang-Guo Wei, and Lu-Qiu Hong
Parka: A Parallel Implementation of BLAST with MapReduce . . . . . . . 185
Li Zhang and Bing Tang
Research of Moving Target Tracking Algorithms in Video
Surveillance System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
Xu Lei, Peng Yueping, and Liu Man
Study on Tumble Behavior Recognition Based on Mining
Algorithm for Potential Behavior Association . . . . . . . . . . . . . . . . . . . . 198
Zhang Qiusheng, Lin Mingyu, and Ju Jianping
Image Geometric Correction Based on Android Phone Sensors . . . . . . . 204
Zhengyi Xiao, Zhikai Fang, Liansheng Gao, and Baotao Xu
Fish-Eye Camera Model and Calibration Method . . . . . . . . . . . . . . . . . 215
Cuilin Li, Guihua Han, and Jianping Ju
Algorithm for Digital Location and Recognition of Digital
Instrument in Complex Scenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
Hao Zhou, Juntao Lv, Guoqing Yang, Zhimin Wang, Mingyang Liu,
and Junliang Li
The Visualization Analysis on Present Research Situation
and Trend of Tackling Overcapacity of Energy . . . . . . . . . . . . . . . . . . . 231
Na Zheng, Dangguo Shao, Lei Ma, Yan Xiang, Ying Xu, Wei Chen,
and Zhengtao Yu
Hybrid Algorithm for Prediction of Battlefield Rescue Capability
of Brigade Medical Aid Station . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Wen-Ming Zhou, San-Wei Shen, Wen-Xiang Xia, Chun-Rong Zhang,
and Hai-Long Deng
The Design of Optimal Synthesis Filter Bank and Receiver
for the FBMC System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Yan Yang and Pingping Xu
A Tool for IMA System Configuration Verification and Case Study . . . 254
Lisong Wang, Ying Zhou, Mingming Wang, and Jun Hu
Using Deep ConvNet for Robust 1D Barcode Detection . . . . . . . . . . . . . 261
Jianjun Li, Qiang Zhao, Xu Tan, Zhenxing Luo, and Zhuo Tang
Study of Agricultural Machinery Operating System Based
on Beidou Satellite Navigation System . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Fanwen Meng, Dongkai Yang, Youquan Wang, and Yuxiang Zhang
xii Contents

Detection Algorithm of Slow Radial Velocity Ship based

on Non-Negative Matrix Factorization by Over-The-Horizon Radar . . . 274
Hui Xiao, Jun Yuan, Shaoying Shi, and Runhua Liu
A v-Twin Bounded Support Tensor Machine for Image Classification . . . 280
Biyan Dai, Huiru Wang, and Zhijian Zhou
Super-resolution Reconstruction of Face Image Based
on Convolution Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
Wenqing Huang, Yinglong Chen, Li Mei, and Hui You
An Image Retrieval Algorithm Based on Semantic
Self-Feedback Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
Lang Pei, Jia Xu, and Jinhua Cai
Using Supervised Machine Learning Algorithms to Screen Down
Syndrome and Identify the Critical Protein Factors . . . . . . . . . . . . . . . . 302
Bing Feng, William Hoskins, Jun Zhou, Xinying Xu, and Jijun Tang
A Multi-view Approach for Visual Exploration of Temporal
Multi-dimensional Vehicle Experiment Data . . . . . . . . . . . . . . . . . . . . . 309
Xianglei Zhu, Haining Tong, Shuai Zhao, Quan Wen, and Jie Li
Impact Analysis of Geometry Parameters of Buoy on the Pitching
Motion Mechanism and Power Response for Multi-section
Wave Energy Converter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
Biao Li and Hongtao Gao
Binary Tree Construction of Multiclass Pinball SVM Via Farthest
Centroid Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
Qiangkui Leng, Fude Liu, and Yuping Qin
Research on Fractal Feature Extraction of Radar Signal Based
on Wavelet Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
Shen Lei, Han Yu-sheng, and Wang Shuo
A Method of Moving Target Detection Based on Scaling Background . . . 337
Yu-chen Tang and Xu-dong Yang
Facial Expression Recognition Based on Deep Learning: A Survey . . . . 345
Ting Zhang
A Fast Connected Components Analysis Algorithm
for Object Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
Dai Dehui and Li Zhiyong
Bibliometrics and Visualization Analysis of Knowledge Map
in Metallurgical Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
Ying Xu, Yan Xiang, Dangguo Shao, Zhengtao Yu, Na Zheng, Wei Chen,
and Lei Ma
Contents xiii

Hierarchical Decision Tree Model for Human Activity Recognition

Using Wearable Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
Cheng Xu, Jie He, and Xiaotong Zhang
Human Motion Monitoring Platform Based on Positional
Relationship and Inertial Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373
Jie He, Cunda Wang, Cheng Xu, and Shihong Duan
Three-Order Computational Ghost Imaging
and Its Inverse Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
Wenbing Zeng, Yi Xu, and Dong Zhou
Face Recognition Using Deep Convolutional Neural Network
in Cross-Database Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
Mei Guo, Min Xiao, and Deliang Gong

E-Enabled Systems
Stock Market Forecasting Using S-System Model . . . . . . . . . . . . . . . . . 397
Wei Zhang and Bin Yang
Online Detection Approach to Auto Parts Internal Defect . . . . . . . . . . . 404
Lin Mingyu and Ju Jianping
Globalized Translation Talent Training Model based
on Artificial Intelligence and Big Data . . . . . . . . . . . . . . . . . . . . . . . . . . 410
Fang Yang
A Study on the Relationship between Enterprise Education
and Training and Operational Performance—A Cases Study
of a Multinational Group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
Yung Chang Wu, Lin Feng, and Shiann Ming Wu
Online Handwritten Character Recognition of New Tai Lue Based
on Online Random Forests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426
Yong Yu, Pengfei Yu, Haiyan Li, and Hao Zhou
Web Information Transfer Between Android Client and Server . . . . . . 435
Baoqin Liu
Real-time Dynamic Data Analysis Model Based
on Wearable Smartband . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442
Xiangyu Li, Chixiang Wang, Haodi Wang, and Junqi Guo
Investigation and Analysis on the Influencing Factors of Consumers’
Trust to Fresh Agricultural Products in E-commerce . . . . . . . . . . . . . . 450
Yipeng Li and Yong Zhang
A PROUD Methodology for TOGAF Business Architecture Modeling . . . 457
Feng Ni, Fang Dai, Michael J. Ryoba, Shaojian Qu, and Ying Ji
xiv Contents

Application of SVM Based on Information Entropy

in Intrusion Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464
Nuo Jia and Dan Liu
Research of Music Recommendation System Based on User
Behavior Analysis and Word2vec User Emotion Extraction . . . . . . . . . . 469
Qiuxia Li and Dan Liu
Exploration of Information Security Education
of University Students . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 476
Shi Wang, Yongxin Qu, Likun Zheng, Yawen Xiao, and Huiying Shi

Mobile and Wireless Communication

Research on the Path Monitoring Capability of Wireless
Multimedia Sensor Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
Zhao Jing, Liu Zhuohua, and Xue Songdong
Preopen Set and Preclosed Set on Intuitive Fuzzy Topological Spaces . . . 490
Zhen-Guo Xu, Rui-Dan Wang, Zhi-Hui Chen, and Ying Zhao
Neural Network Method for Compressed Sensing . . . . . . . . . . . . . . . . . 497
Zixin Liu, Yuanan Liu, Nengfa Wang, and Lianglin Xiong
A Survey on Microwave Surface Emissivity Retrieval Methods . . . . . . . 503
De Xing, Qunbo Huang, Bainian Liu, and Weimin Zhang
Design of Multi-mode Switching Damping Shock Absorber
for Active Suspension and Ride Comfort Test . . . . . . . . . . . . . . . . . . . . 510
Zhao Jing-bo, Han Bing-yuan, Bei Shao-yi, and Liu Hai-mei
Research on the System of Patrol Unmanned Aerial Vehicle (UAV)
Docking on Charging Pile based on Autonomous Identification
and Tracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517
Zinan Qiu, Kai Zhang, and Yuhan Dong
On Data Analysis in Forest Fire Induced Breakdown
of the Transmission Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524
Jiaqing Zhang, Bosi Zhang, Hui Xie, Minghao Fan, and Liufang Wang
Ultrasonic Guided Wave Testing Method of Gun Barrel Crack
Defects Based on L (0, 2) Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 530
Jin Zhang, Xin Wang, Ying Wei, and Yang Shen
Target Re-identification Based on Dictionary Learning . . . . . . . . . . . . . 539
Gong Lianyou and Shi Guochuan
Research on High Impedance Fault Detection Method
of Distribution Network Based on S Transform . . . . . . . . . . . . . . . . . . . 547
Li Mengda
Contents xv

Research on the Protocols of VPN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554

Shuguang Zhang, Ailan Li, Hongwei Zhu, Qiaoyun Sun, Min Wang,
and Yu Zhang
Penetration Level Permission of Small Hydropower Station
in Distributed Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 560
Xin Su, Xiaotian Xu, Genghuang Yang, and Xiayi Hao
Research on Smooth Switching Method of Micro-Grid Operation . . . . . 566
Wenbin Sun, Genghuang Yang, Xin Su, and Xiaotian Xu
A New BDD Algorithm for Fault Tree Analysis . . . . . . . . . . . . . . . . . . . 573
Wei Liu, Yong Zhou, and Hongmei Xie
A Sparse Nonlinear Signal Reconstruction Algorithm
in the Wireless Sensor Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 580
Zhao Xia-Zhang and Yong-Jiang
Chinese POS Tagging with Attention-Based Long Short-Term
Memory Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 586
Nianwen Si, Hengjun Wang, Wei Li, and Yidong Shan
A New Weighted Decision Making Method for Accurate Sound
Source Localization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
Yingxiang Sun and Jiajia Chen
The Efficiency Factors of Point-to-Point Wireless Energy
Transfer System in a Closed Satellite Cavity . . . . . . . . . . . . . . . . . . . . . 600
Yue Yin, Xiaotong Zhang, Fuqiang Ma, and Tingting Zou
Personalized POI Recommendation Model in LBSNs . . . . . . . . . . . . . . . 608
Zhong Guo and Ma Changyi
Verification of CAN-BUS Communication on Robots Based
on xMAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614
Xiujuan Cao
Network Traffic Classification Using Machine Learning Algorithms . . . 621
Muhammad Shafiq, Xiangzhan Yu, and Dawei Wang
Fault Analysis and Fault Diagnosis Method Review
on Active Distribution Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 628
Zhang Tong, Liu Jianchang, Sun Lanxiang, and Yu Haibin
Extensive Survey on Networked Wireless Control . . . . . . . . . . . . . . . . . 634
Li Lanlan, Wang Xianjv, Ci Wenyan, and Chalres Z. Liew
xvi Contents

Internet and Cloud Computing

Research and Design of Smart Home System Based
on Cloud Computing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643
Huiyi Cao, Shigang Hu, Qingyang Wu, Zhijun Tang, Jin Li,
and Xiaofeng Wu
Website Structure Optimization Based on BS Matrix . . . . . . . . . . . . . . 649
Yonghong Xie, Can Cui, Aziguli, and DaoLe Li
The Effects of Bank Employees’ Information Security Awareness
on Performance of Information Security Governance . . . . . . . . . . . . . . 657
Shiann Ming Wu, Dongqiang Guo, and Yung Chang Wu
A Comparative Study on Agglomeration Effects of the Central Cities
of Three Urban Agglomerations in China– A Case Study
of Producer Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664
Qingmin Yuan, Xiao Luo, and Jian Li

Intelligent Systems
A Dynamic Node’s Trust Level Detection Scheme for Intelligent
Transportation System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671
Ge-ning Zhang
An Application of Multi-Objective Genetic Algorithm Based
on Crossover Limited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 677
Shi Lianshuan and Chen YinMei
A Method for Extracting Objects in Physics Using Unit
Semantic Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684
Yanli Wang and Pengpeng Jian
A Novel Matching Technique for Two-Sided Paper
Fragments Reassembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 691
Yi Wei, Lumeng Cao, Wen Yu, and Hao Wu
An Improved Memetic Algorithm with Novel Level Comparison
for Constrained Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 698
Xinghua Qu, Wei Zhao, Xiaoyi Feng, Liang Bai, and Bo Liu
FPGA Based Real-Time Processing Architecture for Recurrent
Neural Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 705
Yongbo Liao, Hongmei Li, and Zongbo Wang
Publication Recommendations of Manuscripts Using Improved
C4.5 Decision Tree Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 710
Didi Jia, Wu Xie, Zhiyong Chen, and Baohua Qiang
Contents xvii

Design Replica Consistency Maintenance Policy for the Oil and Gas
Pipeline Clouding SCADA Multiple Data Centers Storage System . . . . . 715
Miao Liu, Jia Shimin, and Yuan Mancang
SPSS-Based Research on Language Learning Strategy Use . . . . . . . . . . 722
Yang Xu
Study on Intelligent Monitoring of Power Transformer Based
on FPGA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 728
Fu-Sheng Li, Xin-Dong Li, Hong-Xue Bi, and Guang Jin
Mutual Fund Performance Analysis Using Nature Inspired
Optimization Techniques: A Critical Review . . . . . . . . . . . . . . . . . . . . . 734
Zeenat Afroz, Smruti Rekha Das, Debahuti Mishra,
and Srikanta Patnaik

Others
Synthesis Algorithm based on the Pre-evaluation of Quantum
Circuits for Linear Nearest Neighbor Architectures . . . . . . . . . . . . . . . . 749
Dejun Wang, Zhijin Guan, Yingying Tan, and YiZhen Wang
Study of an SIR Epidemic Disease Model with Special Nonlinear
Incidence and Computer Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . 756
Xiuchao Song, Miaohua Liu, Hao Song, and Guohong Liang
An Improvement Response Surface Method Based on Weighted
Regression for Reliability Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 761
Xingchen Yu and Zhangxue Gang
The Greedy Algorithm and Its Performance Guarantees
for Solving Maximization of Submodular Function . . . . . . . . . . . . . . . . 767
Yunxia Guo, Guohong Liang, and Jia Liu
Model of Evaluation on Quality of Graduates from Agricultural
University Based on Factor Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 772
Xuemei Zhao, Xiang Gao, Ke Meng, Xiaojing Zhou, Xiaoqiu Yu,
Jinhua Ye, Yan Xu, Hong Tian, Yufen Wei, and Xiaojuan Yu
Distributed Collaborative Control Model Based on Improved
Contract Net . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779
Zhanjie Wang and Sumei Wang
K Distribution Clutter Modeling Based on Chebyshev Filter . . . . . . . . . 785
Bin Wang and Fengming Xin
Influence of Liquid Film Thickness on Dynamic Property
of Magnetic-Liquid Double Suspension Bearing . . . . . . . . . . . . . . . . . . . 792
Zhao Jianhua, Wang Qiang, Zhang Bin, and Chen Tao
xviii Contents

Characteristics Analysis on Open-Type Liquid Hydrostatic Guideway

with Unequal Area Oil Pocket . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 800
Zhao Jianhua, Wang Qiang, Zhang Bin, and Chen Tao
Influence of Liquid Film Thickness on Static Property
of Magnetic-Liquid Double Suspension Bearing . . . . . . . . . . . . . . . . . . . 809
Zhao Jianhua, Wang Qiang, Zhang Bin, and Chen Tao
The Research on the Thinking of Large Data for Agricultural
Products Market Consumption in Beijing . . . . . . . . . . . . . . . . . . . . . . . 817
Chen Xiangyu, Gong Jing, Yu Feng, and Chen Junhong
The Application of Decision Tree in Workflow . . . . . . . . . . . . . . . . . . . 824
Lisong Wang, Yifan Chu, and Min Xu
Constructing and Analysis of the State Spaces of Workflow
Process Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 831
Lisong Wang, Yifan Chu, Min Xu, Yongchao Yin, and Ping Zhou
Analysis on the Causes of Bad PCBA Heavy Tin . . . . . . . . . . . . . . . . . . 839
Junjie Lv
Characteristics of Solder Paste and Reflow Process Analysis . . . . . . . . . 844
Junjie Lv and Xu Li
Design and Implementation of ARM7 Instruction Set Based
on GDB Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 848
Tao Yongchao, Wu Xianghu, and Qu Mingcheng
Simulation and Research on the Rotor Flux Linkage Model
of Asynchronous Motor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 854
Xiayi Hao, Genghuang Yang, Xin Su, and Xiaotian Xu
Application of Fuzzy PID Control for Oil Temperature
in Overvoltage Withstand Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 861
Qiang Fang, Xin Su, Xiaotian Xu, and Genghuang Yang
Study on High Overload Characteristics of Ultrasonic
Electric Actuator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 867
Xu Xuerong, Tian Xiu, Fu Hongwei, Wang Yanli, and Hao Yongqin
Design and Analysis of Resonator for the Resonant Accelerometer . . . . 873
Yan Li, Xi Chen, and Yunjiu Zhang
Analysis on Quality and Safety of Toys for Children—Based
on the Survey Data of Beijing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 880
Liu Xia, Liu Bisong, Wu Qian, and Li Ya
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 891
Autonomous Systems
 Prediction of Rolling Force Based on a Fusion
of Extreme Learning Machine and Self
Learning Model of Rolling Force

AZiGuLi1,2, Can Cui1,2, Yonghong Xie1,2(&),

Shuang Ha1,2, and Xiaochen Wang3
1
School of Computer and Communication Engineering, University
of Science and Technology Beijing, Beijing 100083, China
[email protected]
2
Beijing Key Laboratory of Knowledge Engineering
for Materials Science, Beijing 100083, China
3
School of Electronic Engineering, Xidian
University, Xi’an 710071, China

Abstract. Aiming at the rolling force model of hot strip rolling mill, the
forecasting technique based on ELM (extreme learning machine) is presented in
this paper. Initially, the variables associated with control rolling are identified by
the analysis of the traditional formula of rolling force, in order to ensure the
effectiveness of the model, and then apply ELM network to predict models.
Production data is applied to train and test the above network, and compare with
the modified calculated value of rolling force, which got from the self-learning
model of rolling force. The results show that the thickness can be predicted more
rapidly and exactly, which can meet the actual demand of production, when this
model and the rolling force learning model are integrated.

Keywords: ELM
Rolling force model
Self learning

1 Introduction

Today, hot rolling production has more strict requirements on the dimension of steel
strip and the precision of plate-shape, in particular for preset of finishing block. And the
precision of finishing block preset be the center of the control model of finishing block,
rolling force involves a series of domain of operation, such as device parameter, rolling
parameter of adjustment and rolling enforcement [1]. And rolling force is related to the
thickness model control, strip shape model control and temperature model control,
therefore the control of rolling force is key to hot rolling production. The traditional
rolling force is counted according to the mathematical model, and the certain physical
quantities of mathematical model is counted according to the experiment many times,
which make the accuracy of forecast not high enough.
And ELM Algorithm have absolute advantages in such sectors as precision and
predicting speed. Therefore, paper will apply the advantage of ELM to rolling force
prediction. Rolling force in the every pass set different values according to the different

© Springer International Publishing AG 2018

F. Xhafa et al. (eds.), Advances in Intelligent Systems and Interactive
Applications, Advances in Intelligent Systems and Computing 686,
https://doi.org/10.1007/978-3-319-69096-4_1
4 AZiGuLi et al.

rolled piece parameters, because the multiple passes are set on rolling force. For
precision rolling, it need to set the rolling force of seven pass, so precision rolling need
to predict the rolling force of seven pass, in turn, according to ELM [2].

2 ELM Algorithm

ELM Algorithm is a simple and effective single invisible layer feed back network
SLFNs learning algorithm, which is firstly proposed by Prof. G.B. Huang [3] in 2004.
The learning rate of the tradition feed-forward [4] neural network far less than the
actual demand, which has been caused by two reasons. The reason one is that the
slowly learning algorithm, which based on gradient, has been widely applied to the
training of neural network, the all network parameters must be repeat adjusted for each
iteration. One the other hand, SVN and transformed smallest squared SVN has applied
to binary classification due to its powerful classification capability, and the traditional
SVN can not be used directly in regression and multi-class classification, which the
problem has not been solved though having proposed different SVN variant.
ELM turned to broad SLFNs, and hidden layers of ELM need not to adjust.
Parameters of hidden nodes is not only independent of training data, but also is
independent of each others [5]. The standard feed forward neural network having
hidden nodes has the ability to the separation and common approximate. ELM features
high efficiency learning, and generalization ability is better than traditional feed for-
ward neural network [4].
Taking the typical structure as an example, as shown in Fig. 1, the network is
comprised of three parts: input layer, hidden layer and output layer.

Fig. 1. Typical structure of SLFNs

Figure 1, in the model, there are n input variable, it means that nerve cell number of
output layer is n. This indicates that the input vector of sample is x1 ; x2 ; . . .; xn , the
variable of output sample is m, this indicates that the output vector of sample is
y1 ; y2 ; . . .; yn . The connection of neighboring layers is determined by joining factor x
and b, and the linked weights matrix are formed by these factors [5].
 Prediction of Rolling Force Based on a Fusion of ELM 5

Hidden layers nodes of n and standard SLFNs math expression of active function
g(xÞ is as followed:
XN 0
i¼1
bi gðxi
xj þ bi Þ ¼ oj ; j ¼ 1; . . .; N ð2:1Þ

xi
x j inner product of vectors and xj
xi

xi weight matrix of hidden layer i and input layer,xi ¼ ½xi1 ; xi2 ; . . .; xin T ,
bi weight matrix of hidden layer i and output layer, bi ¼ ½bi1 ; bi2 ; . . .; bim T ,
bi threshold value of hidden layer i,

When formula is employed to approximate the number N of samples in zero error,

P 0  
that can be expressed as Ni¼1 oj  tj  ¼ 0 [5]. Existing xi , bi and bi makes formula
become
XN0
i¼1
bi gðxi
xj þ bi Þ ¼ tj ; j ¼ 1; . . .; N ð2:2Þ

Sample formula can be abbreviated:

Hb ¼ T ð2:3Þ

H; b and T can be expressed as two formula:

 
H x1 ; x2 ; . . .xN 0 ; b1 ; b2 ; . . .bN 0 ; x1 ; x2 ; . . .xN
2 3
gðx1 x1 þ b1 Þ gðx2 x1 þ b2 Þ . . . gðx1 x1 þ bN 0 Þ
6 gðx1 x2 þ b1 Þ gðx2 x2 þ b2 Þ . . . gðx1 x2 þ bN 0 Þ 7
6 7 ð2:4Þ
¼66 .. .. .
7
7
4. . . . . .. 5
gðx1 xN þ b1 Þ gðx2 xN þ b2 Þ . . . gðx1 xN þ bN 0 Þ NN 0
2 3 2 3
bT1 tT1
6 7 6 .7
6 .7
b ¼ 6 .. 7 ;T ¼ 6
4 .. 5
7 ð2:5Þ
4 5
bTN0 tTN
N 0 m Nm

H is output matrix of hidden layer [3], the number of hidden layer should be much
0
less than the number of sample in most cases, it means that when N  N, H is
rectangle matrix [5]. When learning algorithm, H can be calculated and can be made
6 AZiGuLi et al.

stable through given the any value of x and b, and parameter b should be determined.
0
Through formula (2.3), the least-square solution b can be get:
 0 
 
Hb  T  ¼ minkHb  T k ð2:6Þ

The minimum norm and the least-square solution can be get according to the
formula (2.3):
0
b ¼ H þT ð2:7Þ

A few important property can be get according to the generalized inverse matrix:
0
Least training error. The least-square solution of formula (2.3) is b from formula
(2.7). Although almost of learning algorithm hope to reach least training error, most of
it can not be done because of local minimum value and infinite iteration algorithm.
0
Weight minimum norm and optimal generation ability. b from formula (2.7) is
minimal norm of the least-square solution. For SLFNs, magnitude of weight is very
0
important, the smaller weight, the more powerful generation ability. It means b has
more better generation ability.
0
The minimal norm b of the least-square solution from (2.3) is unique.
The sample is trained by ELM [3], according to the random weight x and threshold
b, output matrix H can be get according to the set of hidden layers nodes and activation
function, corresponding prediction models can be get according to the formula of feed
forward network. Mainly ELM algorithm including:
Input weight x, b are given randomly, and corresponding hidden layers nodes are
get according to the real cases.
Select activation function g(X), calculate the hidden layer output matrix H [2].
0 0
Calculate the output weight b ,b ¼ H þ T.
For ELM, input weight x and threshold b can be generated at random, the nodes of
hidden layers are adjusted correspondingly to reach optimal output conclusion. This
study shows that compared with traditional BP network, ELM has the excellence of
faster learning speed, better generalization ability, and avoid the traditional algorithm
problems, including locally optimal tragedy, sensitive selection of learning rate and so
on [2].
So far, ELM has obtained approval by most researchers, and has been used in so
many fields, such as biology identification technology, image-processing, signal pro-
cessing, big data analysis, robot and internet. ELM could be used in the following
respects:
ELM can be trained in hidden layers feed forward network of widespread type, the
one layer of ELM [4] has been trained by hidden layer, which used in characteristics
learning, cluster, regression and classify. The entire network can be seen as no
adjusting ELM.
ELM can move into many local regions of multi hidden layer feed forward net-
work, and work together with other study architecture or model.
 Prediction of Rolling Force Based on a Fusion of ELM 7

Each hidden layer nodes of ELM are sub-networks formed by some nodes, and
hidden layer can form local receptive domain [3].
In each hidden layer, the connection from the input layer to hidden layer node can
be full connection, and random local connection can be generated by the connection
according to the different continuous probability distribution. For the network [5], if a
limited number of hidden layer nodes and the connection have changed, the power of
the network still can make stable.
Compared with present mainstream president algorithm, the most advantage of
ELM is that the hidden neurons of the entire learning machine can not adjust, and have
high efficiency. Some algorithms of neural network have been applied to the steel
industry, like BP, RBF and Hopfield, but for ELM, there is a little application in steel
production industry for now. In this paper, ELM is applied to the steel rolling pro-
duction, and together with other methods, hot continuous rolling exit thickness and
rolling force are used to model for guiding steel production [5].

3 Rolling Force Mathematical Model

Rolling force can be calculated through some combination of the mathematical model
and empirical data, therefore, there are a lot of formula for rolling force. For now, the
correlative formula of rolling force has focus on the SIMS formula of OROWAN, and
researchers use the formula to calculate result in different rolling force condition, the
get different deformation formula, the original rolling force formula of SIMS follows
[5]:
0
P ¼ Blc Qp KKT ð3:1Þ

P rolling force, KN,

0
lc flattened roll and horizontal projection length of strip contact arc, mm,
Qp the impact influence of stress state is caused by friction on contact arc,
K decide the inner components of mental material and the deformation condition,
mental deformation resistance about the deformation temperature, deformation
speed and the deformation degree,
KT coefficient of front and back tensions on the influence of rolling force,
B bandwidth, mm

By some of these formulas and existed relationship among all variables, corre-
sponding physical makeup of the test samples are found by the database of Beihai steel
production. Because set-up calculation of Beihai steel production, modified calculation
and adaptive learning in steel production, that some of data need be discarded and the
measurement data need be chosen as samples, when samples are chosen. And
according to the corresponding formula of rolling force, variables are chosen such as
roll speed, roll rotate, forward slide coefficient, back slide coefficient, rolling force
measured value, FET measured temperature, FDT measured temperature, exit thickness
gauge measured value (deviation), exit width gauge, roll gap measured value,
8 AZiGuLi et al.

bend-roller force measured value, roll speed measured value, median width slab, entry
thickness measured value, entry temperature, exit temperature and roll gap.

4 Rolling Force Model Self Learning of Finishing Mill

The corresponding technology parameters for the rolling force model of finishing mill
is different depending on the site environment and the equipment situation [6]. In order
to ensure the shape and thick of the production, the computer of the model need be
optimized based on the data of the actual rolling. The rolling force method is modeled
according to the analysis of the deformation resistance and the coefficient of stress
state. And the initial parameters are set, AGC control is used for model, which can
achieve continuous self learning, to adjust rolling force. As compared with self learning
of the other parameters, the biggest difference is that rolling force can not compare the
measured rolling force of the N pass steel with the original estimate rolling force.
Because bounce is a remarkable feature of mil, when measured rolling force become
the ups and downs at the change, because of the elastic deformation, the value of actual
thickness is more or less than the value of the original prediction thickness. So the
value set in rolling force and the measured value of rolling force can be get in two
different technology parameters, which can not be compared. And the self learning
model used the measured rolling force to correct the model parameters, rolling force
model recalculated the rolling force depending on correction factor, this step calcula-
tion is called correction calculation. So compared with the measured value and the
value of rolling force got from the correction calculation, the feature of the model can
be reflected.
Through the above process of the rolling force model, in production, initially, when
slab is required to roll through the first finishing mill, the measured value of rolling
force in the first pass can be get, and the parameters of rolling force model are corrected
by the model according to the measured value of rolling force in the first pass, then, the
value of rolling force is recalculated by the original model, and the parameters of the
production line in the second pass are set by the value of rolling force which was got
according to the correction calculation [8], the second roll is finished. When the second
roll is finished, the rolling force are corrected to calculate by repeating the step of the
first roll, finally, the value of the rolling force is controlled by repeating the self
learning in order to the guide the steel production.

5 Forecasting Experiment and Result Based on Rolling Force

of ELM

According to the above the finishing mill principle, initially, through the formula of the
rolling force, affecting factors about the rolling force are found and regarded as the
input variable in order to predict for the rolling force in the first pass, the measured
value of rolling force in the first pass which is got in the steel production and the
corresponding affecting data about the rolling force are regarded as the input variable
into the next ELM model in order to predict the rolling force in the second pass, by
 Prediction of Rolling Force Based on a Fusion of ELM 9

analogy, when predicting the rolling force in the seventh pass, the prediction rolling
force in the top six pass is got by the model in order to guide the production, so the
measured value of rolling force in the top six pass which is got in the steel production
and the corresponding affecting data about the rolling force are regarded as the input
variable into the next ELM [6] model in order to predict the rolling force in the seventh
pass, finally, the all of the prediction data in the seven passes are got, the multiple ELM
models superposition are realized.
First of all, divide the 522 filtrated groups of data which is related to the rolling
force into training set and test set, with 472 group as the training set, the remaining 50
group as a test set, to predict the first pass rolling force of the finishing mill. Due to the
number of nodes in hidden layer can produce certain influence on prediction accuracy,
so many experiments are needed to determine the optimal number of nodes. The
conclusion can get through the experiment, when the number of hidden layer nodes is
80, network performance achieves the best. The activation function is “sig” function.
Based on the data, the study found some dimensions of the data which is related to the
rolling force are small, while some are very large. If different dimensions are not
processed, the rate of convergence of the network will be reduced. First, process the
input data as normalization, put the rolling force data to be quantitative without the
outline, and then carries on the training and obtain the comparison between the actual
rolling force with predict rolling force model.
From the picture the most of points of first time prediction of rolling force and
actual rolling force are close, occasionally there are some larger errors. The reason of
these errors is the rolling force value is bigger, so the influence of the amplitude is
bigger. For the traditional rolling force model, the comparison with the measured
rolling force can not be done by setting calculation of rolling force. Correct calculation
of rolling force and the measured rolling force should be compared [7]. Therefore
corrected average standard deviation of the calculated rolling force and ELM prediction
model can be seen in the Fig. 2.

Fig. 2. The predicted value of the rolling force and the actual value of the rolling force for ELM
10 AZiGuLi et al.

The first mill prediction precision of rolling force is slightly higher than the pre-
cision of rolling force obtained by modified computing. Because correction calculation
of rolling force is for the guidelines to rolling mill, so the predicted rolling force can be
used in place of the correct calculation of rolling force to guide production. Because the
order of magnitude of rolling force unit is big, the experiment will increase the
numerical calculation, so orders of magnitude unit need to be downsized, and then
according to the calculation results restore the order of magnitude of the corresponding
unit. Make the first mill rolling force production actual value from ELM model with
other variables as the input of the second mill rolling force prediction model [7], the
actual value of rolling force by second mill again make the third mill modeling and so
on to get seven mills of ELM superposition model. The predicted results compare with
the measured values and get the average standard deviations, and compare with average
standard deviation fixed calculated (Table 1).

Table 1. The error value of first pass rolling

The correction calculate of rolling force The prediction of rolling force
RMSE (103) 0.083703207 0.037146671

The first mill prediction precision of rolling force is slightly higher than the pre-
cision of rolling force obtained by modified computing [8]. Because correction cal-
culation of rolling force is for the guidelines to rolling mill, so the predicted rolling
force can be used in place of the correct calculation of rolling force to guide production
[8]. Because the order of magnitude of rolling force unit is big, the experiment will
increase the numerical calculation, so orders of magnitude unit need to be downsized,
and then according to the calculation results restore the order of magnitude of the
corresponding unit. Make the first mill rolling force production actual value from ELM
model with other variables as the input of the second mill rolling force prediction
model [5], the actual value of rolling force by second mill again make the third mill
modeling and so on to get seven mills of ELM superposition model [4]. The predicted
results compare with the measured values and get the average standard deviations, and
compare with average standard deviation fixed calculated (Table 2).

Table 2. The error value of each pass rolling force for precision rolling
Pass The prediction of The correction
rolling force calculation of rolling
force
Time (s) RMSE (103) Time (s) RMSE (103)
1 0.004 0.03708315 – 0.08370321
2 0.001 0.03686208 – 0.05952837
3 0.001 0.05461632 – 0.04789749
4 0.001 0.03943099 – 0.02624345
5 0.001 0.04748756 – 0.03547987
6 0.001 0.03010246 – 0.03407099
7 0.001 0.02245528 – 0.07047431
 Prediction of Rolling Force Based on a Fusion of ELM 11

6 Conclusion

From the above experimental data, the use of ELM for continuous prediction of seven
passes in the results, the first and second mills of the rolling force prediction accuracy
compared with the correction precision of the calculation model is higher; the third to
the fifth mills of prediction precision deviation is bigger [8], the deflection prediction
accuracy is lower than those of correction calculation model, the effect is poorer. Mills
of 6 and 7 of the predicted results compared with the correction results found that 6 and
7 of the finishing passes by using machine learning to predict the effect of is better than
the effect of the modified calculation.
If all seven passes of finishing mill group through ELM rolling force prediction
model, then in the third to the fifth mills due to big errors may lead the final exported
steel does not meet the actual requirements. So according to the above experimental
results to improve the model, through the ELM and self learning of system integration
method to predict the rolling force, build the model as shown in Fig. 3. Two mills
before can use ELM overlay to predict the rolling force, and according to the predicted
rolling force to guide production [9]. The third to the fifth mills on-line correction can
through the system of self learning, Then the sixth, seven mills again apply ELM
rolling force prediction model.
Through improving model to forecast the rolling force, the time of forecast can be
greatly shortened and solve the real-time demanding of rolling technology problem,
and get the strong robustness. High accuracy model will has a broad application
prospect.

References
1. Sun, Y.K.: Model of Control Hot Strip Mill, pp. 124–163. Metallurgical Industry Press,
Beijng (2002)
2. Huang, G.B., Zhu, Q.Y., Siew, C.K.: Extreme learning machine: theory and applications.
Neuro-Computing 70, 489–501 (2006)
3. Huang, G.B., Zhu, Q.Y., Siew, C.K.: Extreme learning machine: a new learning scheme of
feed-forward neural networks. In: Proceedings of International Joint Conference on Neural
Networks (IJCNN 2004), pp. 985–990 (2004)
4. Lonen, J., Kamarainen, J.-K., Lampinen, J.: Differential evolution training algorithm for feed-
forward neural networks. Neural Process. Lett. 7(1), 93–105 (2003)
5. Huang, G.B., Chen, L., Siew, C.K.: Universal approximation using incremental constructive
feed-forward networks with random hidden nodes. IEEE Trans. Neural Netw. 17(4), 879–892
(2006)
6. Huang, G.B., Chen, L.: Convex incremental extreme learning machine. Neurocomputing 70,
3056–3062 (2007)
7. Xie, J., Jiang, S., Xie, W., et al.: An efficient global K-means clustering algorithm. J. Comput.
6(2), 271–279 (2011)
8. Al-Zoubi, M.B., Hudaib, A., Huneiti, A., et al.: New efficient strategy to accelerate k-means
clustering algorithm. Am. J. Appl. Sci. 5(9), 1247–1250 (2008)
9. Gaemperle, R., Mueller, S.D., Koumoutsakos, P.: A parameter study for differential evolution.
In: Grmela, A., Mastorakis, N.E. (eds.) Advances in Intelligent Systems, Fuzzy Systems,
Evolutionary Computation, pp. 293–298. WSEAS Press (2002)
 User Behavior Profiles Establishment
in Electric Power Industry

Wei Song1, Gang Liu1, Di Luo2(&), and Di Gao1

1
State Grid Jibei Electric Power Science Research Institute, Nanjing, China
2
School of Computer and Communication Engineering, University of Science
& Technology Beijing, Beijing, China
[email protected]

Abstract. The big data of electric power industry contains the information of
users’ values, credits, and behavior preferences. On the basis of the data, electric
power companies can provide personal services as well as increasing profits. In
this paper, we proposed a labeling system based on the clustering algorithms and
Gradient Boost Decision Tree (GBDT) algorithm to establish user behavior
profiles for State Grid Group of China, including basic information labels,
behavior labels, behavior description labels, behavior prediction, and user
classification. The experimental results showed that the approach can describe
the behavior features of users in the electric power industry effectively.

Keywords: Behavior profile
Behavior prediction
Clustering
Gradient

boost decision tree
Electric power industry

1 Introduction

There are many kinds of behaviors in electric power industry including electricity
consuming, paying, capacity changing, illegally electricity using, etc. Although the
electric power companies have a lot of users’ behavior data, they are numerical and
hard to understand directly. Some rules are needed to turn the initial numerical data to
semantic labels which could easily illustrate the users’ feature. Here are some of the
labels: ‘Zhangjiakou’, ‘Low monthly consumption’, ‘capacity change frequently’,
‘Recover from capacity suspend every 7 months’, ‘High risk of arrearage’, ‘Low user
value’. These labels are called user profiles. Profiling the users is the basis of providing
the personal services, and its applications in the electric power industry are various.
First, companies can select users by labels, which are more detailed and semantic. Also,
it’s easy to recognize the occurrence of the user behaviors so that companies can easily
monitor the user behaviors and give advise. At the same time, as the companies handle
the feature of the user behaviors, precision marketing will be reasonable and the
companies can arrange business better. What the most important is, according to the
user profiles of electric power industry, the companies can get the semantic credit
report of the users. User profile is commonly used in IT industry, but we can not build
the user profiles in the electricity industry using the same approach in the IT industry

© Springer International Publishing AG 2018

F. Xhafa et al. (eds.), Advances in Intelligent Systems and Interactive
Applications, Advances in Intelligent Systems and Computing 686,
https://doi.org/10.1007/978-3-319-69096-4_2
Exploring the Variety of Random
Documents with Different Content
mounted his new courser, and rode several times round the camp,
shouting in triumph, and putting his horse through all his paces.
The principal wealth of the savages of the west consists in horses, of
which each chief and warrior possesses a great number, that may be
seen grazing about their camp. The horses of the Crows are
principally of the Maroon race of the prairies.[134] They have also
many horses which they have stolen from the Scioux, the
Sheyennes, and other Indians of the south-west, which they had in
their turn stolen from the Spaniards of Mexico. The Crows are
considered the most indefatigable marauders of the desert; they
traverse the mountains in all directions, bringing to one side what
they have taken at the other. The name of Atsharoke, or Crow, has
been given to them on account of their robberies.[135] They are
practised from their infancy in this sort of larceny, and they acquire a
surprising dexterity in it; their glory augments with the number of
their captures, so that a finished robber is in their eyes a hero. I
accompanied for two days, these savages, who, I think, were the
finest Indians I had met in all my travels. They passed the whole
time in rejoicings and feasting. You will not be scandalized, I trust,
when I tell you that I was present at twenty different banquets. I
was scarcely seated in one cabin, when I was called to partake of
the festive entertainment in another.
We arrived, at last, at the first fort belonging to the Fur Company.
The Americans, who have here a trading post, received us most
cordially. At this place I was to part with 52 my faithful Flat Heads. I
therefore told them, that, having before me a country still more
exposed to the incursions of the Black Feet, the Assiniboins, the Big
Bellies, the Arikaras, and Scioux, all of whom are declared enemies
of their tribe, I would no longer peril their lives, on account of my
personal safety; that as for my life, I placed it in the hands of God,
and that I felt a persuasion it would be preserved, in order that,
accompanied by new Missionaries, I might immediately return to
them. I exhorted them for the last time to remain faithful to the
Great Spirit. We embraced each other, wishing, mutually, a happy
return; and shortly after, accompanied by my faithful Fleming, I
disappeared from their sight amidst the solitary defiles. We were to
pass over several hundred miles of country, where no road is yet
traced, and, like the navigator on the boundless ocean, with no
other guide than the compass. For a long time we followed the
course of the Yellow Stone, except when perpendicular rocks
arrested our progress and obliged us to take a circuit. In many
places we discovered forts which the savages are in the habit of
raising for defence, or for concealing themselves, when they are at
war, or waiting for their prey. Perhaps, at the moment of our
passing, they were not without enemies. What a solitude, with its
horrors and dangers! but it possesses one real advantage: with
death constantly before our eyes, we irresistibly feel, without the
possibility of illusion, that we are entirely in the hands of God,
without any support but Him, without any other refuge than his
paternal providence; it is then easy to make to Him the sacrifice of a
life which belongs less to us than to the first savage who wishes to
take it, and to form the most generous resolutions of which man is
capable. It was really the best spiritual retreat that I made in my life.
The second day of the journey, on awaking, I perceived, at the
distance of a quarter of a mile, the smoke of a great 53 fire—a point
of a rock was all that separated us from a detachment of Indians.
Without a moment's delay we saddled our horses and set off,
galloping with all speed along the ravines and beds of dried up
torrents. We rode that day, without resting, more than fifteen
leagues, and we did not encamp until two hours after sunset, lest
the savages, having observed our track, should think of pursuing us.
—The same fear prevented us from lighting a fire, which obliged us
to dispense with supper. I wrapped myself in my blanket, stretched
myself on the grass beside my companion, and having
recommended myself to God, I endeavored to beguile hunger by
sleep. My grenadier, more courageous than I, soon snored like a
steam engine in full play.
The next morning we were on our way at day-break; we advanced
with caution, for the country appeared full of danger. Towards mid-
day we met a new subject of alarm—we found a buffalo, which had
been killed about two hours previously. We thrilled at the sight,
when we thought that the enemy was not far off; and yet we had
reason to thank the Lord for having prepared the food for our
evening meal. The following night we encamped among rocks, which
are the retreat of tigers and bears. I have already said that the dens
of the wild beasts inspire incomparably less terror to the traveller
than the hut of the savage. I this time slept heavily and well. We
always commenced our journey early in the morning, and each day
had new dangers to face, and to meet occasionally the fresh traces
of men and horses. One day we had to cross a field of tents, which
had been recently abandoned; the fires were not quite extinguished;
but happily we met no one. At length we saw again the Missouri at
the very place, where an hour before, a hundred families of the
Assiniboins had passed over it. The foregoing is only a sketch of the
54 long and perilous journey which we made from the fort of the
Crows to fort Union, situated at the mouth of the Yellow Stone river.
[136]

All the country watered by this river abounds in game; I do not think
that there is in all America another place better suited for hunting:
we were continually amidst vast herds of buffalos; we frequently
discovered groups of majestic elks bounding over the plains, whilst
clouds, if I may say so, of antelopes were flying before us with the
swiftness of the wind. The Ashata, or Big Horn, alone appeared not
to be disturbed at our presence: we saw them in groups, reposing
on the edges of the precipices, or sporting on the points of the steep
rocks. The black-tailed roebuck, so richly dressed in its brown coat,
frequently excited our admiration, by its elegant shape, and abrupt,
animated movements, in which it appears scarcely to touch the
earth with its feet.[137] I have already spoken of the grizzly bears,
which are here to be met with in abundance, as well as the wolves,
panthers, badgers and wild cats. Often the traveller sees the prairie
hen and the cock of the mountain start up from the midst of the
heath. The lakes and rivers are covered with swans, geese and
ducks: the industrious beaver, the otter, and the muskrat, together
with the fishes, are in peaceable possession of their solitary waters.
The Arikaras and the Big Bellies, who had been described to us as
most dangerous, received us as friends, whenever we met them on
our way. Before setting out for war, they observe a strict fast, or
rather they abstain from all food for four days. During this interval
their imagination is excited to madness; and, either from the effect
of weakness, or the warlike projects which fill their minds, they
pretend that they have extraordinary visions. The elders and sages
of the tribe are called upon to interpret these reveries; 55 and they
pronounce them to be more or less favorable to the undertaking.
Their explanations are received as oracles, according to which the
expedition is scrupulously regulated. Whilst the preparatory fast
endures, the warriors make incisions in their bodies, and bury in the
flesh, under the shoulder-blade, pieces of wood, to which they
attach leather thongs, by which they are suspended from a stake,
fixed horizontally over the brink of a chasm a hundred and fifty feet
deep. They even sometimes cut off one or two fingers, which they
offer as a sacrifice to the Great Spirit, in order that they may return
loaded with scalps.[138]
In a recent expedition against the Scioux, the Arikaras killed twenty
warriors of the hostile tribe, and piled up the corpses in the middle
of their village. The solemn dance of victory then commenced, at
which men, women, the aged, and children assisted. After having
celebrated, at length, the exploits of the brave, they rushed, like wild
beasts, upon the mangled and bloody bodies of the Scioux, parcelled
them amongst themselves, and fixed the hideous trophies to the end
of long poles, which they carried in proud triumph around the
village.
It is impossible to form an idea of the cruelty that presides over the
barbarous revenge of those tribes, who are constantly occupied in
mutual destruction. As soon as the savages learn that the warriors of
a rival nation have set out for the chase, they unexpectedly attack
the enemy's defenceless camp, and massacre the women, old men,
and children in the cradle. Wo to the men who are spared; their
agony is deferred in order to render it more terrible. At other times
they lie in wait in their enemy's path, and allow the detachment to
pass on, until they have in their power such a portion of it as must
infallibly become their 56 prey; whereupon they raise the death cry,
and pour upon the enemy a shower of balls, arrows, and pieces of
rock; this movement is the signal of extermination: the battle
becomes a massacre: the sights of horror which would freeze the
heart of any civilized man, serve only to inflame the fury of the
savage: he outrages his prostrate rival, tramples on his mangled
carcass, tears off his hair, wallows in his blood with the delight of a
tiger, and often devours the quivering limbs of the fallen, while they
have scarcely ceased to exist.
Such of the vanquished as have not fallen in the combat are
reserved to adorn the triumph, and are conducted prisoners to the
village of the conquerors. The women come to meet the returning
warriors, amongst whom they seek with anxious looks their
husbands and brothers: if they discover them not, they express their
grief by terrific howling. One of the warriors soon commands silence;
he then gives the details of the fortunate expedition; describes the
place selected for the ambuscade, the consternation of the waylaid
tribe, the bravery of the assailants, and recounts the number of the
dead and of the captives. To this recital, which is made with all the
intoxication of victory, succeeds the calling over the names of the
warriors: their absence tells they are no more. The piercing cries of
the women are then renewed; and their despair presents a scene of
frenzy and grief, which exceeds all imagination. The last ceremony is
the proclaiming of victory. Every one instantly forgets his own
misfortunes; the glory of the nation becomes the happiness of all; by
an inconceivable transition, they pass in a moment from frantic grief
to the most extravagant joy.
I know not what terms to use in order to describe the torments
which they inflict on the wretched prisoners: one 57 plucks off their
nails, another tears away their flesh; red hot irons are applied to
every part of their bodies; they are flayed alive, and their palpitating
flesh is devoured as food.[139] The women, who, in other nations,
are more accessible to the feelings of pity than the men, here shew
themselves more thirsty for revenge, and more ingenious in the
barbarous refinement of cruelty. Whilst this horrible drama goes on,
the chiefs are gravely seated about the stake at which the victim is
writhing. The latter appears to be only intent on conquering his
anguish: often has the prisoner been seen to brave his executioners,
and with a stoic coolness exclaim, "I fear not death; those who are
afraid of your torments are cowards; a woman of my tribe would
despise them. Shame upon my enemies; they have not even the
power to force from me a tear. In order to take me, they supplied
their weakness by strategy; and now, to revenge themselves, they
have assembled an entire people against one man, and they are
unable to triumph over him—the cowards! Oh, if they were in my
place, how I would devour them, how I would sip from their
accursed skulls the last drop of their blood!"
The great village of the Arikaras is only ten miles distant from that of
the Mandans. I was surprised to see around their habitations large
and well cultivated fields of maiz. The latter Indians still manufacture
earthen vases,[140] similar to those which are found in the ancient
tombs of the savages of the United States, and which, according to
antiquaries, are presumed to have belonged to a race much more
ancient than that which now peoples the desert of the west. The
jugglers of the Arikaras enjoy a good reputation, and exercise
considerable influence over their credulous countrymen; they
pretend to have communication with the spirit 58 of darkness.[141]
They will fearlessly plunge their arm into boiling water, having
previously rubbed it with a certain root; they also swallow, without
any ill effect, substances on fire, as well as shoot arrows against
themselves. The following is one of the most singular of their tricks,
and one which the Indian sorcerer was unwilling to perform in my
presence, because my medicine (meaning my religion) was superior
to his. He had his hands, arms, legs, and feet, tied with well-knotted
cords; he was then enclosed in a net, and again in a buffalo's skin.
The person who tied him had promised him a horse if he extricated
himself from his bonds. In a minute after, the savage, to the
amazement of the spectators, stood before him perfectly free. The
commandant of the neighbouring fort offered him another horse, if
he would reveal to him his secret. The sorcerer consented, saying,
"Have thyself tied; I have at my command ten invisible spirits: I will
detach three of them and put them at thy service: fear them not,
they will accompany thee everywhere, and be thy tutelary genii."
The commandant was disconcerted, or unwilling to make the trial,
and thus the matter terminated.[142]
The last observation which I have to make concerns the redoubtable
tribe of the Scioux. Whoever, amongst these savages, dies in a
quarrel provoked by drunkenness, or as 59 the victim of the revenge
of a fellow countryman, receives not the ordinary honours of burial;
he is interred without ceremony and without provisions. The most
glorious death for them is to expire in fighting the enemies of their
nation. Their bodies are, in that case, rolled in buffaloes' skins and
placed upon a raised platform, near their camps or highways.[143]
From some conversations I have had with the chiefs of this tribe, I
have every reason to believe that a mission would produce amongst
them the most consoling effects.
I arrived, at length, at Council Bluffs. It would be vain for me to
attempt to express what I felt, on finding myself again in the midst
of my brethren: I had travelled two thousand Flemish leagues
amongst the most barbarous nations, where I had no sooner
escaped one danger than I met with another. From Council Bluffs to
Westport, a frontier city of the Missouri, I pursued my journey
without obstacle or accident. At Independence,[144] I took the public
conveyance, and on the eve of the new year, I embraced my dear
Fathers of the University of St. Louis.
Recommending myself to your prayers,
I am yours, &c.

P. J. De Smet.
 LETTER III
Banks of the Platte, 2d June, 1841.
Rev. and Very Dear Father Provincial:
Behold us at last on our way towards the long wished for "Rocky
Mountains," already inured to the fatigues of the journey and full of
the brightest hopes. It is now afternoon and we are sitting on the
banks of a river, which, it is said, has not its equal in the world. The
Indians call it Nebraska or Big Horn; the Canadians give it the name
of la Platte, and Irving designates it as the most wonderful and
useless of rivers. The sequel will show that it deserves these various
affixes. It was to enjoy the freshness and beauty of its scenery that
we travelled more than twenty miles this morning, without breaking
our fast, through a wilderness without a single rivulet to water our
jaded horses, who must therefore rest where they are till to-morrow.
I am far from regretting the delay as it will give me an opportunity
of commencing a letter which, I know, will interest you.
Like all the works of God, our humble beginnings have not been
unattended with trials: our journey had even well nigh been
indefinitely postponed by the unexpected non-arrival of two caravans
on which we had confidently relied; one of hunters, for the American
Fur Company; the other an exploring expedition belonging to the
United States, at the head of which we expected to see the
celebrated M. Nicolet.[145] Happily God inspired two estimable
travellers, 61 of whom more hereafter, and afterwards sixty others,
to take the same route as ourselves, some for health, others for
science, or pleasure; but the greater number to seek their fortune in
the too highly boasted land of California. This caravan formed an
extraordinary mixture of different nations, every country of Europe
having in it a representative, my own little band of eleven persons
hailing from eight.[146]
The difficulties of setting out once overcome, many others followed
in succession. We had need of provisions, fire-arms, implements of
every kind, waggons, guides, a good hunter, an experienced captain,
—in a word, whatever becomes necessary when one has to traverse
a desert of eight hundred leagues, and expects nothing but
formidable obstacles to surmount, and thieving, and sometimes
murderous, enemies to combat,—and swamps, ravines and rivers to
cross, and mountains to climb, whose craggy and precipitous sides
suddenly arrest our progress, compelling us to drag our beasts of
burden up their steep ascents. These things are not done without
toil and money, but thanks to the generous charity of our friends in
Philadelphia, Cincinnati, Kentucky, St. Louis and New Orleans,[147]
which place I visited in person and which is always at the head of
the others when there is a question of relieving the necessities of
the poor, or showing compassion and munificence to any who may
be in need of assistance, we were enabled by the resources thence
supplied, and by a portion of the funds allowed by the Lyons
Association in behalf of the Indian Missions, to undertake this long
journey.
You have already learned from my letters of the past year, that I was
specially sent among the Flat Heads to ascertain their dispositions
towards the "Black Robes," whom they had so long desired. I
therefore started from 62 St. Louis in April, 1840, and arrived on the
banks of the Colorado precisely at the moment when a band of Flat
Heads reached that point on their way to meet me. It was the
rendezvous I had given them. Besides the Flat Heads I visited during
that journey, many other tribes, such as the Pends-d'oreilles (Ear
Rings), Nez Perces (Pierced Noses), Cheyennes, Serpents, Crows,
Gros ventres or Minatarees, Ricaras, Mandans, Kanzas, the
numerous nations of the Scioux, &c. Finding every where such good
dispositions, I resolved, notwithstanding the approach of winter and
frequent attacks of fever, in order to second the visible designs of
the divine mercy in favor of so many souls, to commence my journey
across the immense ocean of mountains and prairies. I have
travelled without any other guide than a compass, without any
protection from nations hostile to the whites, but a veteran from
Ghent, formerly a grenadier of the Empire, any other provisions in
an arid desert, than what powder and ball and a strong confidence
in God might procure us. I shall not here repeat what I have already
communicated to you, of my adventures and the result of this
mission. It will suffice to say, that the unexpected quickness of my
return to St. Louis, the excellent health I enjoyed, even though it
was the midst of winter, and the consoling accounts I had to give of
my reception by the Flat Heads, &c. &c., all contributed to make the
most lively impression on the hearts of our brethren. Almost every
one thought himself called to share the labors of a mission which
offered so many attractions to their zeal. After due deliberation, the
fellow-laborers allotted me were five in number, namely two Fathers,
Rev. Mr. Point[148] of La Vendee, as zealous and courageous for the
salvation of souls as his compatriot, La Roche Jacquelin[149] was in
the service of his lawful sovereign; Rev. Mr. Mengarini, recently from
63 Rome, specially selected by the Father General himself, for this
mission, on account of his age, his virtues, his great facility for
languages and his knowledge of medicine and music;[150] and three
lay-brothers, two Belgians, Claessens and Huet, and one German, of
whom the first is a blacksmith, the second a carpenter, and the third
a tinner, or a sort of factotum;[151] all three industrious, devoted to
the Missions and full of good will. They had long ardently desired to
be employed on these missions and I thank God that had the choice
been left to myself, I could have made none better. Thus launched
into the midst of this interminable Far West, how often did I repeat
these beautiful lines of Racine:
O Dieu, par quelles routes inconnues aux mortels
Ta Sagesse conduit tes desseins eternels!

In seven days from my departure from St. Louis, namely, on the

30th of April, I arrived at Westport, a frontier town on the West of
the United States. It took us seven days, on board a steamboat,[152]
to perform this journey of 900 miles, no unfair average of the time
required to travel such a distance on the Missouri, at the breaking up
of the winter, when, though the ice is melted, the water is still so
low, the sand banks so close together and the snags so numerous
that boats cannot make greater headway.... We landed on the right
bank of the river, and took refuge in an abandoned little cabin,
where a poor Indian woman had died a few days before, and in this
retreat, so like to that which once merited the preference of the
Saviour and for which was thenceforth to be substituted only the
shelter of a tent in the wilderness, we took up our abode until the
10th May—occupied as well we might be in supplying the wants
created by the burning of our baggage waggon on board the
steamboat, the sickness of one of our horses 64 which we were
compelled to leave after us, and the loss of another that escaped
from us at the moment of landing.
We started, then, from Westport, on the 10th of May, and after
having passed by the lands of the Shawnees and Delawares, where
we saw nothing remarkable but the college of the Methodists,[153]
built, it is easy to divine for what, where the soil is richest; we
arrived after five day's march on the banks of the Kanzas river,
where we found those of our companions, who had travelled by
water, with a part of our baggage.[154] Two of the relatives of the
grand chief had come twenty miles from that place to meet us, one
of whom helped our horses to pass the river in safety, by swimming
before them, and the other announced our arrival to the principal
men of the tribe who waited for us on the opposite bank. Our
baggage, waggons and men crossed in a pirogue, which, at a
distance, looked like one of those gondolas that glide through the
streets of Venice. As soon as the Kanzas understood that we were
going to encamp on the banks of the Soldier's River,[155] which is
only six miles from the village, they galloped rapidly away from our
Caravan, disappearing in a cloud of dust, so that we had scarcely
pitched our tents when the great Chief presented himself with six of
his bravest warriors, to bid us welcome. After having made me sit
down on a mat spread on the ground, he, with much solemnity, took
from his pocket a Portfolio containing the honorable titles that gave
him a right to our friendship and placed them in my hands. I read
them, and having, with the tact of a man accustomed to the
etiquette of savage life, furnished him the means of smoking the
Calumet, he made us accept for our guard the two braves who had
come to meet us. Both were armed like warriors, one carrying a
lance and a buckler, and the other a bow and arrows, with a naked
sword and a collar 65 made of the claws of four bears which he had
killed with his own hand. These two braves remained faithful at their
post during the three days and three nights that we had to wait the
coming up of the stragglers of the caravan. A small present which
we made them at our departure, secured us their friendship.

Kanza Village
On the 19th we continued our journey to the number of seventy
souls, fifty of whom were capable of managing the rifle—a force
more than sufficient to undertake with prudence the long march we
had to make. Whilst the rest of our company inclined to the West,
Father Point, a young Englishman and myself turned to the left, to
visit the nearest village of our hosts.[156] At the first sight of their
wigwams, we were struck at the resemblance they bore to the large
stacks of wheat which cover our fields in harvest-time. There were
of these in all no more than about twenty, grouped together without
order, but each covering a space of about one hundred and twenty
feet in circumference, and sufficient to shelter from thirty to forty
persons. The entire village appeared to us to consist of from seven
to eight hundred souls—an approximation which is justified by the
fact that the total population of the tribe is confined to two villages,
together numbering 1900 inhabitants. These cabins, however
humble they may appear, are solidly built and convenient. From the
top of the wall, which is about six feet in height, rise inclined poles,
which terminate round an opening above, serving at once for
chimney and window. The door of the edifice consists of an
undressed hide on the most sheltered side, the hearth occupies the
centre and is in the midst of four upright posts destined to support
the rotunda; the beds are ranged round the wall and the space
between the beds and the hearth is occupied by the members 66 of
the family, some standing, others sitting or lying on skins, or yellow
colored mats. It would seem that this last named article is regarded
as a piece of extra finery, for the lodge assigned to us had one of
them.[157]
It would be difficult to describe all the curiosities we beheld during
the hour we passed among these truly strange beings; a Teniers
would have envied us. What most excited our attention was the
peculiar physiognomy of the greater number of these personages,
their vivacity of expression, singular costume, diversity of
amusement and fantastic attitudes and gestures. The women alone
were occupied, and in order to attend to their various duties with
less distraction, they had placed those of their papooses who were
unable to walk, on beds or on the floor, or at their feet, each tightly
swathed and fastened to a board, to preserve it from being injured
by surrounding objects. This machine, which I shall not call either
cradle or chair, is carried, when they travel, either on the back, after
the fashion of the gypsies and fortune-tellers in Europe, or at their
side, or more frequently, suspended from the pummel of the saddle,
while they lead or drive their ponies, laden with the rest of their
goods and chattels. With such encumbrances they manage to keep
pace with their husbands, who generally keep their horses at a
gallop. But let us return to our wigwam. How were the men
occupied? When we entered, some were preparing to eat, (this is
their great occupation when they are not asleep) others were
smoking, discharging the fumes of the tobacco by their mouths and
nostrils, reminding one of the funnels of a steamboat; they talked,
they plucked out their beard and the hair of their eye-brows, they
made their toilette; the head receiving particular attention. Contrary
to the custom of the other tribes, who let the hair on their heads
grow, (one of 67 the Crows has hair eleven feet long) the Kanzas
shave theirs, with the exception of a well curled tuft on the crown,
destined to be wreathed with the warrior's plume of eagle's feathers,
the proudest ornament with which the human head can be adorned.
[158] While we were smoking I could not help watching the motions
of a young savage, a sort of dandy, who ceased not to arrange, over
and over again, his bunch of feathers before a looking glass,
apparently unable to give it the graceful finish he intended.—Father
Point, having suffered his beard to grow, soon became an object of
curiosity and laughter, to the children—a beardless chin and well
picked brows and eye-lashes being, among them, indispensable to
beauty. Next come the Plume and Slit-ears, with their pendants of
beads and other trinkets. This is but a part of their finery, and the
pains thus taken to reach the beau-ideal of personal decorations, are
but a faint specimen of their vanity. Do you wish to have an idea of a
Kanza satisfied with himself in the highest degree? Picture him to
yourself with rings of vermillion encircling his eyes, with white, black,
or red streaks running down his face, a fantastic necklace, adorned
in the center with a large medal of silver or copper, dangling on his
breast; bracelets of tin, copper, or brass, on his arms and wrists; a
cincture of white around his waist, a cutlass and scabbard,
embroidered shoes or mocasins on his feet; and, to crown all, a
mantle, it matters not for the color, thrown over the shoulders and
falling around the body in such folds or drapery as the wants or
caprice of the wearer may direct, and the individual stands before
you as he exhibited himself to us.
As for dress, manners, religion, modes of making war, &c., the
Kanzas are like the savages of their neighborhood, with whom they
have preserved peaceful and friendly relations 68 from time
immemorial. In stature, they are generally tall and well made. Their
physiognomy is manly, their language is guttural, and remarkable for
the length and strong accentuation of the final syllables. Their style
of singing is monotonous, whence it may be inferred that the
enchanting music heard on the rivers of Paraguay, never cheers the
voyager on the otherwise beautiful streams of the country of the
Kanzas.
With regard to the qualities which distinguish man from the brute,
they are far from being deficient. To bodily strength and courage
they unite a shrewdness and address superior to other savages, and
in their wars or the chase, they make a dexterous use of fire arms,
which gives them a decided advantage over their enemies.
Among the chiefs of this tribe are found men really distinguished in
many respects. The most celebrated was "White Plume," whom the
author of the Conquest of Grenada represents as a man of great
powers of mind and chivalrous character.[159] He was endowed with
uncommon intelligence, frankness, generosity and courage. He had
been particularly acquainted with Rev. Mr. De la Croix, one of the
first Catholic Missionaries that visited that part of the West, and
conceived for him and his colleagues, the "Black Robes" profound
esteem.[160] His feelings towards the Protestant Missionaries were
far different. He had neither esteem nor veneration for them or their
reformation. When on a certain occasion one of them spoke to him
of conversion; "conversion," said the unsophisticated savage, "is a
good thing when the change is made for something good. For my
part, I know none such but what is taught and practised by the
Black Robes. If then you desire me to change, you must first quit
your wife and then put on the habit I shall show you, and then we
shall 69 see further." This habit was a priest's cassock, which a
missionary had left him with the memory of his virtues.—We
presume we need not add that these hard conditions were not
complied with by the preacher.
It is not to be inferred from the apparent pleasantry of this remark
that the chief spoke lightly of Religion; on the contrary, the Kanzas,
like all the Indian tribes, never speak on the subject without
becoming solemnity. The more they are observed the more evident
does it become that the religious sentiment is deeply implanted in
their souls, and is, of all others, that which is most frequently
expressed by their words and actions. Thus, for instance, they never
take the calumet, without first rendering some homage to the Great
Spirit. In the midst of their most infuriate passions they address him
certain prayers, and even in assassinating a defenceless child, or a
woman, they invoke the Master of life. To be enabled to take many a
scalp from their enemies, or to rob them of many horses, becomes
the object of their most fervid prayers, to which they sometimes add
fasts, macerations and sacrifices. What did they not do last spring,
to render the heavens propitious? And for what? To obtain the
power, in the absence of their warriors, to massacre all the women
and children of the Pawnees! And in effect they carried off the scalps
of ninety victims, and made prisoners of all whom they did not think
proper to kill. In their eyes, revenge, far from being a horrible vice,
is the first of virtues, the distinctive mark of great souls, and a
complete vindication of the most atrocious cruelty. It would be time
lost to attempt to persuade them that there can be neither merit,
nor glory, in the murder of a disarmed and helpless foe. There is but
one exception to this barbarous code, it is when an enemy
voluntarily seeks a refuge in one of their villages. As long as 70 he
remains in it, his asylum is inviolable—his life is more safe than it
would be in his own wigwam. But wo to him if he attempt to fly—
scarcely has he taken a single step, before he restores to his hosts
all the imaginary rights which the spirit of vengeance had given
them to his life! However cruel they may be to their foes, the Kanzas
are no strangers to the tenderest sentiments of piety, friendship and
compassion. They are often inconsolable for the death of their
relations, and leave nothing undone to give proof of their sorrow.
Then only do they suffer their hair to grow—long hair being a sign of
long mourning. The principal chief apologised for the length of his
hair, informing us, of what we could have divined from the sadness
of his countenance, that he had lost his son. I wish I could represent
to you the respect, astonishment and compassion, expressed on the
countenances of three others, when they visited our little Chapel for
the first time.[161] When we showed them an "Ecce Homo" and a
statue of our Lady of the seven Dolours, and the interpreter
explained to them that that head crowned with thorns, and that
countenance defiled with insults, were the true and real image of a
God who had died for the love of us, and that the heart they saw
pierced with seven swords, was the heart of his mother, we beheld
an affecting illustration of the beautiful thought of Tertullian, that the
soul of man is naturally Christian! On such occasions, it is surely not
difficult, after a short instruction on true faith and the love of God, to
excite feelings of pity for their fellow creatures in the most ferocious
bosoms. What were the Iroquois before their conversions, and what
have they not since become? Why do the Kanzas and so many other
tribes on the confines of civilization, still retain that savage ferocity
of manners? Why have the great sums expended in their behalf by
Protestant philanthropy 71 produced no satisfactory results? Why are
the germs of civilization so thickly scattered among these tribes, as it
were, stricken with sterility? Ah! it is doubtless, because something
more than human policy and zeal of Protestantism is necessary to
civilize the savages and make them Christians. May the God of
Mercies, in whom we alone place all our trust, bless our undertaking
and enable us to predict that our sweat, mixed with the fertilizing
dew of heaven, will fall auspiciously on this long barren earth, and
make it produce something else besides briars and thorns! When we
took leave of our hospitable hosts, two of their warriors, to one of
whom they gave the title of Captain, escorted us a short distance on
the road, which lay through a vast field which had been cleared and
planted for them by the United States, but which had been ravaged
before the harvest home—sad proof of what we have stated above.

Interior of a Kanza Lodge

Our escort continued with us until the day following, and would have
remained with us still longer, did they not fear the terrible reprisals
of the Pawnees, for the massacre committed some months
previously. Having therefore received our thanks and a portion of
tobacco, they resumed the road to their village, just in time to
escape the vengeance of a party of Pawnees, whom we met two
days later, in quest of the Kanzas!
The Pawnees are divided into four tribes, scattered over the fertile
borders of the Platte River.[162] Though six times more numerous
than the Kanzas, they have almost on every occasion been
conquered by the latter, because they are far inferior to them in the
use of arms, and in strength and courage. Yet as the party just
mentioned seemed to have adopted decisive measures, and as their
thirst of revenge had been stimulated to the highest degree by the
still fresh recollection of what their mothers, their wives and children
72 had suffered, we had reason to fear for the Kanzas. Already we
fancied that we saw the blood streaming on all sides, when, two
days after we had passed them, we saw them return to meet us.
The two first who approached us, excited our attention, the one by a
human scalp, which hung suspended from the neck of his horse, the
other by an American flag, which he had wrapped around his body,
in the form of a cloak. This kind of attire made us tremble for the
fate of our hosts; but the captain of the caravan having asked them
by signs concerning the result of their expedition, they informed us
that they had not even seen the enemy, and that they suffered much
from the cravings of hunger. We gave to them, and to about fifteen
others who followed them, both victuals and tobacco. They
devoured the victuals, but did not smoke; and, contrary to the
custom of the Indians, who generally expect to get a second meal
after the first, they left us in a manner which indicated that they
were dissatisfied. The suddenness of their departure, their refusal to
smoke the calumet, the unexpected return of their party, the
neighborhood of their villages, and their well known love of plunder
—in short, every thing induced us to fear that they had some design
to make an attempt, if not upon our persons, at least upon the
baggage; but, God be praised, not one re-appeared after the
departure of the party.
Though addicted to the practice of lying and stealing, yet, what must
appear wonderful, the Pawnees are in some respects true believers,
with regard to the certainty of a future life, and display a pharisaical
punctuality in the observance of their superstitious rites. Dancing
and music, as well as fasting, prayer and sacrifice, form an essential
part of their worship. The most common worship among them is
that which they offer to a stuffed bird, filled with 73 herbs and roots,
to which they attribute a supernatural virtue.[163] They protest that
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com