See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/323379498

FPGA-based online power quality monitoring

system for electrical distribution network

Article in Measurement · February 2018

DOI: 10.1016/j.measurement.2018.02.058

CITATIONS READS

0 57

6 authors, including:

Özal yıldırım Hüseyin Erişti

Munzur University Mersin University
23 PUBLICATIONS 56 CITATIONS 18 PUBLICATIONS 270 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

harmonic View project

Power quality monitoring systems View project

All content following this page was uploaded by Özal yıldırım on 06 April 2018.

The user has requested enhancement of the downloaded file.

 Measurement 121 (2018) 109–121

Contents lists available at ScienceDirect

Measurement
journal homepage: www.elsevier.com/locate/measurement

FPGA-based online power quality monitoring system for electrical T

distribution network
⁎
Ozal Yildirima, , Bekis Eristib, Huseyin Eristic, Sencer Unald, Yavuz Erold, Yakup Demird
a
Computer Engineering Department, Engineering Faculty, Munzur University, Tunceli, Turkey
b
Electrical and Energy Department, Vocational School of Technical Sciences, Mersin University, Mersin, Turkey
c
Electrical and Electronics Engineering Department, Engineering Faculty, Mersin University, Mersin, Turkey
d
Electrical and Electronics Engineering Department, Engineering Faculty, Firat University, Elazig, Turkey

A R T I C L E I N F O A B S T R A C T

Keywords: In this paper, a new generation Internet-based Power Quality Monitoring System (IPQMS) that transmits real-
Power quality monitoring time power quality (PQ) data over the internet has been developed. The monitoring system has the ability to
FPGA measure PQ parameters in accordance with the related standards. This monitoring system includes many
Embedded systems hardware and software designs and presents an efficient structure. PQ parameters are determined by signal
Signal processing
processing algorithms that are applied to the current and voltage signals continuously obtained from the elec-
trical power network. These signal processing algorithms are performed as embedded functions in the FPGA
device. The PQ data obtained from the measurement points are transmitted to a server by UDP/IP commu-
nication protocol that is implemented in the FPGA device. The monitoring, reporting and permanently storing
tasks are accomplished with the real-time automation software and web applications that running on the server
computer. With its innovative software and hardware designs, the proposed monitoring approach presents a very
useful monitoring structure that can be used in the PQ field.

1. Introduction In order to solve PQ problems that occur on the power system, the
parameters of the system have to be monitored constantly and in ac-
The concept of PQ has gained considerable importance as power cordance with the standards. PQ events are identified and analyzed
electronics based devices begin to be used more commonly in our daily with the monitoring systems placed at particular points on the power
lives. PQ is described as “the determining of electrical quantities at a system. These monitoring systems provide important data about such
certain point in a power system according to a series of reference operational and preservation issues as determination of the perfor-
technical parameters” [1]. There has been a significant increase in the mance of the system on which measurement is taken, the identification
number of studies on PQ recently [2–9]. The reasons for this increase of disturbance type and source on the system, determination of neces-
can be listed as [10–12]; PQ issues in any part of a power system affects sary operational conditions [15,16]. PQ monitoring systems are eval-
not only the part where it occurs but also the whole system due to the uated in two categories. The systems in the first category perform the
interconnected nature of the system. PQ has become a research field measurements of such electrical parameters as voltage, current and
from medium voltage to low voltage with an increase in the integration power as well as overvoltage, low voltage and power failure. The sys-
of renewable energy resources to grid systems. Consumers are given the tems in the second category classify events according to such standards
freedom to choose their electrical energy provider with the enabling of as IEEE, IEC and display these results on their own panels or computers.
competition in production and distribution of electricity and PQ is an These systems also perform some functions such as power quality
important criterion in this choice. Recently, the demand for power was measurement, storage and analysis of the results [17]. Monitoring is a
significantly increased and this leads a decrease in the power quality. primarily needed process to solve power quality problems and classi-
There has been an increase in the usage of power electronics based fication these problems [18–21]. Signal processing methods such as
devices that generate harmonic in parallel with technological devel- root-mean-square (RMS) [22], wavelet transform (WT) [23,19], S-
opments. Smart grid technologies have been became popular and transform [21,24] with machine learning approaches such as artificial
monitoring studies are critical for such systems [13,14]. neural networks (ANN) [3,8], support vector machine (SVM) [19] and

⁎
Corresponding author.
E-mail addresses: [email protected], [email protected] (O. Yildirim).

https://doi.org/10.1016/j.measurement.2018.02.058
Received 17 May 2017; Received in revised form 22 February 2018; Accepted 23 February 2018
Available online 24 February 2018
0263-2241/ © 2018 Elsevier Ltd. All rights reserved.
O. Yildirim et al. Measurement 121 (2018) 109–121

extreme learning machines (ELM) [25,21] are used in the classification function as an embedded system independent of a computer. It is
of PQ disturbances. On the other hand, PQ disturbances can be divided convenient for mass production because it has been designed as an
into two types, single and complex. Complex PQ disturbances usually embedded system.
involve a combination of two or more single types [26]. For this reason, • The monitoring system fulfills remote access technologies require-
it is difficult to identify and classify such PQ problems. In the study ments that are expected of new generation monitoring systems and
[26], successful results were obtained using empirical mode decom- can be accessed online. The developed monitoring system can
position (EEMD) and rank wavelet SVM (r-WSVM) for complex PQ monitor all the data concerning PQ systems that are placed at more
disturbance classification. However, there are different studies in the than one measurement points.
literature for complex PQ disturbance detection [27–31]. • Developed system offers many advantages to users with its easy-to-
The studies related to power quality monitoring systems include use software. Users can report PQ data at their own measurement
many monitoring structures with different methods. When these studies points. PQ data can constantly be recorded by its large scale data-
are examined, the measurements of PQ problems taken by Electrical base.
Power Research Institute at 300 different points within the area of ac-
tivity of 24 different distribution companies in the USA between 1992 The rest of the paper is organized as follows; basic information
and 1994 have proved to be the comprehensive PQ monitoring systems about PQ and common PQ issues are mentioned in the second part.
in this field. Important data have been obtained about PQ by studying Hardware and software design processes of the new generation mon-
the results of these measurements [32]. The same institution conducted itoring system designed within the framework of this study are given in
a different program called DPQ-II between 2001 and 2002 [33,34]. This detail in the third part. The fourth part provides information on the
research showed that voltage sag, voltage swell, transients, harmonics application of the monitoring system in the field and the data obtained
and grounding problems were the most common problems consumers from field applications. Finally, the results of the study have been ad-
faced. Similarly, Canada Electrical Association started a three-year dressed and various suggestions have been made.
study to investigate the power quality in Canada in 1991. Measure-
ments were taken at 550 points for three years. The obtained results 2. Power quality and common power quality issues
have been classified in terms of residential, commercial and industrial
loads and different conclusions about power quality were reached for Different definitions are given for PQ by various people and in-
these loads [35]. In a study done in [36], a web-based monitoring stitutions. Besides, the sources in this field use the term “power quality”
system called I-Grid was introduced. A DSP-based power quality mon- in different ways. The most common terms for PQ are “service quality”,
itoring device called I-Sense was developed within the framework I- “current quality”, “supply reliability”, “feeding quality” and “consumption
Grid. This monitoring device was planned to be placed at 50.000 dif- quality” [48]. PQ according to the IEEE 1159 standard is “supplying
ferent points in the USA and Canada. In [37], besides a low cost mea- power to sensitive electronical devices and doing the grounding in a
surement system, the problems and tendencies in developing dis- convenient way for these devices” [49]. IEEE Std 100–1988 defines PQ
tributed measurement systems have been investigated by using smart as supplying power to electronical devices and doing the grounding in a
sensors. In this study, a web-based measurement system with smart convenient way for the functioning of these devices and other equip-
sensors is presented. In [38], an embedded energy monitoring system ments connected to the system with a wiring system.
approach is presented for monitoring and analyzing energy parameters Power systems are designed to function within a certain frequency
in industrial processes. In [39–43], web-based studies for the mea- and amplitude. Significant deviations in the form of a wave or in fre-
surement and analysis of PQ parameters are presented. In [44], a DSP- quency and amplitude constitute a potential PQ problem. There is a
FPGA-based real time classification system for classifying power quality very close relationship between current and voltage. Although gen-
parameters has been developed. In [45], a prototype application of a erators produce a nearly perfect sinusoidal voltage, the current that
power quality monitoring and measurement system has been realized passes through impedance causes various disturbances in the form of
by using FPGA. In [46], a PQ monitoring system developed on the voltage fluctuation [50]. For example;
CompactRIO controller was proposed. In this system, they used Lab-
VIEW software to monitor and detect disturbances on the power system. • Short circuit currents that occur in the system may cause a voltage
The monitoring system can record events for further analysis. However, leap or a complete voltage cut depending on the situation.
there is no online monitoring structure from different measuring points • Current pulses that occur in a power system due to lighting may
in this system. In [47], a ZigBee routing optimization strategy was often cause sudden high voltage pulses on insulation and other
proposed for an online power quality monitoring system. This system is events such as short circuits.
tested only in the simulation environment and the results are shared. • The currents disturbed because of harmonic generating loads also
Also, this system is proposed only a communication structure for online cause disturbances on the voltages in a system. Therefore, end users
monitoring PQ data. become subjected to a disturbed voltage.
The power quality monitoring systems in the literature are usually
developed from a single point and using ready packaged softwares such There is various approaches to the classification of PQ problems.
as LabVIEW. In this study, a completely innovative online monitoring Each of these approaches uses a certain feature to classify a problem.
system is proposed and original PQ monitoring softwares are designed. Some of these classify PQ events under two captions as “steady state”
This system can monitored PQ parameters from multi measuring points and “non-steady state”. While the most important factor in PQ events is
in real-time. In this study, a new generation monitoring approach for claimed to be the event duration in some regulations such as ANSI
PQ monitoring systems is presented. In this approach, hardware and C84.1, such standards as IEEE 519 use duration and waveform that
software designs for the system that enables web-based monitoring of consists of amplitude to classify PQ issues. Another standard, IEC, uses
power quality parameters have been realized. The advantages of this the frequency interval of PQ data to classify the problems. The IEC
approach can be summed up as follows: 61000-2-5 standard uses frequency interval and evaluates PQ problems
under three main categories. These categories are low frequency (< 9
• The proposed monitoring system, with the innovative approaches it kHz), high frequency (> 9 kHz) and electrostatic discharge events [48].
involves, stands out among traditional PQ monitoring systems. It The characteristics of PQ disturbances also define the type of PQ pro-
can be constantly updated due to the advantages provided by its blems. Detailed definitions can be obtained about the changes that take
original designs. place in the main components of sine wave such as voltage, current and
• The PQ monitoring system based on the FPGA technology can frequency, and the characteristics of PQ issues. The most common

110
O. Yildirim et al. Measurement 121 (2018) 109–121

Va Ia Va Ia Va Ia
Vb Ib Vb Ib Vb Ib
Vc Ic Vc Ic Vc Ic
Vn Vn Vn

Measurement Measurement Measurement

Point I Point II Point III
Signal Card Signal Card Signal Card

FPGA PQ FPGA PQ FPGA PQ

Voltage Curr Pow
Monitoring Voltage Curr Pow
Monitoring Voltage Curr Pow
Monitoring
Harmonics RMS Device Harmonics RMS Device Harmonics RMS Device
Events Freq Events Freq Events Freq

Network Network Network

Device Device Device

Internet/Local
Real Time PQ Datas

Firewall
PQ Web Pages

Network Devices

PQ Automation Internet
Server

Users

PQ Database

Monitoring Center

Fig. 1. Basic working principle of the developed PQ monitoring system.

issues are voltage sag, voltage swell, interruption and harmonics. The current and voltage signals converted from signal input module
are transmitted to the digital signal processing (DSP) module of the
monitoring system. The signal processing algorithms run on the FPGA
3. The proposed power quality monitoring system hardware. PQ data that have been measured in compliance with such
PQ standards as IEC 61000-4-30 are transmitted instantly to the mon-
In this paper, a new monitoring system to monitor PQ parameters itoring center by means of UDP/IP communication protocol. The UDP/
online has been developed. Both hardware and software devices have IP protocol is faster than TCP/IP and its permits a continuous packet
been developed within this monitoring system for the purpose of stream. For this reason we preferred this protocol. The instant data that
monitoring PQ. The monitoring system has been designed in a way to come from measurement points are selected with automation software
continuously monitor the PQ parameters that have been placed at the and stored in the database on the server. Also, various modules have
measurement points of the monitoring system. The basic working been created on the automation for the purpose of monitoring both
principle of the developed monitoring system is given in Fig. 1. instant parameters and past reportings of the measurement points. A
Internet-based Power Quality Monitoring System (IPQMS) transmits web application has been prepared to remotely access PQ data at the
the PQ parameters obtained from the measurement points over the measurement points of IPQMS. Users or system administrators can
internet or a local area network simultaneously to a monitoring center. connect to the monitoring center with this web application from any
PQ data can be monitored simultaneously or retroactively with the point where internet connection is available. Thus, users or system
computer software created in the monitoring center. Monitoring system administrators can easily access to instant or past parameters of the
has been designed to have seven inputs at each measurement point; measurement points.
three for current, three for voltage and one for neutral input. To reduce
the signals of current and voltage to suitable levels faultlessly, various
current and voltage sensory circuits are placed on signal input unit. 3.1. Hardware structure of the monitoring system
Also, an Analog Digital Converter (ADC) integrated circuit is located on
signal input module to convert analog current and voltage signals into There are a few design approaches commonly used for monitoring
digital values. The ADC integrated circuit on the signal input module systems. In these approaches, such hardware as DSP, microprocessors
samples current and voltage signal belonging to each phase with 6.4 and data acquisition cards are used either separately or as hybrids.
kHz sampling frequency. These approaches have some advantages and disadvantages in

111
O. Yildirim et al. Measurement 121 (2018) 109–121

Analog Signals Signal Conversion Digital Signal Processing

Va
Analog to Digital Arithmetic
Voltage Signal I/O
Conversion Unit Logic Unit
Sensor Circuit Process
Vb (ALU)
v v
10101010111
11101000010
Vc 00100101011
11101100111 FPGA Data
(PQ Algorithms)
Ia
Current Signal
Sensor Circuit
Ib Power Supply/Supply
A A
Memory UDP/IP
Circuit
Unit Interface
Ic

V
Power Signal

Fig. 2. The block diagram for the hardware structure of the proposed system.

Voltage Sensor Circuit Current Sensor Circuit

Digital Signal Processing

Modul (FPGA)

Power Supply Circuit Analog/Digital

Convertion Module

Fig. 3. Hardware elements of the PQ monitoring system.

comparison to one another in terms of cost, software developing pro- the voltage sensor circuit is suitable for the highest voltage and fre-
cess, communication and peripheral units. In this study, FPGA hard- quency of the supply voltage to be analyzed and is designed so that the
ware has been preferred in the design of proposed monitoring system characteristic and the accuracy can be kept at 1.2 times the maximum
because it can work at high speeds, offer much more flexible software voltage. A peak factor of at least 1.5 is sufficient for the measurement
structures compared to DSP and microcontrollers and use commu- procedure. The LV25-P voltage sensors of the LEM Company were used
nication functions effectively. Hardware structure of the monitoring for the voltage conversion on the voltage signal sensor circuit of the
system can be divided into five subtitles. These are voltage signal sensor designed monitoring system. The benefits such as high accuracy, line-
circuit, current signal sensor circuit, analog/digital conversion module, arity and high bandwidth are the criteria of choice for the LV25-P
digital signal processing module and power source/supply circuit. The voltage sensors in this study.
block diagram for the hardware structure of the monitoring system is The current sensor circuit is an electronic circuit designed to carry
given in Fig. 2. Hardware elements of the internet-based PQ monitoring out accurate and continuous measurement of current through the net-
system is given in Fig. 3. work. On the current signal sensor circuit, current sensors and various
The reduction of the current and voltage signals to suitable levels for electronic elements are required for the operation of these sensors. In
conversion is realized with voltage signal sensor and current signal the selection of current sensors; accuracy, stability, frequency response,
sensor circuits. These analog signals reduced to suitable levels are isolation are taken into account. In this study, current sensors of 1000
transformed to digital information with ADC. Then, running the algo- amper 9660 model belonging to HIOKI Company were used.
rithms of signal processing on these digital current and voltage data is The Analog digital converter module is used to convert analog
performed by DSP module and transmitted to the output with UDP/IP current and voltage signals obtained from current and voltage sensor
protocol. Selection of electronic elements in the hardware design has circuits to digital signals. The ADC device on this module transforms
been realized sensitively and the necessary tests have been conducted. analog voltage and current signals into digital format that can be de-
The power quality monitoring system has four voltage inputs (three tected by the FPGA processor on the signal processing side. The Max186
phases and neutral). These voltage inputs are connected to the voltage device is used in the study is an Analog digital converter device with 12
bus bars at the measuring points. As it is mentioned in the standards, bit resolution, 8 different channels, high bandwidth and high

112
O. Yildirim et al. Measurement 121 (2018) 109–121

Power Quality
Monitoring Softwares
ISE - VHDL C#- ASP.NET

Computer
FPGA Based
Designs Softwares
Signal Conversions
Digital Signal Processing
UDP/IP Interface Realtime WEB
Memory Organization Automation Applications
I/O Port Controls Softwares

Instant Monitoring Screen User Processing

Reports Tools Monitoring Screen
Database Process Query Modules
UDP/IP Package Decomposition Graphical Tools
Disturbance Event Alert Reports Tools

Fig. 4. Software structure of the PQ monitoring system.

conversion rate serial interface. a suitable form. In this work, a 32-bit IEEE-754 floating-point number
The digital signal processing module measures the power quality system is used. In the data conversion submodule, the signal data of the
parameters according to the standards and transfers these parameters to channels are passed through the channels and passed through the ne-
the computer side PQ database. The digital signal processing module cessary conversion processes and transferred to the main module, from
consists of an FPGA processor to process the signal data and peripherals which it is transmitted to the other submodules. Another parameter that
of the processor. The fact that FPGA hardware has a rich structure in needs to be set for the ADC is the sampling frequency. In this study, it
terms of parallel processing capability and control of peripherals has was preferred to obtain the current and voltage signals as having a 6.4
been the main choice in this study. Virtex5 LX110T model FPGA de- kHz sampling frequency in terms of both the efficiency of the FPGA-side
velopment card of Virtex family produced by Xilinx Company was used software and the memory size in the computer-side database.
in the IPQMS. Within the signal processing module, various signal processing al-
gorithms are included for the purpose of determining the power quality
parameters in accordance with the standards. The digital current and
3.2. Software structure of the monitoring system
voltage information obtained from the signal conversion module is
entered as input data to the algorithms in the signal processing module.
PQ monitoring system softwares have generally been created with
The signal processing module includes; sub-units such as the effective
two different platforms. PQ software include both signal processing
value of the current and voltage signal of the measuring point, network
software on the PQ monitoring system and PQ monitoring software in
frequency information, active power, reactive power, apparent power,
the computer. FPGA designs of the monitoring system have been cre-
power factor, harmonic information and detection of power quality
ated with the platform ISE by using VHDL hardware description lan-
disturbances. Instantaneous current and voltage values obtained ac-
guage. Softwares on the computer side have been designed by using
cording to the measurement ranges specified in the power quality
Microsoft ASP.NET programming languages. A figure showing the
standards are stored in the memory units. Calculation of necessary
software structure of the PQ monitoring system is given in Fig. 4.
parameters is performed by using the signal information on these
The FPGA designs on the monitoring system consists of such parts as
memory units.
the signal processing algorithms to be run in the system and the transfer
In the UDP/IP communication part, the data packages are trans-
of the results obtained from these algorithms to computer-based soft-
ferred by means of Ethernet interface to the server whose IP address has
ware. Although the actions done in the FPGA environment look like
been identified by using the UDP/IP communication protocol. The
they are of a software nature, all the designs are, in fact, done in terms
softwares that are created computer-based consist of sub-softwares that
of hardware. The general structure of the FPGA designs is given in
enable the instant acquisition and storage of data from the monitoring
Fig. 5.
system and serving these data to users by means of visual interfaces.
Each of the embedded softwares in FPGA designed as different
The UDP/IP communication submodule is the part that enables the
modules works in parallel. In the main module (top level), the I/O
transmission of the power quality monitoring system data that obtained
connections that enable the communications of the embedded softwares
from the measuring points to computer side automation. The trans-
with the physical environment and the connections of all the sub soft-
mission of the parameters obtained from the measuring points to the
wares with one another are made possible. In the signal conversion
monitoring center can be achieved through a designated local network
part, binary digital voltage and current values in the form of “1”-“0”
or the internet communication. For this aim, Ethernet interface and
and coming from the system are transformed to network values and
UDP/IP communication interface software are embedded in the mon-
transmitted to other parts. In the DSP part, signal processing algorithms
itoring system. In this way, the monitoring system can send the data
used to determine PQ parameters are found.
from the measurement points to the monitoring center using only an
In addition, I/O controls of the ADC integration are provided under
internet connection cable. The block diagram showing the UDP/IP
this module. The binary encoded digital information from the ADC is
based communication structure of IPQMS is shown in Fig. 6.
transferred from the serial data bus to the signal conversion module. In
Ethernet communication, which is very advantageous for embedded
order to process the binary encoded data in signal processing algo-
systems, eliminates the disadvantages of serial communication
rithms and computer-based software, these data must be converted into

113
O. Yildirim et al. Measurement 121 (2018) 109–121

Signal Conversion
SCLK values from ADC
CS Module Voltage/Current values
Busy binary floating point
Dout 10010110101 0.8787
request signals 10101010101 20.30232
Physical I/O Ports

Din
11110101010 -19.49494
Ground
Buttons Digital Signal
Ethernet
Main Module
(Top Level) Processing
Module

Packages
Packages
UDP/IP Data flow -Voltage, current values

UDP
-RMS calculation
interface module PQ params -Filters
-Active,reactive power
-Fourier Transform
-Harmonics
-Disturbance detection

Fig. 5. The general structure of the FPGA designs.

protocols such as RS232 which has low speed for data transmission. In Table 1
addition, Ethernet communication with the remote connection feature The measurement intervals of the PQ parameters and the data size.
makes it very easy to interfere with the device from the outside. In
Measured value Measurement interval Data size Phase
embedded systems, programming the Ethernet interfaces from the be-
ginning is quite long and requires hard work. For this reason companies Instant voltage wave shape 8 cycle 12,288 byte All phases
usually present some cores. The Xilinx Virtex5 FPGA development Instant current wave shape 8 cycle 12,288 byte All phases
board contains the Tri-Mode Ethernet MAC (TEMAC) core. TEMAC RMS of voltages 10 cycle, 3 s, 10 min 12 byte All phases
RMS of currents 10 cycle, 3 s, 10 min 12 byte All phases
makes it very easy for devices to communicate with network equipment Active power 10 cycle, 3 s, 10 min 12 byte All phases
such as switches and routers. With TEMAC core, Ethernet designs at 10/ Reactive power 10 cycle, 3 s, 10 min 12 byte All phases
100/1000 Mb/s in IEEE 802.3 specification can be realized. In this Power factor 10 cycle, 3 s, 10 min 12 byte All phases
study, Xilinx TEMAC core are modified for UDP/IP communication of Frequency 10 s 4 byte Single phase
1…53. Voltage harmonics 10 cycle, 3 s, 10 min 636 byte All phases
IPQMS.
1…53. Current harmonics 10 cycle, 3 s, 10 min 636 byte All phases
Voltage THD 10 cycle, 3 s, 10 min 12 byte All phases
3.3. PQ database structure Current THD 10 cycle, 3 s, 10 min 12 byte All phases
Disturbance event 0.5 cycle 24,576 byte All phases
The common point among the computer-based softwares in the
proposed system is the database they use. Both softwares commonly use
of the parameters whose 3-s averages have been obtained are recorded
the database where PQ parameters are stored. The data regarding the
in the database.
measurement intervals of the parameters stored in the database of the
monitoring system and the size they take in the database are given in
Table 1. 3.4. Real-time PQ monitoring automation software
The measurement intervals of the PQ parameters are realized in
accordance with the values established in the IEC-61000-4-30 stan- Real-time PQ monitoring automation software consists of visual
dards. By means of the updatable characteristic of the monitoring forms that enable real time monitoring of the data and storage of these
system, storage intervals of parameters can be changed. Instant current data in the database. The automation software that is created with
and voltage waveforms are not continuously recorded in the database. Microsoft Visual C# programming platform makes it possible for users
When necessary, users can record these waveforms in a time interval to monitor the PQ parameters from the measurement points graphically
they prefer. Measurements are generally done in 10-period interval and and digitally. An image of the real-time monitoring automation soft-
3-s averages of these measurements are obtained. The 10-min averages ware is given in Fig. 7.

Fig. 6. Block diagram of IPQMS's UDP/IP based communication structure.

114
O. Yildirim et al. Measurement 121 (2018) 109–121

Fig. 7. Real-time PQ monitoring automation.

Fig. 8. The PQ web application designed for this study.

More than one different measurement points can be defined in the the automation screen. Also, the disturbances that occur at each mea-
automation. Real time parameters of each measurement point are surement point are provided to users as warnings on the main screen.
presented to system users with instant digital information screen. Such
values as voltage, current, active power, reactive power, frequency and
power factor from measurement points are displayed on the instant 3.5. PQ web application
digital information screen. Some menus, which can realize various re-
portings in addition to the information from the center, are placed on A web-based application has been developed in order to instantly
monitor the data from measurement points where internet connection

115
O. Yildirim et al. Measurement 121 (2018) 109–121

Current Clamp

Voltage Probe

Our Proposed
Monitoring System

A) Engineering Faculty

B) Labarotories C) Vocational School

Fig. 9. The connections of PQ monitoring system to the distribution panels at measurement points. (a) Engineering Faculty, (b) Laboratories, (c) Vocational School.

Fig. 10. 10-min changes in the RMS values of three phase voltages at the Engineering Faculty measurement point.

116
O. Yildirim et al. Measurement 121 (2018) 109–121

Fig. 11. 10-min changes in the RMS values of current at the Engineering Faculty measurement point in 5 days.

Fig. 12. Frequency changes at Vocational School measurement point in 7 days.

is available. The web-page that has been created by using ASP.NET 4. Applications and measurement results
technology running on the server and provides access to the data. With
this web application that enables monitoring of PQ on the internet, the The PQ monitoring system designed for this study has been installed
users defined in the system can access detailed information about their on the distribution panels in the Engineering Faculty, Laboratories and
own measurement points. The web application designed for this study is Vocational School at Tunceli University campus. The data acquired
given in Fig. 8. from the measurement points have been continuously transferred to the
The menus prepared on the web page provide users with both di- monitoring center installed in the Electric-Electronics Laboratory. The
gital and graphical data about measurement points. The instant para- monitoring system has been installed to receive three phase voltage,
meter information data about the measurement points defined for the three phase current and one neutral signal. Special voltage probes that
user are displayed on the home page. This information data are updated are very easy to use are connected to the bus bar of each phase for
every three seconds. Like the real-time automation software, the web measuring voltage signal. Current connections are realized by taking
application of the monitoring system, too, contains sub menus that can into account the direction of the current that passes through the con-
report the data in detail. Users can access instant data from any mea- ductor and by connecting it in a way that current sensors encircle the
surement point that they choose. They can also perform graphical and conductor. The current sensors used in this study are especially con-
digital reporting on the web. venient in that it is easy to install in the network and it can be installed
without having to cut the circuit. The distribution panels that are re-
lated to the measurement points where IPQMS was installed and the
system’s connections to the distribution panels are shown in Fig. 9.

117
O. Yildirim et al. Measurement 121 (2018) 109–121

Fig. 13. The waveforms and RMS changes of a voltage sag detected by the PQ monitoring system.

Table 2
Voltage disturbances that occur at measurement points.

Laboratories Engineering faculty Vocational school

Disturbance Date Disturbance Date Disturbance Date

Interruption 08.08.2014 13:43:25 Voltage Sag 19.08.2014 13:41:35 Interruption 28.08.2014 11:30:14
Voltage Swell 08.08.2014 13:43:34 Voltage Sag 19.08.2014 13:52:42 Voltage Sag 29.08.2014 07:12:02
Voltage Sag 09.08.2014 06:08:23 Voltage Sag 19.08.2014 15:43:30 Voltage Sag 29.08.2014 17:36:39
Voltage Sag 09.08.2014 13:17:33 Interruption 21.08.2014 18:54:05 Voltage Sag 30.08.2014 04:50:12
Interruption 09.08.2014 17:24:20 Voltage Swell 21.08.2014 19:00:47 Voltage Sag 30.08.2014 05:26:18
Voltage Sag 10.08.2014 17:17:39 Voltage Sag 22.08.2014 14:17:03 Voltage Sag 31.08.2014 08:42:31
Voltage Swell 10.08.2014 17:24:05 Voltage Sag 02.09.2014 05:59:47
Voltage Sag 11.08.2014 07:11:16 Voltage Sag 03.09.2014 14:47:58

Fig. 14. Changes in 3rd, 5th, and 7th voltage harmonics as 10-min averages from the phase A of the measurement point at the Engineering Faculty.

118
O. Yildirim et al. Measurement 121 (2018) 109–121

Fig. 15. 3rd, 5th, and 7th harmonic values from the phase A of the Vocational School measurement point.

Fig. 16. Active and reactive power changes at the Engineering Faculty measurement point.

4.1. PQ measurement at measurement points and results 4.1.1. RMS changes

Fig. 10, shows 10-min averages of the RMS value changes of the
Measurement results have generally been presented under five ca- three phase voltages that have been obtained from the Engineering
tegories; RMS changes, frequency, disturbance events, harmonics and Faculty measurement point.
power components. RMS changes have separately been measured for For current measurements, the changes in the RMS value have been
the three phases of current and voltage signals at measurement points. measured for the measurement time of each phase as 10 min averages.
Frequency measurements have been obtained from the single phase of In Fig. 11, the changes in the effective values of three phase currents
the voltage signals. PQ disturbances that occur in voltage have been obtained from the 5 day measurements at the Engineering Faculty
determined for three phase voltage. For harmonics measurements, as measurement point is shown.
many harmonic components as 53th have been measured for each phase
of both voltages and currents. For the measurements of power com- 4.1.2. Frequency measurements
ponents, active power changes, reactive power changes and power Frequency values of the each measurement point have been mea-
factor changes on the “Phase A” of each measurement point has been sured over a single phase within the PQ monitoring system. In the
given. The real-time measurement data have been recorded in a 10 min Fig. 12, the frequency changes in 7 days at Vocational School mea-
time interval as specified by standards. Thus a more productive usage of surement point is given.
the database has been achieved. The measurement results obtained
from field applications are given in the subsequent parts of the paper. 4.1.3. Voltage disturbance events
The PQ event detection system that has been designed in accordance

119
O. Yildirim et al. Measurement 121 (2018) 109–121

with the measurement intervals specified by IEC 61000-40-30 detects signal processing techniques and machine learning algorithms.
the PQ disturbances that occur in the voltages at measurement points in
real time. The waveform of the PQ disturbance event consists of 16 Acknowledgement
periods in total; 8 periods before the event and 8 periods after the
event. The PQ event detection system can determine voltage sag, vol- This study was supported within the scope of the SANTEZ project
tage swell and interruption events successfully. In the Fig. 13, the wa- numbered STZ0320-12. We thank the Ministry of Science, Industry and
veform of a voltage sag detected by the PQ disturbance detection Technology of the Turkish Republic for their support.
system and the changes in the value of voltage RMS are given.
During the measurement times, a total number of 22 PQ disturbance References
at three measurement points by the monitoring system. When voltage
events are examined, it can be seen that the disturbance type occurring [1] IEC 61000-4-30, Electromagnetic Compability (EMC) – Part 4: Testing and mea-
often at measurement points is the voltage sag. The graphical and di- surement techniques – Section 30: Power quality measurement methods,
International Electrotechnical Commission, Switzerland, 2003.
gital data of disturbance events in voltage can easily be accessed within [2] H. Liu, F. Hussain, Y. Shen, Power quality disturbances classification using com-
the proposed monitoring system. In Table 2, the numeric distribution pressive sensing and maximum likelihood, IETE Techn. Rev. (2017) 1–10.
from voltage disturbances obtained at measurement points are given. [3] C.B. Khadse, M.A. Chaudhari, V.B. Borghate, Conjugate gradient back-propagation
based artificial neural network for real time power quality assessment, Int. J. Electr.
Power Energy Syst. 82 (2016) 197–206.
4.1.4. Harmonic measurements [4] R. Kumar, B. Singh, D.T. Shahani, C. Jain, Dual-tree complex wavelet transform-
The measurements of the harmonics at measurement points have based control algorithm for power quality improvement in a distribution system,
IEEE Trans. Ind. Electron. 64 (2017) 764–772.
revealed as many as 1st to 53rd harmonic components from both vol- [5] D. Li, K. Yang, Z.Q. Zhu, Y. Qin, A novel series power quality controller with re-
tage and current signals. These harmonic data can be reported selec- duced passive power filter, IEEE Trans. Ind. Electron. 64 (2017) 773–784.
tively over the real-time PQ automation system. In Fig. 14, the changes [6] A.S. Bubshait, A. Mortezaei, M. Simoes, T. Busarello, Power quality enhancement
for a grid connected wind turbine energy system, IEEE Trans. Ind. Appl. (2017).
in 3rd, 5th, and 7th voltage harmonics as 10-min averages from the
[7] S. Ali, K. Wu, K. Weston, D. Marinakis, A machine learning approach to meter
“Phase A” of the measurement point at the Engineering Faculty are placement for power quality estimation in smart grid, IEEE Trans. Smart Grid 7
presented. (2016) 1552–1561.
In Fig. 15, the 3rd, 5th, and 7th current harmonic value changes [8] S. Khokhar, A.A.M. Zin, A.P. Memon, A.S. Mokhtar, A new optimal feature selection
algorithm for classification of power quality disturbances using discrete wavelet
from the “Phase A” of the Vocational School measurement point are transform and probabilistic neural network, Measurement 95 (2017) 246–259.
shown. [9] Y. Shen, F. Hussain, H. Liu, D. Addis, Power quality disturbances classification
based on curvelet transform, Int. J. Comput. Appl. (2017) 1–10.
[10] S.K. Khadem, M. Basu, M. Conlon, Power quality in grid connected renewable en-
4.1.5. Measurement of power components ergy systems: role of custom power devices, in: International Conference on
The measurements of power components include active power, re- Renewable Energies and Power Quality, Granada, Spain, 2010.
active power, apparent power and power factor parameters. [11] M. Farhoodnea, A. Mohamed, H. Shareef, H. Zayandehroodi, Power quality impacts
of high-penetration electric vehicle stations and renewable energy-based generators
Measurements of power components on the power system are very on power distribution systems, Measurement 46 (2013) 2423–2434.
important to characterize system. The constant monitoring and storage [12] B. Kang, System design for real-time measuring of power quality and harmonics
of the energy the system spends give network managers a great ad- distortion using digital signal processor, J. Kor. Inst. Inform. Commun. Eng. 20
(2016) 1283–1289.
vantage. In Fig. 16, the changes in the active and reactive power [13] M. Jaradat, M. Jarrah, A. Bousselham, Y. Jararweh, M. Al-Ayyoub, The internet of
components recorded at the Engineering Faculty measurement point energy: smart sensor networks and big data management for smart grid, Procedia
are presented. Comput. Sci. 56 (2015) 592–597.
[14] L. Morales-Velazquez, R. de Jesus Romero-Troncoso, G. Herrera-Ruiz, D. Morinigo-
Sotelo, R.A. Osornio-Rios, Smart sensor network for power quality monitoring in
5. Conclusion electrical installations, Measurement 103 (2017) 133–142.
[15] H. Erişti, Development of a wavelet transform and support vector machine based
In this paper, a new internet-based system called IPQMS has been event receognition technique for power quality, PhD Thesis, Fırat University
Natural and Applied Sciences Institue, Elazıg/Turkey, 2010.
developed to monitor PQ continuously and in accordance with the re- [16] S.M. Blair, C.D. Booth, G. Williamson, A. Poralis, V. Turnham, Automatically de-
lated standards. Various hardware and software designing this mon- tecting and correcting errors in power quality monitoring data, IEEE Trans. Power
itoring system have been created. In terms of software, the designing of Delivery 32 (2017) 1005–1013.
[17] D.J. Won, I.Y. Chung, J.M. Kim, S.I. Moon, J.C. Se, J.W. Cho, Development of power
a new generation PQ monitoring system has been done. This mon- quality monitoring system with central processing scheme, IEEE Power Engineering
itoring system is easily installed at measurement points and enables the Society Summer Meeting, vol.18, 2002, pp. 915–919.
continuous acquisition of PQ data from these points. Also, it can per- [18] A.F. Zooba, M.M. Canteli, R. Bansal, Power quality: monitoring, analysis and en-
chancement, Intech Open Access, Croatia, 2011.
form the real time analyses of signals by means of signal processing [19] H. Erişti, Ö. Yıldırım, B. Erişti, Y. Demir, Optimal feature selection for classification
algorithms it includes. FPGA devices, which has become popular in of the power quality events using wavelet transform and least squares support
signal processing recently, has been used in the new generation PQ vector machines, Int. J. Electr. Power Energy Syst. 49 (2013) 95–103.
[20] M. Hajian, A.A. Foroud, A new hybrid pattern recognition scheme for automatic
system. FPGA hardware has been preferred in this study because of its
discrimination of power quality disturbances, Measurement 51 (2014) 265–280.
suitable architecture for signal processing. [21] H. Erişti, Ö. Yıldırım, B. Erişti, Y. Demir, Automatic recognition system of under-
Three different measurement points have been selected for the lying causes of power quality disturbances based on s-transform and extreme
learning machine, Int. J. Electr. Power Energy Syst. 61 (2014) 553–562.
testing of IPQMS. These are the Engineering Faculty, Vocational School
[22] P.G.V. Axelberg, I.Y.-H. Gu, M.H.J. Bollen, Support vector machine for classification
and Laboratories at Tunceli University campus. The monitoring system of voltage disturbances, IEEE Trans. Power Deliv. 22 (2007) 1297–1303, http://dx.
has been installed at these points for the durations specified for each doi.org/10.1109/TPWRD.2007.900065.
measurement point and continuous data have been obtained. The data [23] C.A. Naik, P. Kundu, Power quality index based on discrete wavelet transform, Int.
J. Electr. Power Energy Syst. 53 (2013) 994–1002, http://dx.doi.org/10.1016/j.
from the measurement points have been transmitted in real time to the ijepes.2013.06.024.
monitoring center established in the Electric-Electronic Laboratory. The [24] O.P. Mahela, A.G. Shaik, Recognition of power quality disturbances using S-trans-
data obtained from the field applications of IPQMS have showed that form based ruled decision tree and fuzzy C-means clustering classifiers, Appl. Soft
Comput. J. 59 (2017) 243–257, http://dx.doi.org/10.1016/j.asoc.2017.05.061.
our proposed monitoring system is very successful in terms of applic- [25] T. Chakravorti, P.K. Dash, Multiclass power quality events classification using
ability and convenience. The proposed monitoring system can be used variational mode decomposition with fast reduced kernel extreme learning ma-
in the smart grid studies because power quality is essential for internal chine-based feature selection, IET Sci. Meas. Technol. 12 (2017) 106–117.
[26] Z. Liu, Y. Cui, W. Li, A classification method for complex power quality disturbances
intelligent mechanisms of a healthy grid. In addition, the proposed using EEMD and rank wavelet SVM, IEEE Trans. Smart Grid 6 (2015) 1678–1685.
system provides a platform that can be used for complex PQ operations [27] Z. Liu, Q. Zhang, An approach to recognize the transient disturbances with spectral
such as analysis of complex PQ problems by implementing appropriate kurtosis, IEEE Trans. Instrum. Meas. 63 (2014) 46–55, http://dx.doi.org/10.1109/

120
O. Yildirim et al. Measurement 121 (2018) 109–121

TIM.2013.2277513. energy monitoring and analysis in industrial processes, Energy 115 (2016)
[28] Z. Liu, Q. Hu, Y. Cui, Q. Zhang, A new detection approach of transient disturbances 811–819.
combining wavelet packet and Tsallis entropy, Neurocomputing 142 (2014) [39] L. Moreira, S. Leitão, Z. Vale, J. Galvão, Use of Web Based Meters to Improve Energy
393–407, http://dx.doi.org/10.1016/j.neucom.2014.04.020. Efficiency and Power Quality in Buildings. In Doctoral Conference on Computing,
[29] Z. Liu, Q. Zhang, Z. Han, G. Chen, A new classification method for transient power Springer. International Publishing, Electrical and Industrial Systems, 2015, pp.
quality combining spectral kurtosis with neural network, Neurocomputing 125 337–344.
(2014) 95–101, http://dx.doi.org/10.1016/j.neucom.2012.09.037. [40] I. Adam, A. Mohmmed, H. Sansui, Embedded web-based power quality data logger,
[30] D.D. Ferreira, J.M. De Seixas, A.S. Cerqueira, A method based on independent J. Eng. Technol. 1 (2011) 58–62.
component analysis for single and multiple power quality disturbance classification, [41] Y.H. Chen, J.C. Cheng, L.C. Hwang, J.H. Teng, S.Y. Cheng, Embedded web server
Electr. Power Syst. Res. 119 (2015) 425–431, http://dx.doi.org/10.1016/j.epsr. based power quality recorder, TENCON 2005, 21–24 Nov., 2005, pp. 1–6.
2014.10.028. [42] M. Zhang, K. Li, A power quality monitoring system over the internet, Int. Conf.
[31] A.K. Puliyadi Kubendran, L. Ashok Kumar, Detection and classification of complex Inform. Sci. Eng. (2009) 1577–1580.
power quality disturbances using S-transform amplitude matrix–based decision tree [43] M.E. Monteiro, E.S. Moura, A.B. Drago, L.O. Bonino, J.G. Pereiraf, P.F.S. Amaral, D.
for different noise levels, Int. Trans. Electr. Energy Syst. 27 (2017). http://doi.org/ S.L. Simonetti, P.F. Rosa, G.D. Bazelatto, An internet-based power quality mon-
10.1002/etep.2286. itoring system, vol. 1, in: IEEE International Symposium on Industrial Electronics
[32] E.W. Gunther, H. Mehta, A survey of distribution power system power quality ISIE'03, 9–11 June, 2003, pp. 333–336.
preliminary results, IEEE Trans. Power Deliv. 10 (1995) 322–329. [44] M. Zhang, K. Li, H. Yisheng, Dsp-fpga based real-time power quality disturbances
[33] C.J. Melhorn, A. Maitra, W. Sunderman, M. Waclawiak, A. Sundaram, Distribution classifier, Metrol. Measure. Syst. 17 (2011) 205–216.
system power quality assessment phase-II: voltage sag and interruption analysis, [45] A.D. Femine, D. Gallo, C. Landi, M. Luiso, Power quality monitoring instrument
Copyright IEEE, 2005, PCIC-2005-13. with fpga transducer compensation, IEEE Trans. Instrum. Meas. 58 (2009)
[34] B.H.S. McNulty, Power quality problems and renewable energy solutions, analysis 3149–3158.
of the prospects for renewable PQ solutions in Massachusetts, Massachusetts [46] M.V.E. Xavier, W. do Couto Boaventura, M.D.I. Noronha, High performance power
Renewable Energy Trust, USA, 2002. quality monitoring system, in: IEEE 17th International Conference on Harmonics
[35] M.B. Hughes, J.S. Chan, Canadian national power quality survey results, IEEE Proc. and Quality of Power (ICHQP), October 2016, pp. 974–979.
Transmiss. Distribut. Conf. (1996) 45–51. [47] Z. Teng, M. Zhang, T. Jiang, J. Xu, Q. Zhang, A power quality online monitoring
[36] D. Divan, G. Luckjiff, W. Brumsickle, I-GridTM: Infrastructure for nationwide real- system oriented ZigBee routing optimization strategy, Wireless Netw. 22 (2016)
time power monitoring, IEEE 37th IAS Annual Meeting. vol. 3, 2002, pp. 1869–1875.
1740–1745. [48] E.F. Fuchs, M.A.S. Masoum, Power quality in power systems and electrical ma-
[37] J.J. Gonzales de la Rosa, A.M. Munoz, A. Gil de Castro, V.P. Lopez, J.A.S. Castillejo, chines, Academic Press, 2008.
A web-based distributed measurement system for electrical power quality assess- [49] C. Sankaran, Power quality, CRC Press, 2002.
ment, Measurement 43 (2010) 771–780. [50] R.C. Dugan, M.F. McGranaghan, S. Santoso, H.W. Beaty, Electrical Power Systems
[38] J.A. Back, L.P. Tedesco, R.F. Molz, E.O.B. Nara, An embedded system approach for Quality, McGraw-Hill, 2004.

121

View publication stats