Hindawi

Mathematical Problems in Engineering

Volume 2022, Article ID 1463006, 22 pages
https://doi.org/10.1155/2022/1463006

Research Article
Consistency Improvement Method of Pairwise Matrix Based on
Consistency Ratio Decreasing Rate and Attribute Weighting
Method Considered Decision Makers’ Levels in Analytic Hierarchy
Process: Application to Hip Joint Prosthesis Material Selection

Won-Chol Yang ,1 Hyon-Song Kang ,2 Gyong-Su Ri ,2 and Jin-Sim Kim 1

1
Faculty of Materials Science and Technology, Kim Chaek University of Technology, 60 Kyogu, Pyongyang,
Democratic People’s Republic of Korea
2
Faculty of Mathematics, Kim Jong Tae Haeju University of Education, 44 Sami, Haeju, Democratic People’s Republic of Korea

Correspondence should be addressed to Won-Chol Yang; [email protected]

Received 10 June 2022; Revised 30 August 2022; Accepted 21 September 2022; Published 17 October 2022

Academic Editor: Harish Garg

Copyright © 2022 Won-Chol Yang et al. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is
properly cited.
Analytic hierarchy process (AHP) is a well-known attribute weighting method in multiattribute decision-making. Its major
requirement is to satisfy the consistency of pairwise matrix (PM). To solve this problem, we first propose a new consistency
improvement method of PM based on consistency ratio (CR) decreasing rate. In this method, we calculate the CR decreasing rates
of all the PMs reconstituted by replacing all elements of the PM with the lower and upper neighbouring 9-point scales and find the
element with maximum CR decreasing rate, and then modify it to its lower or upper neighbouring scale. Second, we develop third-
order approximate polynomial for random consistency index using least square method. It enables to determine the RI value
according to the number of attributes without a numerical table. Third, we propose the final PM determining method and final
attribute weighting method considered decision makers’ levels based on the CR values of the individual PMs in case several
decision makers perform their own pairwise comparisons. We test the performances of the proposed and some previous
consistency improvement methods with two numerical examples. The results demonstrate that the proposed method improves
the consistency of PM better and faster with smaller amount of modification than that of the previous methods, while it modify the
elements of the PM to 9-point scales, necessarily. We apply the proposed method to hip joint prosthesis material selection. The
proposed methods may be widely used in practical applications of AHP.

1. Introduction comparisons between the attributes instead of the direct

allocation of the weights [3]. The details of AHP have been
Determining a reasonable attribute weights plays a vital role described in the literature.
in multiattribute decision-making (MADM) and multi- AHP has been widely applied to calculate the subjective
objective optimization (MOO) because the decision-making attribute weights in many practical MADM and MOO
and optimization results may differ according to the attri- problems [4, 5]. Soni et al. [6] determined the criteria
bute weights. Analytic hierarchy process (AHP) is well- weights using AHP for material selection of reinforced
known attribute weighting method [1]. AHP determines the sustainable composites by recycling waste plastics and agro-
attribute weights based on pairwise comparison evaluation waste. Zhong et al. [7] constructed a cost evaluation system
data for the pairs of attributes. The 9-point scales are used to with 5 indices in the first level and 22 indices in the second
transform the decision maker’s judgments into numerical level using AHP. Radulescu et al. [8] calculated the overall
quantities [2]. The essential feature of AHP is the pairwise weights of the criteria as a linear combination of the
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2 Mathematical Problems in Engineering

individual weights obtained from the group AHP and ex- et al. [21] proposed ANTAHP method using ant optimi-
tended entropy weighting method for evaluating the fourth zation algorithm to reconstitute the inconsistent PM by
wave of COVID-19 pandemic. Peng and Wu [9] determined minimizing the distance between the primary and modified
the comprehensive weights of each index using AHP to score PMs. Wu et al. [22] improved the inconsistency of the PM
the index system for evaluating the benefit development of using marginal optimization method. ,e method is based
the offshore wind power after the cancellation of the public to increase (or decrease) all elements by a fixed value, and it
subsides. Wei et al. [10] constructed the evaluation model calculates the marginal effect of each modification. Zeshui
using AHP and fuzzy comprehensive evaluation. Mathew and Cuiping [23] proposed a consistency improvement
et al. [11] calculated the weights of the criteria using fuzzy method based on auto-adaptive process. In the method, the
AHP and determined the final ranking of the alternatives element aij of the inconsistent PM A is replaced by bij � aijα
using spherical fuzzy TOPSIS. Rawa et al. [12] proposed an (wi /wj )1−α, where w � ((w1 , . . . , wi , . . . , wn )T is the weight
economical-technical-environmental operation for power vector obtained from A. ,e generated matrix B � [bij] has a
networks with wind-solar-hydropower generation by using reduced CR. ,is process is repeated until the consistency is
AHP and improved grey wolf algorithm. ,ey used a satisfied. Cao et al. [24] proposed a heuristic method to
weighted sum strategy using AHP to transform the multi- modify inconsistent PM. ,ey decomposed the primary PM
objective problem into a normalized single objective one. as the Hadamard product of the consistent PM and a re-
Okudan and Budayan [13] used fuzzy AHP for conducting ciprocal deviation matrix. ,ey constituted a modified PM
the evaluation of project characteristics affecting risk oc- by convex combination of the reciprocal zero deviation
currences in the construction projects. Dano [14] analysed matrices. ,ey proposed auto-adaptive modification algo-
the impacts of flash hazards using AHP and identified the rithm using such convex combinations. Yang et al. [25]
most effective methods to reduce the flash flood impacts modified the inconsistent PM by combining the particle
using expert’s opinions in Jeddah. Chen et al. [15] dealt with swarm optimization and Taguchi method. Benı́tez et al. [26]
the uncertainty of the wind power, the load demand, and the proposed a linearization method to provide the closest
multiobjective function by using fuzzy chance constraint consistent PM to the inconsistent PM by using orthogonal
programming and improved AHP. projection in a linear space.
For group decision-making problems, hesitant fuzzy set, For consistency improvement, it needs to pay attention
complex intuitionistic fuzzy set, and probabilistic hesitant to guarantee a good balance between improvement of
fuzzy set have been introduced to handle and model the consistency and preservation of primary information.
uncertainty and vagueness in decision-making very effec- However, the previous methods are lacking in guaranteeing
tively, reflect the importance of different numerical values such balance. On the other hand, the elements of the PM are
more clearly, elicit the decision makers’ knowledge, and 9-point scales because the pairwise comparison is performed
develop more effective decision-making model. [16–18] Rani by means of 9-point scales {1/9, 1/8, . . ., 8, 9} in the con-
and Garg [16] proposed a novel algorithm for multiattribute ventional AHP. However, in the previous methods, the el-
group decision-making using complex intuitionistic fuzzy ements of the reconstituted PMs are no 9-point scales, and
values. Jin et al. [17] proposed a decision-making model therefore, the PMs obtained from the previous methods are
using probabilistic hesitant fuzzy preference relations for no inherent ones. To deal with this shortcoming, we propose
reflecting clearly the importance of different numerical a new consistency improvement method according to the
values and eliciting the decision makers’ knowledge in the following principles:
group decision-making problems. Liu et al. [18] calculated
the probabilities of elements in the probabilistic hesitant (i) ,e amount of consistency improvement of the PM
fuzzy element and the probability of risk status by using two should be as large as possible, and the deviation
nonlinear programming models. Khan et al. [19] proposed a between the primary and reconstituted PMs and the
performance measure using an MADM method based on the number of the modified elements should be as small
complex T-spherical fuzzy power aggregation operators. Liu as possible.
et al. [20] developed a novel correlation coefficient to
(ii) ,e elements of the inconsistent PM should be
measure the strength of the relationship between the hesitant
replaced with the lower or upper neighbouring 9-
fuzzy sets.
point scales, and the elements of the reconstituted
Although AHP is a useful tool for attribute weighting, it
PM should be 9-point scales.
has some drawbacks. One drawback is that it is difficult to
conduct pairwise comparison in practical applications. Yang When two or more decision makers take part in the
et al. [5] proposed a simplest questionnaire to conduct the pairwise comparison between the attributes, the PMs and the
pairwise comparison, easily and conventionally. Another attribute weights may differ according to their knowledge
drawback is that it is difficult to satisfy the consistency of and opinions. ,erefore, it is necessary to constitute a final
pairwise matrix (PM) in practical applications. To determine PM by synthesizing the individual PMs obtained from each
the reasonable attribute weights using AHP, the consistency decision makers and determine the final attribute weighting
of PM must be satisfied. When it does not satisfy the from the final PM. It needs to consider the decision makers’
consistency, it is need to repair the primary PM. In order to levels to constitute the final PM. However, there is no
satisfy the consistency of the PM, some researchers proposed reasonable objective method to determine the decision
the reconstitution methods of the inconsistent PM. Girsang makers’ levels, while the previous methods are generally
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Mathematical Problems in Engineering 3

subjective ones. To overcome this shortcoming, we propose a 2. Methodology

new objective method to determine the decision makers’
levels based on the CR values of their PMs. ,e worse the 2.1. CR Decreasing Rate-Based Consistency Improvement
consistency of the PM is, the more mistakes the decision Method of Inconsistent PM. Let CR(A) be the CR value of the
maker’s judgment has, and therefore, we can regard that the PM A � (aij)n×n.
CR value reflects the decision maker’s level. ,e main steps of a new consistency improvement
We propose a new consistency improvement method of method are as follows:
pairwise matrix based on consistency ratio decreasing rate Step 1. For each element alm (l � 1, n − 1, m � l + 1, n)
and attribute weighting method considered decision makers’ in the upper triangular matrix of the PM A � (aij)n×n,
levels in analytic hierarchy process and apply the methods to reconstitute the PM A−lm � (a−ij )n×n by replacing the
hip joint prosthesis material selection. element alm with the lower neighbouring scale h−lm , and
,e novelties and advantages of the proposed methods then calculate its CR value CR(A−lm ) and CR decreasing
are as follows: rate dC Rr−lm as follows:
(i) In the CR decreasing rate-based consistency im- dC R−
provement method of inconsistent PM, the ele- dC Rr−lm � 􏼌􏼌 􏼌 􏼌􏼌 lm
􏼌 􏼌􏼌 , (1)
ments of the inconsistent PM are modified to the 􏼐􏼌􏼌alm − hlm 􏼌􏼌 + 􏼌􏼌1/alm − 1/h−lm 􏼌􏼌􏼑
−

adjacent 9-point scales, and all the elements of the

reconstituted PM are the 9-point scales, not real where
numbers. ,is method improves the consistency dC R−lm � CR(A) − CR A−lm 􏼁,
more, better, and faster with smaller number of
elements and smaller amount of modification, and it 1 −
guarantees a very good balance between consistency a−lm � h−lm ,a−ml � , a � a , i,j � 1,n, (i, j) ≠ (l,m).
a−lm ij ij
improvement and information preservation of the
(2)
primary PM.
(ii) ,e approximate formula is used to determine the ,e lower and upper neighbouring scales of the 9-point
RI value according to the number of attributes, scales {1/9, 1/8, . . ., 8, 9} are shown in Table 1.
not the numerical table for RI. It enables to test the Step 2. Find the maximum value of the CR decreasing
consistency of PM without a numerical table for rate dC Rr−r1 s1 from {dC Rr−lm ; l � 1, n − 1, m � l + 1, n}
RI. as follows:
(iii) In the final attribute weighting method considered
decision makers’ levels, the CR value is used as an dC Rr−r1 s1 � max 􏼈dC Rr−lm 􏼉. (3)
1≤l<m≤n
objective measure that reflects the decision maker’s
level. It is possible to determine the decision Step 3. For each element alm (l � 1, n − 1, m � l + 1, n)
makers’ levels objectively, not subjectively. It en- in the upper triangular matrix of the PM A � (aij)n×n,
ables to determine the attribute weights, more reconstitute the PM A+lm � (a+ij )n×n by replacing the
scientifically and reasonably. In this method, the element alm with the upper neighbouring scale h+lm , and
elements of the final PM are also 9-point scales, and then calculate its CR value CR(A+lm ) and CR decreasing
it enables to preserve the inherent characteristics of rate dC Rr+lm as follows:
AHP.
dC R+lm
,e rest of this paper is organized as follows: In dC Rr+lm � 􏼌􏼌 +
􏼌
􏼌 􏼌􏼌 􏼌
+ 􏼌 , (4)
􏼐􏼌􏼌alm − hlm 􏼌􏼌 + 􏼌􏼌1/alm − 1/hlm 􏼌􏼌􏼑
Section 2, we describe a new consistency improvement
method of PM based on CR decreasing rate, a devel-
where
opment method of approximate polynomial for RI
according to number of attributes, and a constitution dC R+lm � CR(A) − CR A+lm 􏼁,
method of final PM and final attribute weighting method
considered decision makers’ levels. In Section 3, we 1 +
a+lm � h+lm ,a+ml � , a � a , i,j � 1,n, (i, j) ≠ (l,m).
describe the numerical test results of the proposed a+lm ij ij
method and its application to hip joint prosthesis ma- (5)
terial selection. In Section 4, we present the conclusions.
In Appendix section, we describe the attribute weighting Step 4. Find the maximum value of the CR decreasing
method using AHP with the simplest questionnaire, rate dC Rr+r2 s2 from {dC Rr+lm ; l � 1, n − 1, m � l + 1, n}
development method of approximate polynomial by least as follows:
square method, and three well-known MADM methods
(MADMs) such as simple additive weighting (SAW) dC Rr+r2 s2 � max 􏼈dC Rr+lm 􏼉. (6)
1≤l<m≤n
method, technique for order preference by similarity to
ideal solution (TOPSIS) method, and grey relational Step 5. If dC Rr−r1 s1 ≥ dC Rr+r2 s2 , then reconstitute the
analysis (GRA) method used in this work. PM B � (bij )n×n by replacing the element ar1 s1 with the
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
4 Mathematical Problems in Engineering

Table 1: Lower and upper neighbouring scales of the 9-point scales.

9-Point scale 1/9 1/8 1/7 1/6 1/5 1/4 1/3 1/2 1 2 3 4 5 6 7 8 9
Lower neighbouring scale 1/9 1/9 1/8 1/7 1/6 1/5 1/4 1/3 1/2 1 2 3 4 5 6 7 8
Upper neighbouring scale 1/8 1/7 1/6 1/5 1/4 1/3 1/2 1 2 3 4 5 6 7 8 9 9

lower neighbouring scale h−r1 s1 in the PM A � (aij)n×n, ,e greater the values of dCR(A,B) and dCRr(A,B) are,
where the better the performance of the consistency improvement
1 method is. ,e smaller the value of dM(A,B) is, the better the
br1 s1 � h−r1 s1 , bs1 r1 � , bij � aij , reconstituted PM preserves the information of the primary
br1 s1
(7) PM. ,e CR decreasing rate dCRr(A,B) becomes the major
measure to evaluate the performance of the consistency
i, j � 1, n, (i, j) ≠ r1 , s1 􏼁. improvement method of PM from among above three
measures.
If dC Rr−r1 s1 < dC Rr+r2 s2 , then reconstitute the PM B �
(bij )n×n by replacing the element ar2 s2 with the upper
neighbouring scale h+r2 s2 in the PM A � (aij)n×n, where 2.2. Development Method of Approximate Polynomial for RI
1 According to the Number of Attributes. We develop m-th
br2 s2 � h+r2 s2 , bs2 r2 � , bij � aij , order approximate polynomial for RI as the following form:
br2 s2
(8)
RI � RIm (n) � a0 + a1 n . . . + am−1 nm− 1 + am nm , (13)
i, j � 1, n, (i, j) ≠ r2 , s2 􏼁.
with the data set {(n, RIn), n � 3, 15.}, where n is the number
Step 6. Calculate the CR value CR(B) of the recon- of attributes and RIn is the corresponding RI value (Table 2).
stituted PM B � (bij )n×n. ,e MAE, MRE, and MSE of the m-th order approxi-
Step 7. If CR(B) > CR0, then A � B and go to Step 1. mate polynomial are as follows:
Step 8. If CR(B) ≤ CR0, then calculate the principal 1 15 􏼌􏼌􏼌 􏼌􏼌
eigenvector (v � v1 , . . . , vj , . . . , vn )T from the recon-
MAEm � 􏽘 􏼌RIn − RIm (n)􏼌􏼌,
13 n�3
stituted PM B � (bij )n×n.
􏼌 􏼌
Step 9. Calculate the attribute weights (w1 , . . . , 1 15 􏼌􏼌􏼌􏼌RIn − RIm (n)􏼌􏼌􏼌􏼌
wj , . . . , wn )T by normalizing (v � v1 , . . . , vj , . . . , vn )T MREm � 􏽘􏼌 􏼌􏼌 × 100%, (14)
13 n�3􏼌􏼌 RIn 􏼌
as follows:
vj 1 15
wj � n , j � 1, n. (9) MSEm �
2
􏽘 RIn − RIm (n)􏼁 .
􏽐k�1 vk 13 n�3

We call this method CR decreasing rate-based method. ,e main steps to develop the approximate polynomial
Let A � (aij)n×n and B � (bij)n×n be the primary incon- for RI value according to the number of attributes are as
sistent PM and reconstituted PM, respectively. follows:
To evaluate the performance of the consistency im-
Step 1. Develop six approximate polynomials
provement method of the inconsistent PM, we use the
RI � RIm(n); m � 1, 6 with the data {(n, RIn), n � 3, 15.}
following three measures:
(Table 2).
(i) CR decreasing amount, Step 2. Evaluate the MAE, MRE, and MSE of six ap-
dC R(A, B) � CR(B) − CR(A), (10) proximate polynomials RI � RIm(n); m � 1, 6.
Step 3. Select the suitable approximate polynomial
where CR(A) and CR(B) are the CR values of the RI � RIr(n) with acceptable MAE, MRE, and MSE from
primary PM A and the reconstituted PM B, among six approximate polynomials RI � RIm(n); m �
respectively. 1, 6.
(ii) Matrix deviation,
n n 􏼌
􏼌 􏼌􏼌 2.3. Constituting Method of Final PM Constituting Method
dM(A, B) � 􏽘 􏽘􏼌􏼌􏼌aij − bij 􏼌􏼌􏼌. (11) and Final Attribute Weighting Method Considered Decision
i�1 j�1 Makers’ Levels. Let M be the number of the decision makers.
,e main steps to constitute the final PM and deter-
(iii) CR decreasing rate, mining the final attribute weights considered decision
makers’ levels are as follows:
dC R(A, B)
dC Rr(A, B) � . (12) Step 1. Constitute the simplest questionnaires by M
dM(A, B)
decision makers and constitute M PMs
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Mathematical Problems in Engineering 5

Table 2: RI value according to the number of attributes [28].

n 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
RI 0.0 0.0 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49 1.51 1.53 1.56 1.57 1.59

{A(m) � (a(m)ij )n×n ; m � 1, M } from the simplest Step 7. Constitute the composite PM B(0) � (b(0)
ij )n×n as
questionnaires, where A(m) � (a(m) ij )n×n is the PM the geometric mean of the individual PMs
constituted from m-th decision maker’s questionnaire. {B(m) � (b(m)
ij )n×n ; m � 1, M} as follows:
Step 2. Test the consistency of the PMs. M
hm
Step 3. Reconstitute the consistent PM by modifying b(0) (m)
ij � 􏽙 􏼐bij 􏼑 , i, j � 1, n. (16)
m�1
the inconsistent PM using the CR decreasing rate-based
consistency improvement method.
,e elements of B(0) � (b(0) ij )n×n may be no 9-point
Denote the consistent PMs as B(m) � (b(m) ij )n×n ; m � scales.
1, M.
Step 8. Constitute the final PM A(0) � (a(0) ij )n×n by
Step 4. Calculate the CR value CR(m) of the PM B(m) � transforming the elements of B(0) � (b(0) ij n×n to the
)
(b(m)
ij )n×n ; m � 1, M. nearest neighbouring 9-point scales.
Step 5. Calculate the inverse values of the CR values ,e final PM A(0) � (a(0) ij )n×n reflects not only the
{ICR(m) � 1/CR(m); m � 1, M.}. decision makers’ pairwise judgments but also their
Step 6. Normalize the inverse values of the CR values levels.
{ICR(m); m � 1, M} and determine the normalized Step 9. Determine the final attribute weights
values as the decision makers’ levels {hm; m � 1, M}. (w1 , . . . , wj , . . . , wn )T by normalizing the principal
eigenvector (v � v1 , . . . , vj , . . . , vn )T of the final PM
ICR(m)
hm � (j)
. (15) A(0) � (a(0) ij )n×n .
􏽐M
j�1 ICR
Figure 1 shows the flowchart of the proposed methods.
Commonly, the greater the CR value is, the worse the
consistency of the PM is. ,e worse the consistency of 3. Results and Discussion
the PM is, the more mistakes the decision maker’s 3.1. Numerical Test Results of the Proposed Consistency Im-
judgment has. ,erefore, we can regard that the inverse provement Method. We test the performance of proposed
value of the CR value of the PM reflects the decision consistency improvement method of PM by applying it to
maker’s level. ,is is why we assign the normalized two examples and compare with the previous methods.
inverse values of CR values to the decision makers’
levels. Example 1. ,e primary PM [21, 22] is as follows:

1 1
1 5 3 7 6 6
⎜
⎛
⎜
⎜
⎜ 3 4 ⎟ ⎞
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ 1 1 1 ⎟
⎜
⎜
⎜
⎜ 1 5 3 3 1/7 ⎟
⎟
⎟
⎟
⎟
⎜
⎜
⎜ 5 3 5 ⎟
⎟
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ ⎟
⎜
⎜
⎜
⎜
⎜ 1 1 ⎟⎟
⎟
⎟
⎟
⎜
⎜ 3 1 6 3 4 6 ⎟
⎟
⎜
⎜
⎜
⎜ 3 5 ⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎜ ⎟
⎜
⎜
⎜
⎜ 1 1 1 1 1 1 1 ⎟ ⎟
⎟
⎟
⎟
⎜
⎜
⎜ 1 ⎟
⎟
⎜
⎜
⎜
⎜ 3 5 6 3 4 7 8 ⎟ ⎟
⎟
⎟
⎟ , CR � 0.169087. (17)
⎜
⎜
⎜ ⎟
⎟
⎜ ⎟
⎜
⎜
⎜
⎜ 1 1 1 1 1 1 ⎟ ⎟
⎟
⎟
⎟
⎜
⎜ 3 1 ⎟
⎜
⎜
⎜
⎜
⎜ 6 3 3 2 5 6 ⎟ ⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ ⎟
⎜
⎜
⎜
⎜ 1 1 1 1 1 ⎟ ⎟
⎟
⎟
⎜
⎜ 4 2 1 ⎟
⎟
⎜
⎜
⎜
⎜ 6 2 4 5 6 ⎟ ⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎜
⎜ 1 1 ⎟⎟
⎟
⎟
⎟
⎜
⎜ 3 5 7 5 5 1 ⎟
⎟
⎜
⎜
⎝ 6 2 ⎟⎟
⎟
⎠
4 7 5 8 6 6 2 1
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
6 Mathematical Problems in Engineering

Select the attributes suited to given MADM problems

Select the decision makers that take part in the pairwise

comparison between the attributes

Record the simplest questionnaires by all the decision

makers

Constitute PMs from the simplest questionnaires from all

the decision makers

Determine the RI value according to the number of

attributes using the approximate polynomial

Calculate the CR values of all the PMs and test the

consistency

Reconstitute the inconsistent PMs using the CR

decreasing rate-based consistency improvement method

Constitute the final PM and determine the final attribute

weights considered decision makers’ levels

Figure 1: Flowchart of the proposed methods.

,e reconstituted PM using the ANTAHP method [21] is

as follows:

1 1
1 4.2 2.4 7.4 5.8 5.4
⎜
⎛
⎜
⎜
⎜
⎜ 2.4 3.4 ⎟
⎞
⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎜ ⎟
⎜
⎜
⎜
⎜ 1 1 1 ⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ 1 4.4 2.6 2.4 1/4 ⎟
⎟
⎜
⎜ 4.2
⎜
⎜ 3 6.8 ⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜ ⎟
⎜
⎜
⎜
⎜
⎜ 1 1 ⎟⎟
⎟
⎟
⎟
⎜
⎜ 3 1 6.8 3.4 4.2 4.4 ⎟
⎟
⎜
⎜
⎜
⎜ 2.4 4.2 ⎟
⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜ ⎟
⎜
⎜
⎜
⎜
⎜ 1 1 1 1 1 1 1 ⎟ ⎟
⎟
⎟
⎟
⎜
⎜ 1 ⎟
⎟
⎜
⎜
⎜
⎜ 7.4 4.4 6.8 2.8 3.4 7 9 ⎟ ⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟ , CR � 0.099838. (18)
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ 1 1 1 1 1 1 ⎟ ⎟
⎟
⎜
⎜ 2.8 1 ⎟
⎟
⎜
⎜
⎜
⎜
⎜ 5.8 2.6 3.4 1.8 5 6.8 ⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎜ 1 1 1 1 1 ⎟ ⎟
⎟
⎟
⎟
⎜
⎜ 3.4 1.8 1 ⎟
⎟
⎜
⎜
⎜
⎜
⎜ 5.4 2.4 4.2 5 6.8 ⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜ ⎟
⎜
⎜
⎜
⎜ 1 1 ⎟⎟
⎟
⎟
⎜
⎜ 2.4 4 7 5 5 1 ⎟
⎟
⎜
⎜
⎜
⎜
⎜ 4.4 2 ⎟⎟
⎟
⎟
⎟
⎜
⎜
⎝ ⎟
⎟
⎠
3.4 6.8 4.2 9 6.8 6.8 2 1
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Mathematical Problems in Engineering 7

,e reconstituted PMs using Wu method [22] are as In case of d � 0.1 (where d is the fixed modifying
follows: amount),

1 4.3 2.3 77 53 5.3 0.4545 0.303

⎜
⎛
⎜ ⎟
⎞
⎟
⎜
⎜
⎜
⎜
⎜ 023256 1 1 03333 43 0.2326 0.1587 ⎟
2.3 ⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎜
⎜
⎜
⎜ 043478 3.003 30003 1 67 4.3 5.3 0.2326 ⎟⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎜
⎜
⎜
⎜ 0.12987 0.23256 023256 014925 1 07692 0.1471 0.1149 ⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟, CR � 0.099955. (19)
⎜ ⎟
⎜
⎜
⎜
⎜ 0.18868 043478 043478 027027 22999 1 0.2326 0.1493 ⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎜
⎜
⎜
⎜ 0.18868 043478 023256 33003 13001 1 0.2326 0.1493 ⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜ ⎟
⎜
⎜
⎜
⎝ 2.2002
⎜ 42992 018868 67981 42992 42992 1 0.3846 ⎟
⎟
⎟
⎟
⎠
3.3003 63012 42992 87032 66979 66979 2.6001 1

In case of d � 0.01,

1 4.28 2.28 7.72 5.28 5.28 0.4386 0.3049

⎜
⎛
⎜
⎜ ⎟
⎞
⎟
⎟
⎜
⎜
⎜
⎜ 0.23364 1 0.3367 4.28 2.28 0.23360 0.1592 ⎟
2.28 ⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ ⎟
⎜
⎜
⎜
⎜
⎜ 0.4386 2.97 1 6.72 3.72 4.32 5.28 0.2336 ⎟⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ ⎟
⎜
⎜
⎜
⎜
⎜ 0.12953 0.23364 0.14881 1 0.4386 0.3049 0.1484 0.1147 ⎟⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟ , CR � 0.099825. (20)
⎜
⎜ ⎟
⎜
⎜
⎜
⎜ 0.18939 0.4386 0.26882 2.28 1 0.7812 0.2336 0.1488 ⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎜
⎜ ⎟
⎟
⎜
⎜
⎜
⎜ 0.18939 0.4386 0.23148 3.27981 1.2801 1 0.2336 0.1488 ⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎜
⎜ ⎟
⎟
⎜
⎜
⎝ 2.28
⎜ 4.2808 0.18939 6.7385 4.2808 4.2808 1 0.3802 ⎟
⎟
⎟
⎟
⎠
3.2798 6.2814 4.2808 8.7184 6.7204 6.7204 2.6302 1

In case of d � 0.001,

1 4.282 2.282 7.718 5.282 5.282 0.4382 0.3047

⎜
⎛
⎜
⎜ ⎟
⎞
⎟
⎟
⎜
⎜
⎜
⎜ 0.23354 1 0.3364 4.282 2.282 2.282 0.2335 0.1592 ⎟⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜ ⎟
⎜
⎜
⎜
⎜
⎜ 0.43821 2.9727 1 6.718 3.718 4.328 5.282 0.2335 ⎟ ⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ ⎟
⎜
⎜
⎜
⎜
⎜ 0.12957 0.23354 0.14885 1 0.4382 0.3047 0.1483 0.1147 ⎟
⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟ , CR � 0.099997. (21)
⎜
⎜ ⎟
⎜
⎜
⎜
⎜ 0.18932 0.43821 0.26896 2.2821 1 0.78 0.2335 0.1489 ⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎜
⎜ ⎟
⎟
⎜
⎜
⎜
⎜ 0.18932 0.43821 0.23105 3.2819 1.2821 1 0.2335 0.1489 ⎟
⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎜
⎜ ⎟
⎟
⎜
⎜
⎝ 2.2821
⎜ 4.2827 0.18932 6.7431 4.2827 4.2827 1 0.3802 ⎟
⎟
⎟
⎟
⎠
3.2819 6.2814 4.2827 8.7184 6.7159 6.7159 2.6302 1

,e reconstituted PM using Xu and Wei method [23] is

as follows:
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
8 Mathematical Problems in Engineering

1 4.524 2.339 7.523 5.888 5.686 0.425 0.292

⎜
⎛
⎜ ⎟
⎞
⎟
⎜
⎜
⎜
⎜ 0.221 1 0.326 4.516 2.671 2.580 0.222 0.147 ⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ ⎟
⎜
⎜
⎜
⎜ 0.427 3.067 1 6.749 3.460 4.188 4.155 0.249 ⎟ ⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎜
⎜ 0.133 0.221 0.148 1 0.373 0.287 0.134 0.104 ⎟ ⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟ , CR � 0.096964. (22)
⎜
⎜
⎜
⎜
⎜
⎜ 0.170 0.374 0.289 2.681 1 0.561 0.197 0.147 ⎟ ⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎜
⎜ 0.176 0.388 0.234 3.479 1.784 1 0.204 0.153 ⎟ ⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜
⎝ 2.354
⎜ 4.497 0.241 7.479 5.073 4.899 1 0.501 ⎟ ⎟
⎟
⎠
3.419 6.786 4.024 9.624 6.783 6.551 1.996 1

,e reconstituted PM using Cao method [24] is as

follows:

1 4.44412 2.3682 7.6743 5.8559 5.6079 0.4201 0.2968

⎜
⎛
⎜ ⎟
⎞
⎟
⎜
⎜
⎜
⎜ 0.2252 1 0.3210 4.4224 2.6175 3.5392 0.2268 0.1486 ⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎜ ⎟
⎜
⎜
⎜
⎜
⎜ 0.4223 3.1151 1 6.9149 3.5351 4.2774 4.5105 0.2487 ⎟
⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎜ 0.1303 0.2261 0.1446 1 0.3805 0.2927 0.1313 0.1030 ⎟
⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟ , CR � 0.107316. (23)
⎜
⎜
⎜
⎜
⎜ 0.1708 0.3820 0.2829 2.6278 1 0.5722 0.1960 0.1444 ⎟
⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎜ 0.1783 0.3938 0.2338 3.4166 1.7478 1 0.2055 0.1494 ⎟
⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎝ 2.3804
⎜ 4.4099 0.2217 7.6136 5.1012 4.8661 1 0.5004 ⎟
⎟
⎟
⎠
3.3689 6.7293 4.0215 9.7130 6.9235 6.6949 1.9984 1

,e reconstituted PM using the proposed method is as

follows (CR0 � 0.1.):

1 1
1 5 3 7 6 6
⎜
⎛
⎜
⎜
⎜ 3 4⎟⎞
⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎜
⎜
⎜
⎜
⎜ 1 1 1 1⎟⎟
⎟
⎟
⎟
⎜
⎜ 1 5 3 3 ⎟
⎟
⎜
⎜
⎜
⎜ 5 3 5 7⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎜ ⎟
⎜
⎜
⎜ 1 1⎟⎟
⎟
⎜
⎜
⎜ 3 1 6 3 4 1 ⎟ ⎟
⎟
⎜
⎜
⎜
⎜ 3 5⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ 1 1 1 1 1 1 1⎟⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎜ 1 ⎟
⎜
⎜
⎜
⎜
⎜ 7 5 6 3 4 7 8⎟⎟
⎟
⎟
⎟ , CR � 0.087973. (24)
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎜
⎜ 1 1 1 1 1 1⎟⎟
⎟
⎜
⎜
⎜ 3 1 ⎟
⎟
⎜
⎜
⎜
⎜ 6 3 3 2 5 6⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎜
⎜
⎜
⎜
⎜ 1 1 1 1 1⎟⎟
⎟
⎟
⎟
⎜
⎜ 4 2 1 ⎟
⎟
⎜
⎜
⎜
⎜ 6 3 4 5 6⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ 1⎟⎟
⎜ ⎟
⎜
⎜
⎜ 3 5 1 7 5 5 1 ⎟ ⎟
⎟
⎜
⎜
⎝ 2⎟⎟
⎟
⎠
4 7 5 8 6 6 2 1

Table 3 shows the comparison results between some and 0.013905 by modifying only two elements with the dM of
previous and proposed consistency improvement methods. 5.833333 using the proposed method. ,e CR, dCR, and
,e CR, dCR, and dCRr are, respectively, 0.087973, 0.081114, dCRr are, respectively, 0.096964, 0.072123, and 0.004792 by
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Mathematical Problems in Engineering 9

Table 3: Comparison results between some consistency improvement methods about Example 1.
Wu et al. [22] Proposed
Performance measure ANT-AHP [21] Zeshui and Cuiping [23] Cao et al. [24]
d � 0.1 d � 0.01 d � 0.001 method
CR 0.099838 0.099955 0.099825 0.099997 0.096964 0.107316 0.087973
dCR 0.069248 0.069130 0.069262 0.069090 0.072123 0.061771 0.081114
dM 14.842974 19.457000 20.009000 19.956000 15.050557 16.605557 5.833333
dCRr 0.004665 0.003553 0.0034615 0.003462 0.004792 0.003720 0.013905
Number of modified elements 46 54 56 56 56 56 2

modifying 56 elements with the dM of 15.050557 using the 1 1 1

previous Cao’s method. ⎜
⎛ 1 1 1 1⎟⎞
⎜
⎜
⎜ 3 5 4 ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
Example 2. ,e primary inconsistent PM [26] is as follows: ⎜
⎜
⎜ 1 ⎟
⎟
⎜
⎜
⎜
⎜ 3 1 2 1 3 3⎟⎟
⎟
⎟
⎜
⎜ 2 ⎟
⎟
⎟
1 1 1 ⎜
⎜
⎜ ⎟
⎟
⎟
⎜ 1 1 2 3⎟⎞ ⎜
⎜
⎜ ⎟
⎟
⎟
⎛
⎜
⎜ 3 5 4 ⎟
⎟ ⎜
⎜ ⎟
⎜
⎜
⎜
⎜
⎟
⎟
⎟
⎟
⎜
⎜
⎜
⎜ 5 2 1 4 2 6 5⎟⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜
⎜ ⎟
⎟
⎜
⎜ 1 1 ⎟ ⎜
⎜ ⎟
⎟
⎜
⎜
⎜ 3 1 2 3 3⎟⎟
⎟
⎟ ⎜
⎜
⎜ 1 1 1 ⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎜ 2 3 ⎟
⎟
⎟
⎟
⎜
⎜
⎜
⎜ 1 1 1 1⎟⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜
⎜ 2 4 4 ⎟
⎟
⎟ , CR � 0.009261. (27)
⎜
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜
⎜ ⎟
⎟
⎜
⎜ ⎟ ⎜
⎜ ⎟
⎟
⎜
⎜
⎜
⎜ 5 2 1 4 5 6 5⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎟
⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟ ⎜
⎜ 1 ⎟
⎜
⎜
⎜
⎜
⎜
⎟
⎟
⎟
⎟
⎜
⎜
⎜
⎜ 4 1 4 1 3 2⎟⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜
⎜ 2 ⎟
⎟
⎟
⎜
⎜ 1 1 1 ⎟ ⎜
⎜ ⎟
⎜
⎜
⎜
⎜ 1 1 1 2⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎟
⎟
⎟
⎟
⎟
⎜
⎜
⎜ 2 4 4 ⎟
⎟
⎟ , CR � 0.112090. (25) ⎜
⎜
⎜ ⎟
⎟
⎜
⎜ ⎟
⎟ ⎜
⎜ 1 1 1 ⎟
⎟
⎜
⎜
⎜
⎜
⎜
⎟
⎟
⎟
⎟
⎜
⎜
⎜
⎜ 1 1 1 1⎟⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜
⎜ 3 6 3 ⎟
⎟
⎟
⎜
⎜ 1 ⎟ ⎜
⎜ ⎟
⎜
⎜
⎜
⎜ 4 3 4 1 3 1⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎟
⎟
⎟
⎟
⎟
⎜
⎜
⎜ 5 ⎟
⎟
⎟ ⎜
⎜
⎝ ⎟
⎠
⎜
⎜
⎜ ⎟
⎟
⎟ 1 1 1
⎜
⎜
⎜ ⎟
⎟
⎟ 1 1 1 1
⎜
⎜ ⎟ 3 5 2
⎜
⎜
⎜ 1 1 1 1 1⎟⎟
⎟
⎜
⎜
⎜
⎜ 1 1 ⎟ ⎟
⎟
⎟
⎜
⎜
⎜
⎜ 2 3 6 3 3⎟⎟
⎟
⎟ In case of CR0 � 0.005,
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟ 1 1 1
⎜
⎜
⎝1 ⎟
⎠ 1 1 1 1⎟
1 1 1 ⎜
⎛
⎜ 3 5 4 ⎞
⎟
⎟
1 3 1 ⎜
⎜
⎜ ⎟
⎟
3 3 5 2 ⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜ 1
,e reconstituted PM using the Benı́tez method [26] is as ⎜
⎜
⎜
⎜
⎜ 3 1 2 1 3 3⎟⎟
⎟
⎟
⎟
follows: ⎜
⎜ 2 ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
1 0.526 0.154 0.794 0.471 1.738 1.17 ⎜
⎜ ⎟
⎜
⎛
⎜ ⎟
⎞
⎟
⎜
⎜
⎜
⎜ 5 2 1 4 2 6 5⎟⎟
⎟
⎟
⎟
⎜ ⎟
⎜
⎜
⎜
⎜
⎜ 1.902 1 0.293 1.510 0.896 3.306 2.225 ⎟
⎟
⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎜
⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜ ⎟
⎟
⎟
⎜
⎜ ⎟ ⎜
⎜ 1 1 1 ⎟
⎟
⎜
⎜
⎜
⎜
⎜ 6.487 3.411 1 5.149 3.055 11.28 7.590 ⎟ ⎟
⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎜ 1 1 1 1⎟⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜ 2 4 3 ⎟
⎟
⎟ , CR � 0.0004667. (28)
⎜
⎜ ⎟ ⎜
⎜ ⎟
⎟
⎜
⎜
⎜
⎜ 1.260 0.662 0.194 1 0.593 2.190 1.474 ⎟ ⎟
⎟
⎟
⎟ , ⎜
⎜
⎜
⎜
⎜
⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟ (26) ⎜
⎜ ⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜
⎜ 1 ⎟
⎟
⎜
⎜
⎜ 2.123 1.116 0.327 1.685 1 3.691 2.484 ⎟ ⎟ ⎜
⎜ 4 1 3 1 3 3⎟⎟
⎟
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜
⎜ ⎟
⎟
⎜
⎜
⎜ ⎟
⎟ ⎜
⎜ 2 ⎟
⎟
⎟
⎜ ⎟ ⎜ ⎟
⎜
⎜
⎝ 0.575
⎜ 0.302 0.089 0.457 0.271 1 0.673 ⎟ ⎟
⎟
⎟
⎠
⎜
⎜
⎜
⎜
⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ 1 1 1 ⎟
⎟
0.855 0.449 0.132 0.678 0.403 1.486 1 ⎜
⎜
⎜
⎜ 1 1 1 1⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ 3 6 3 ⎟
⎟
⎟
CR � 0.099677. ⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎝ ⎟
⎠
,e reconstituted PMs using the proposed method are as 1 1 1
1 1 1 1
follows: 3 5 3
In case of CR0 � 0.01, In case of CR0 � 0.001,
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
10 Mathematical Problems in Engineering

1 1 1 CR, dCR, and dCRr are, respectively, 0.000007693, 0.112082,

⎜
⎛ 1 1 1 1⎟⎞ and 0.003194 by modifying 42 elements with the dM of
⎜
⎜
⎜ 3 5 3 ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟ 35.096667 using the previous Benı́tez’s method.
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟ As can be seen in Tables 3 and 4, the proposed method is
⎜
⎜ 1 ⎟
⎜
⎜
⎜
⎜
⎜ 3 1 3 1 3 3⎟⎟
⎟
⎟
⎟ much better than the previous ones from the viewpoints of
⎜
⎜ 2 ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟ CR, dCR, dM, dCRr, and the number of modified elements.
⎜
⎜
⎜ ⎟
⎟
⎜
⎜ ⎟
⎟
⎜
⎜
⎜
⎜ 5 2 1 5 2 6 5⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟ 3.2. Approximate Polynomial for Random Consistency Index
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ 1 1 1 ⎟
⎟ (RI) According to the Number of Attributes. We develop
⎜
⎜
⎜
⎜ 1 1 1 1⎟⎟
⎟
⎟
⎟ approximate polynomial for RI according to the number of
⎜
⎜
⎜ 3 5 3 ⎟
⎟
⎟ , CR � 0.000772. (29)
⎜
⎜
⎜ ⎟
⎟ attributes by using the proposed method in Section 3.2.
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟ Table 5 and Figures 2 and 3 show the performance test
⎜
⎜
⎜ 1 ⎟
⎟
⎜
⎜
⎜
⎜ 3 1 3 1 3 3⎟⎟
⎟
⎟
⎟ results of approximate polynomials for RI with different
⎜
⎜
⎜ 2 ⎟
⎟
⎜
⎜ ⎟
⎟
⎟ orders.
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟ From Table 5 and Figures 2 and 3, we can know that
⎜
⎜ 1 1 1 ⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎜ 1 1 1 1⎟⎟
⎟
⎟
⎟
third-order approximate polynomial is appropriate from the
⎜
⎜
⎜ 3 6 3 ⎟
⎟ viewpoints of the performance and complexity.
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟ ,ird-order approximate polynomial for RI is as follows:
⎜
⎜ ⎟
⎟
⎟
⎜
⎝ 1 1 1 ⎠
1 1 1 1 RI � 0.001311n3 − 0.045332n2 + 0.533262n − 0.615824.
3 5 3
(30)
Table 4 shows the comparison result between Benı́tez
method [26] and the proposed method. Table 6 shows the performance test result of the above
,e CR, dCR, and dCRr are, respectively, 0.000772, third-order approximate polynomial.
0.111318, and 0.005214 by modifying 22 elements with the As the result, the approximate formula of CR is as
dM of 21.35 using the proposed method (CR0 � 0.001). ,e follows:

CI λmax − n λmax − n􏼁
CR � � � . (31)
RI (n − 1)RI 􏽨(n − 1)􏼐0.00311n3 − 0.045332n2 + 0.533262n − 0.615824􏼑􏽩

By using (30) and (31), we can directly calculate the RI unalloyed titanium (A7), Ti-6Al-4V (A8), composites
value and CR value without the numerical table for RI values (fabric reinforced)-epoxy-70% glass (A9), composites
according to the number of attributes and evaluate the (fabric reinforced)-epoxy-63% carbon (A10), and com-
consistency of the constituted PM rapidly. posites (fabric reinforced)-epoxy-62% aramid (A11). ,e
On the other hand, we test the performance of the material attributes are, respectively, tissue tolerance (TT),
following previous approximate formula for RI: [4]. corrosion resistance (CR), tensile strength (TS) (MPa),
1.98(n − 2) fatigue strength (FS) (MPa), relative toughness (RT), rel-
RI � . (32)
n ative wear resistance (RWR), elastic modulus (EM) (GPa),
Table 7 shows the performance test result of the previous specific gravity (SG) (g/cc), and cost (C). Table 8 shows
approximate formula (32). some hip joint prosthesis materials and their properties.
From Tables 6 and 7, we know that the proposed third- [27].
order approximate polynomial (30) has much better per- First, we conduct the correlation analysis between the
formance than the previous approximation formula (32). properties of hip joint prosthesis materials.
Table 9 shows the correlation coefficients between
properties of hip joint prosthesis materials. As shown in
3.3. Application to Hip Joint Prosthesis Material Selection. Table 9, the elastic modulus and specific gravity have the
We use the proposed consistency improvement method of high correlation coefficients with the tissue tolerance, fatigue
PM based on CR decreasing rate to hip joint prosthesis strength, and relative wear resistance.
material selection [27]. ,erefore, we remove two properties such as EM (elastic
,e alternative materials for hip joint prosthesis are, modulus) and SG (specific gravity) from the consideration.
respectively, stainless steels 316 (A1), stainless steels 317 Consequently, we constitute the decision matrix with
(A2), stainless steels 321 (A3), stainless steels 347 (A4), Co- seven attributes: TT, CR, TS, FS, RT, RWR, and C. Table 10
Cr alloys-cast alloy (A5), Co-Cr alloys-wrought alloy (A6), shows the decision matrix.
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Mathematical Problems in Engineering 11

Table 4: Comparison result between Benı́tez method and the proposed method about Example 2.

Benı́tez method [26] Proposed method

Performance measure
CR0 � 0.005 CR0 � 0.01 CR0 � 0.005 CR0 � 0.001
CR 0.000007693 0.009261 0.004667 0.000772
dCR 0.112082 0.10283 0.10742 0.111318
dM 35.096667 15.467 18.05 21.35
dCRr 0.003194 0.006648 0.0059514 0.005214
Number of modified elements 42 14 16 22

Table 5: Performance test results of approximate polynomials for RI with different orders.
Mean absolute Maximum of absolute Mean relative error Maximum of relative Mean-squared error Maximum of
m
error errors (%) errors (%) (%) squared errors (%)
1 0.11101 0.33637 10.756 57.995 0.018312 0.11315
2 0.046806 0.11791 4.3837 20.33 0.0031264 0.013903
3 0.016633 0.038921 1.5316 5.4092 0.00040314 0.0015149
4 0.0078826 0.0228 0.6634 1.7273 9.4729·10−5 0.00051983
5 0.0056381 0.014733 0.42149 1.1161 4.8904·10−5 0.00021705
6 0.0049297 0.01272 0.3941 0.96365 3.9756·10−5 0.0001618

m = 1: MAE = 0.11101, MRE = 10.7556, MSE = 0.018312 m=2: MAE=0.046806, MRE=4.3837, MSE=0.0031264
2 2

1.5 1.5
RI

RI

1 1

0.5 0.5
2 4 6 8 10 12 14 16 2 4 6 8 10 12 14 16
n n
(a) (b)
n
m=3: MAE=0.016633, MRE=1.5316, MSE=0.00040314 m=4: MAE=0.0078826, MRE=0.6634, MSE=9.4729e−05
2 2

1.5 1.5
RI

RI

1 1

0.5 0.5
2 4 6 8 10 12 14 16 2 4 6 8 10 12 14 16
n n
(c) (d)
m=5: MAE=0.0056381, MRE=0.42149, MSE=4.8904e−05 m=6: MAE=0.0049297, MRE=0.3941, MSE=3.9756e−05
2 2

1.5 1.5
RI

RI

1 1

0.5 0.5
2 4 6 8 10 12 14 16 2 4 6 8 10 12 14 16
n n
(e) (f )

Figure 2: Performance test results of approximate polynomials for RI with different orders: (a) m � 1, (b) m � 2, (c) m � 3, (d) m � 4,
(e) m � 5, and (f ) m � 6.
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
12 Mathematical Problems in Engineering

0.12 12 0.02

0.1 10
0.015
0.08 8

MRE (%)
MAE

MSE
0.06 6 0.01

0.04 4
0.005
0.02 2

0 0 0
0 2 4 6 0 2 4 6 0 2 4 6
m m m
(a) (b) (c)

Figure 3: MAE, MRE, and MSE values of the approximate polynomials for RI according to different orders: (a) MAE, (b) MRE, and
(c) MSE.

Table 6: Performance test result of the third-order approximate polynomial for RI (equation (21)).
n RI Calculated RI Absolute error Relative error (%) Squared error
3 0.58 0.61137 0.031374 5.4092 0.0009843
4 0.9 0.87582 0.024176 2.6862 0.00058447
5 1.12 1.0811 0.038921 3.4751 0.0015149
6 1.24 1.235 0.004995 0.40282 2.495·10−5
7 1.32 1.3455 0.02547 1.9295 0.0006487
8 1.41 1.4203 0.01034 0.73331 0.00010691
9 1.45 1.4675 0.017483 1.2057 0.00030564
10 1.49 1.4948 0.0047652 0.31981 2.2707·10−5
11 1.51 1.5101 5.4945·10−5 0.0036387 3.019·10−9
12 1.53 1.5212 0.0087812 0.57394 7.711·10−5
13 1.56 1.5361 0.023876 1.5305 0.00057007
14 1.57 1.5626 0.0073626 0.46896 5.4208·10−5
15 1.59 1.6086 0.018626 1.1715 0.00034694
Mean 0.016633 1.5316 0.00040314
Max 0.038921 5.4092 0.0015149

Table 7: Performance test result of the previous approximate formula for RI (equation (23)).
n RI Calculated RI Absolute error Relative error (%) Squared error
3 0.58 0.66 0.08 13.793 0.0064
4 0.9 0.99 0.09 10 0.0081
5 1.12 1.188 0.068 6.0714 0.004624
6 1.24 1.32 0.08 6.4516 0.0064
7 1.32 1.4143 0.094286 7.1429 0.0088898
8 1.41 1.485 0.075 5.3191 0.005625
9 1.45 1.54 0.09 6.2069 0.0081
10 1.49 1.584 0.094 6.3087 0.008836
11 1.51 1.62 0.11 7.2848 0.0121
12 1.53 1.65 0.12 7.8431 0.0144
13 1.56 1.6754 0.11538 7.3964 0.013314
14 1.57 1.6971 0.12714 8.0983 0.016165
15 1.59 1.716 0.126 7.9245 0.015876
Mean 0.097678 7.6801 0.00991
Max 0.12714 13.793 0.016165
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Mathematical Problems in Engineering 13

Table 8: Alternative materials and their attribute values [27].

Alternative material TT CR TS FS RT RWR EM SG C
A1 10 7 517 350 8 8 200 8 1
A2 9 7 630 415 10 8.5 200 8 1.1
A3 9 7 610 410 10 8 200 7.9 1.1
A4 9 7 650 430 10 8.4 200 8 1.2
A5 10 9 655 425 2 10 238 8.3 3.7
A6 10 9 896 600 10 10 242 9.1 4
A7 8 10 550 315 7 8 110 4.5 1.7
A8 8 10 985 490 7 8.3 124 4.4 1.9
A9 7 7 680 200 3 7 22 2.1 3
A10 7 7 560 170 3 7.5 56 1.6 10
A11 7 7 430 130 3 7.5 29 1.4 5

Table 9: Correlation coefficients between properties of hip joint prosthesis materials.

TT CR TS FS RT RWR EM SG C
TT 1.000 0.161 0.229 0.780 0.530 0.792 0.966 0.961 −0.479
CR 1.000 0.570 0.476 −0.008 0.484 0.175 0.101 −0.127
TS 1.000 0.729 0.287 0.465 0.300 0.268 −0.157
FS 1.000 0.681 0.792 0.848 0.832 −0.492
RT 1.000 0.251 0.624 0.688 −0.624
RWR 1.000 0.815 0.744 −0.129
EM 1.000 0.985 −0.472
SG 1.000 −0.576
C 1.000
In this table, the bold values indicate higher correlation coefficients (ñ>0.8) than others.

1 1 5 5 3 5 5
Table 10: Decision matrix. ⎜
⎛
⎜
⎜ ⎟
⎞
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜ ⎟
Alternative materials TT CR TS FS RT RWR C
⎜
⎜
⎜
⎜ 1 1 5 5 3 5 5⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
A1 10 7 517 350 8 8 1 ⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
A2 9 7 630 415 10 8.5 1.1 ⎜
⎜
⎜
⎜ 1 1 1 1 1 1⎟⎟
⎟
⎟
⎟
⎜
⎜ 1 ⎟
A3 9 7 610 410 10 8 1.1 ⎜
⎜
⎜
⎜
⎜ 5 5 5 5 3 5⎟⎟
⎟
⎟
⎟
A4 9 7 650 430 10 8.4 1.2 ⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
A5 10 9 655 425 2 10 3.7 ⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ 1 1 1 ⎟
A6 10 9 896 600 10 10 4 ⎜
⎜
⎜
⎜ 5 1 3 5⎟⎟
⎟
⎟
⎟
A7 8 10 550 315 7 8 1.7 A (1) ⎜
�⎜
⎜ 5 5 3 ⎟
⎟ . (33)
⎜
⎜ ⎟
⎟
⎟
A8 8 10 985 490 7 8.3 1.9 ⎜
⎜
⎜ ⎟
⎟
⎜
⎜ ⎟
⎟
⎟
A9 7 7 680 200 3 7 3 ⎜
⎜
⎜ 1 1 ⎟
⎟
⎟
⎜
A10 7 7 560 170 3 7.5 10
⎜
⎜
⎜
⎜ 5 3 1 3 5⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ 3 3 ⎟
⎟
⎟
A11 7 7 430 130 3 7.5 5 ⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎜
⎜ 1 1 1 1 ⎟
⎟
⎜
⎜
⎜
⎜ 3 1 3⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ 5 5 3 3 ⎟
⎟
⎟
As shown in Table 10, there are 11 alternative materials ⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
with 7 attributes. Excepting C (cost), six attributes are benefit ⎜
⎜
⎝1 ⎟
⎟
⎠
1 1 1 1
attributes among seven attributes, while C (cost) is cost 5 1
5 5 5 5 3
attribute.
,e material attribute weights are determined by using
,e consistency test result of the PM A(1) is as follows:
the AHP with simplest questionnaire by three decision
makers. Table 11 shows the simplest pairwise comparison λmax � 7.913828, CI � 0.152305, CR � 0115382 > 01. (34)
questionnaire completed by the first decision maker.
From Table 11, the first decision maker’s PM is as ,erefore, we should modify the PM A(1) until CR < 0.1
follows: using the proposed consistency improvement method.
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
14 Mathematical Problems in Engineering

Table 11: Simplest pairwise comparison questionnaire completed by the first decision maker.
Tissue Corrosion Tensile Fatigue Relative Relative wear
Cost
tolerance resistance strength strength toughness resistance
Tissue tolerance 1 1 5 5 3 5 5
Corrosion resistance 1 5 5 3 5 5
Tensile strength 1 5 5 3 5
Fatigue strength 1 3 3 5
Relative toughness 1 3 5
Relative wear
1 3
resistance
Cost 1
It indicates that the value is the maximum. I have no objection to remove the Bold type.

In the first iteration, we first reconstitute the PM by In the next iteration, we first reconstitute the PM B(1)
1 by
replacing all elements of A(1) with the lower neighbouring replacing all elements with the lower neighbouring scales
scales and calculate CR values and CR decreasing rates and calculate the CR values and CR decreasing rates
(Table 12). (Table 14).
From Table 12, the maximum CR decreasing rate is From Table 14, the maximum CR decreasing rate is
0.0077709 (CR � 0.10632) when we replace the element 0.0071198 (CR � 0.096187) when we replace the element
a67 � 3 in the primary PM with the lower neighbouring scale a67 � 3 with the lower neighbouring scale 2.
2. We next reconstitute the PM by replacing all elements
We next reconstitute the PM by replacing all elements of with the upper neighbouring scales and calculate the CR
A(1) with the upper neighbouring scales and calculate the CR values and CR decreasing rates (Table 15).
values and CR decreasing rates (Table 13). From Table 15, the maximum CR decreasing rate is
From Table 13, the maximum CR decreasing rate is 0.010142 (CR � 0.093506) when we replace the element
0.010371 (CR � 0.10449) when we replace the element a37 � 1/4 with the upper neighbouring scale 1/3. As
a37 � 1/5 in the primary PM with the upper neighbouring 0.0071198 < 0.010142, we decide to modify the element
scale 1/4. a37 � 1/4 to the upper neighbouring scale 1/3. ,e recon-
As 0.0077709 < 0.010371, we decide to modify the ele- stituted PM is as follows:
ment a37 � 1/5 of the primary PM to the upper neighbouring
1 1 5 5 3 5 5
scale 1/4. ,e reconstituted PM is as follows: ⎜
⎛ ⎟
⎞
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜1
⎜
⎜ 1 5 5 3 5 5⎟⎟
⎟
⎟
⎟
1 1 5 5 3 5 5 ⎜
⎜
⎜ ⎟
⎟
⎜
⎜ ⎟
⎟
⎟
⎜
⎛
⎜
⎜ ⎟
⎞
⎟
⎟ ⎜
⎜
⎜ ⎟
⎟
⎜
⎜
⎜
⎜
⎟
⎟
⎟
⎟
⎜
⎜
⎜
⎜ 1 1 1 1 1 1⎟⎟
⎟
⎟
⎟
⎜
⎜ 5⎟⎟ ⎜
⎜ 1 ⎟
⎜
⎜
⎜
⎜
1 1 5 5 3 5 ⎟
⎟
⎟
⎟
⎜
⎜
⎜
⎜ 5 5 5 5 3 3⎟⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜ ⎟
⎟
⎜ ⎜ ⎟
⎜
⎜
⎜
⎜ 1 1 1 1 1 1⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜
⎜ 1 1 1
⎟
⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎜ 5 5
1
5 5 3 4⎟
⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎜ 5 1 3 5⎟⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟ B (1) ⎜
⎜
�⎜ 5 5 3 ⎟
⎟
⎟ , CR(1) � 0.093506. < 0.1. (36)
⎜
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜ ⎟
⎟
⎜
⎜
⎜ ⎟
⎟ ⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜
⎜ ⎟
⎟
⎜
⎜ 1 1 1 ⎟ ⎜1 1 ⎟
⎟
⎜
⎜
⎜
⎜ 5 1 3 5⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜
⎜ 5 3 1 3 5⎟
⎟
⎟
⎟
B(1) ⎜
�⎜
⎜ 5 5 3 ⎟
⎟ , CI � 010449 > 01. (35) ⎜
⎜
⎜ 3 3 ⎟
⎟
⎟
1 ⎜
⎜ ⎟
⎟ ⎜
⎜ ⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ 1 1 ⎟
⎟
⎟ ⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ 5⎟⎟ ⎜ 1 1 1 1 ⎟
⎜
⎜
⎜
⎜ 3 3
5 3 1 3 ⎟
⎟
⎟
⎟
⎜
⎜
⎜
⎜ 3 1 3⎟⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟ ⎜
⎜
⎜ 5 5 3 3 ⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟ ⎜
⎜
⎝1 ⎟
⎟
⎠
⎜ 1 1 1 1 ⎟ 1 1 1 1
⎜
⎜
⎜
⎜ 3 1 3⎟⎟
⎟
⎟
⎟ 3 1
⎜
⎜
⎜ 5 5 3 3 ⎟
⎟ 5 5 5 5 3
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎝1 ⎟
⎟
⎠ As CR � 0.093506 < 0.1, the reconstituted PM B(1) sat-
1 1 1 1
4 1 isfies the consistency.
5 5 5 5 3
By the similar way, the reconstituted PMs based on the
As CR � 0.10449 > 0.1, we should modify the PM B(1)
1 PMs constituted by the second and third decision makers are
until CR < 0.1 again. as follows:
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Mathematical Problems in Engineering 15

Table 12: CR decreasing rates by replacing all elements with the lower neighbouring scales.
dC Rr−pq 2 3 4 5 6 7
1 −0.004171 −0.0077567 0.0038198 0.0042379 0.00012511 −0.0028697
2 −0.0077567 0.0038198 0.0042379 0.00012511 −0.0028697
3 −0.0022949 0.00080871 −0.0031613 −0.010273
4 −0.0077696 0.0061679 0.0062029
5 0.00035439 0.0018634
6 0.0077709
It indicates that the value is the maximum. I have no objection to remove the Bold type.

Table 13: CR decreasing rates by replacing all elements with the upper neighbouring scales.
dC Rr+pq 2 3 4 5
1 −0.004171 0.0051298 −0.004292 −0.0058214 −0.001012 0.0013276
2 0.0051298 −0.004292 −0.0058214 −0.001012 0.0013276
3 0.0014299 −0.0021998 0.0007321 0.010371
4 0.0063837 −0.0075791 −0.0065044
5 −0.0025883 −0.0025898
6 −0.0090527
It indicates that the value is the maximum. I have no objection to remove the Bold type.

Table 14: CR decreasing rates by replacing all elements with the lower neighbouring scales.
dC Rr−pq 2 3 4 5 6 7
1 −0.0042129 −0.007257 0.0037917 0.0042349 5.7868e − 05 −0.0032652
2 −0.007257 0.0037917 0.0042349 5.7868e − 05 −0.0032652
3 −0.0026355 0.00050903 −0.0036426 −0.010371
4 −0.0078255 0.0062676 0.0057371
5 0.00031694 0.0014975
6 0.0071198

Table 15: CR decreasing rates by replacing all elements with the upper neighbouring scales.
dC Rr+pq 2 3 4 5 6 7
1 −0.0042129 0.0047588 −0.0042749 −0.0058397 −0.00096572 0.0016549
2 0.0047588 −0.0042749 −0.0058397 −0.00096572 0.0016549
3 0.0018419 −0.0018135 0.0013729 0.010142
4 0.0064268 −0.0076851 −0.0060728
5 −0.0025853 −0.0022606
6 −0.0084573

1 1/3 7 5 3 5 5
⎜
⎛
⎜
⎜ ⎟
⎞
⎟
⎟
⎜
⎜
⎜
⎜
⎜ 3 1 5 5 3 5 7⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ 1 1 1 1 ⎟
⎟
⎜
⎜
⎜
⎜
⎜ 1 1 1⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ 7 7 3 5 ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ 1 1 1 ⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎜ 3 1 3 5⎟⎟
⎟
⎟
⎟
B(2) ⎜
⎜
�⎜
⎜ 5 5 3 ⎟
⎟
⎟ , CR(2) � 0.071796 < 0.1,
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ 1 1 ⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎜ 5 3 1 3 5⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ 3 3 ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ 1 1 1 1 ⎟
⎟
⎟
⎜
⎜
⎜
⎜ 1 1 3⎟⎟
⎟
⎟
⎜
⎜
⎜ 5 5 3 3 ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎝1 1 1 1 1 ⎟ ⎠
1 1
5 7 5 5 3
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
16 Mathematical Problems in Engineering

1 3 7 5 5 7 5
⎜
⎛
⎜
⎜ 1 ⎟
⎞
⎟
⎟
⎜ ⎟
⎜
⎜
⎜
⎜ 1 7 5 5 7 5⎟⎟
⎟
⎟
⎜
⎜
⎜ 3 ⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎜
⎜
⎜ 1 1 1 1 1⎟⎟
⎟
⎜
⎜
⎜
⎜
⎜ 1 1 ⎟ ⎟
⎟
⎟
⎟
⎜
⎜
⎜ 7 7 3 5 3⎟⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ 1 1 1 ⎟
⎟
⎜
⎜
⎜ 3 1 3 3⎟⎟
⎟
⎟
B(3) �⎜
⎜
⎜
⎜
⎜ 5 5 3 ⎟
⎟
⎟
⎟ , CR(3) � 0.076189 < 0.1.
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ 1 1 ⎟
⎟
⎜
⎜
⎜
⎜
⎜ 5 3 1 5 3⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ 5 5 ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ 1 1 1 1 ⎟
⎟
⎜
⎜
⎜
⎜
⎜ 1 1 1⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ 7 7 3 5 ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎝1 1 1 1 ⎠
3 1 1
5 5 3 3
(37)

Based on the CR values {0.093506, 0.071796, 0.076189}, By using (31), the composite PM is as follows:
three decision makers’ levels are determined as follows:
h1 � 0.28331,
h2 � 0.36898, (38)
h3 � 0.34771.

1 09769 63635 5 35831 56206 5

⎜ ⎟
⎛
⎜
⎜
⎜
⎜ 10236 1 56206 5 35831 56206 56609 ⎞⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎜
⎜
⎜
⎜
⎜ 015715 017792 1 028842 02 073253 049995 ⎟
⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
B(0) ⎜
⎜
�⎜
⎜
⎜ 02 02 34672 1 033333 3 41863 ⎟⎟
⎟
⎟
⎟ . (39)
⎜
⎜ ⎟
⎟
⎜
⎜
⎜
⎜
⎜ 027909 027909 5 3 1 35831 41863 ⎟⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎜
⎜
⎝ 017792 017792 13651 033333 027909 1 20475 ⎟⎟
⎠
02 017665 2.0002 0.23887 0.23887 0.4884 1

By transforming the elements of B(0) � (b(0) ij )n×n to the

nearest neighbouring 9-point scales, the final PM is as follows:

1 1 6 5 4 6 5
⎜
⎛
⎜ ⎟
⎞
⎟
⎜
⎜
⎜
⎜
⎜ 1 1 6 5 4 6 6⎟⎟
⎟
⎟
⎟
⎜
⎜ ⎟
⎟
⎜
⎜
⎜
⎜ 1 1 1 1 1 1⎟⎟
⎟
⎟
⎜
⎜ 1 ⎟
⎟
⎜
⎜
⎜
⎜
⎜ 6 6 4 5 2 2⎟⎟
⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜ 1 1 1 ⎟
⎟
⎜
⎜
⎜
⎜
⎜ 4 1 3 4⎟⎟
⎟
⎟
⎟
A(0) �⎜
⎜
⎜ 5 5 3 ⎟
⎟
⎟ . (40)
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ 1 1 ⎟
⎟
⎟
⎜
⎜
⎜ 5 3 1 4 4⎟⎟
⎟
⎜
⎜
⎜ 4 4 ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ 1 1 1 1 ⎟
⎟
⎟
⎜ ⎟
⎜
⎜
⎜
⎜ 2 1 2⎟⎟
⎟
⎟
⎜
⎜
⎜ 6 6 3 4 ⎟
⎟
⎟
⎜
⎜
⎜ ⎟
⎟
⎟
⎝1 1 1 1 1 ⎠
2 1
5 6 4 4 2
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Mathematical Problems in Engineering 17

Table 16: Comprehensive scores of the alternatives from three MADMs.

Alternative materials
MADMs
A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11
SAW 0.532 0.491 0.480 0.493 0.680 0.874 0.617 0.680 0.078 0.042 0.049
TOPSIS 0.508 0.439 0.437 0.439 0.677 0.797 0.608 0.621 0.080 0.045 0.059
GRA 0.640 0.574 0.570 0.575 0.701 0.847 0.652 0.694 0.360 0.344 0.348

0.9

0.8

0.7
Overall performance scores

0.6

0.5

0.4

0.3

0.2

0.1

0
1 2 3
SAW TOPSIS GRA

A1 A5 A9
A2 A6 A10
A3 A7 A11
A4 A8
Figure 4: Comprehensive scores of the alternatives from three MADMs.

,e CR value of the final PM A(0) is CR(0) � 0.064623, ,e final attribute weights calculated from the final PM
and therefore the final PM satisfies the consistency. A(0) are as follows:

0.31665, 0.32205, 0.031323, 0.093804, 0.14673, 0.048864, 0.040575. (41)

With these attribute weights, we can select the best hip Table 17 shows the ranking of the comprehensive scores
joint prosthesis material using the well-known three of the alternatives from three MADMs.
MADMs such as SAW, TOPSIS, and GRA. From Table 17, we know that the final ranking of the
Table 16 and Figure 4 show the comprehensive scores of alternative materials is as follows:
the alternatives from three MADMs.

A6 > A5 > A8 > A7 > A1 > A4 > A2 > A3 > A9 > A11 > A10. (42)

,erefore, we can select A6 (Co-Cr alloys-wrought alloy) (Co-Cr alloys-cast alloy), A8 (Ti-6Al-4V), A7 (unalloyed
as the best hip joint prosthesis material, and the next are A5 titanium), and so on.
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
18 Mathematical Problems in Engineering

Table 17: Ranking of the alternatives using three MADMs.

Alternative materials
MADMs
A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11
SAW 5 7 8 6 2 1 4 3 9 11 10
TOPSIS 5 7 8 6 2 1 4 3 9 11 10
GRA 5 7 8 6 2 1 4 3 9 11 10
Final ranks 5 7 8 6 2 1 4 3 9 11 10

First …
row
… Ci Cj … C11
First
Coloumn

Facing
C1 1

… 1 Recording cell
cell (i, j)
Ci Facing

… 1

Cj 1

… 1

C11 1

Figure 5: Recording method of the pairwise comparison judgment value on the recording cell cell(i, j) [5].

Table 18: Recording value according to verbal judgment. guarantees a very good balance between improve-
ment of consistency and preservation of primary
Recording
Verbal judgment information.
value
Equal importance 1 (ii) ,e approximate polynomial for RI enables to
Weak importance 3 calculate the RI value according to the number of
Strong importance 5 attributes without the numerical table for RI.
Demonstrated importance 7 (iii) ,e proposed method for constituting the final PM
Extreme importance 9 and determining the final attribute weights enables
Intermediate values between two adjacent judgments 2, 4, 6, 8 to reflect the decision makers’ levels well, and it is
useful for group AHP.
,e limitation of this work is that we do not deal with the
4. Conclusions uncertainty of the pairwise comparison judgments. To
We proposed a new consistency improvement method of handle the fuzziness and vagueness effectively, elicit the
inconsistent PM based on CR decreasing rate, developed an decision makers’ knowledge and develop more effective
approximate polynomial for RI according to the number of decision-making model; future work needs to introduce the
attributes, and proposed a method to determine final PM and hesitant fuzzy set, complex intuitionistic fuzzy set, and so on.
final attribute weights considered decision makers’ levels. We
compared the performances of the proposed and previous Appendix
consistency improvement methods with two numerical ex-
amples, and then applied the proposed methods to determine
material attribute weights in hip joint prosthesis material A. Preliminaries
selection.
A.1. Method to determine attribute weights using AHP with
,e main conclusions are as follows:
simplest questionnaire. Let considered attributes be noted
(i) ,e CR decreasing rate-based consistency im- C1, . . ., Cj, . . ., Cn and the weights of these attributes be noted
provement method of PM improves the consistency w1 , . . . , wj , . . . , wn , where n is the number of the attributes
better with much smaller modification amounts and and wj j is the weight of j-th attribute Cj.
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Mathematical Problems in Engineering 19

,e main steps for determining the attribute weights CI

using AHP with simplest questionnaire are as follows: [5] CR � , (A.2)
RI
Step 1. Record the pairwise comparison judgment
where CI is the consistency index of A. It is calculated as
values for the importance of the attributes in the
follows:
simplest questionnaire.
In the simplest questionnaire, the attributes C1, . . ., Cj, λmax − n􏼁
. . ., Cn are listed in the same order at the first row and CI � . (A.3)
(n − 1)
first column, respectively. ,e cell with slash “\” () is
called recording cell. Figure 5 shows the recording RI is the random consistency index (RI) (Table 2). If
method of the pairwise comparison judgment value. CR < CR0, then the PM satisfies the consistency.
,e recording cell cell(i, j) corresponds to i-th attribute Otherwise, the PM should be modified. CR0 is the
Ci at the first column and j-th attribute Cj at the first consistency threshold value. Usually, CR0 is 0.1.
row (i � 1, n − 1, j � i + 1, n).
Step 5. Calculate the attribute weights w1 . . ., wj . . ., wn
Compare i-th attribute Ci at the first column and j-th by normalizing the principal eigenvector
attribute Cj at the first row facing the recording v � (v1 , . . . , vj , . . . , vn )T of the consistent PM A as
cell cell(i,j). If i-th attribute Ci at the first column is follows:
more important over j-th attribute Cj at the first row, vj
then record the judgment value a in the left area of the wj � n , j � 1, n. (A.4)
recording cell as . If j-th attribute Cj at the first row is 􏽐k�1 vk
more important over i-th attribute Ci at the first col-
umn, then record the judgment value a in the right area
of the recording cell as . If two attributes Ci and Cj have A.2. Development method of approximate polynomial by least
an equal importance, then record the number 1 in the square method. To develop approximate polynomial
arbitrary area. Table 18shows the recording value Pm(x) � a0 + a1x + . . . + am−1xm−1 + amxm with a set of data
according to the verbal judgment. (xn, yn), n � l, 2, . . ., N, least square method could be used
By the similar way, record all the recording cells in the [29, 30].
simplest questionnaire. It requires to determine the coefficients a0, a1, . . ., am−1,
Step 2. Constitute the PM A � (aij)n×n based on the am to minimize the following function:
complete questionnaire.
N
,e recording cell cell(i,j) corresponds to the element E � 􏽘 y n − Pm x n 􏼁 􏼁 .
2
(A.5)
aij of the upper triangular matrix in the PM A. If the n�1
state of cell(i,j) is , then aij is a (aij � a). If the state of
cell(i, j) is , then aij is 1/a (aij � 1/a). From the states of From zE/zai � 0; i � 0, m, we have a system of m + 1
all the recording cells, all the elements of the upper linear equations for m + 1 unknowns ai,
triangular matrix are determined in A. ,e lower tri-
N N N
angular elements are always the positive reciprocal of
⎜
⎛
⎜
⎜ N 􏽘 xn · · · 􏽘 xm n ⎞⎟
⎟
⎟ ⎜
⎛
⎜
⎜ 􏽘 yn ⎞ ⎟
⎟
⎟
the upper triangular elements (aji � 1/aij). All the ⎜
⎜ ⎟
⎟ ⎜
⎜ ⎟
⎟
⎜
⎜
⎜ n�1 n�1 ⎟
⎟
⎟ a ⎜
⎜
⎜ n�1 ⎟
⎟
⎟
principal diagonal elements are equal to 1. (aii � 1; i � ⎜
⎜
⎜ ⎟
⎟
⎟
0 ⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜ N N N ⎟
⎟
⎟⎜⎜
⎛
⎜
⎜ ⎟
⎞
⎟
⎟ ⎜
⎜
⎜ N ⎟
⎟
⎟
1, n) ⎜
⎜
⎜ x x2n m+1 ⎟
· · · 􏽘 xn ⎟ ⎟ ⎜
⎜ a ⎟
⎟
⎟ ⎜
⎜
⎜ x y ⎟
⎟
⎟
⎜
⎜ 􏽘 n 􏽘 ⎟
⎟ ⎜
⎜
1 ⎟
⎟ � ⎜
⎜ 􏽘 n n ⎟
⎟ .
⎜
⎜
⎜ ⎟
⎟ ⎜
⎜ ⎟
⎟
⎟ ⎜
⎜
⎜ ⎟
⎟
⎟
Step 3. Determine the principal eigenvector ⎜
⎜
n�1 n�1 n�1 ⎟
⎟
⎟ ⎜
⎜
⎝ · · · ⎟
⎠ ⎜
⎜ n�1 ⎟
⎟
⎜
⎜ ··· ··· ··· ··· ⎟ ⎟ ⎜
⎜ ··· ⎟ ⎟
v � (v1 , . . . , vj , . . . , vn )T of A � (aij)n×n using the ei- ⎜
⎜
⎜ ⎟
⎟
⎟
⎜
⎜
⎜
⎜
⎟
⎟
⎟
⎟
⎜
⎜
⎜ N N N ⎟
⎟
⎟ am ⎜
⎜ N ⎟
⎟
genvector method. ⎜
⎝ 2m ⎟
⎠ ⎜
⎜
⎝ ⎟
⎟
⎠
􏽘 xm n 􏽘 xn
m+1
· · · 􏽘 xn m
􏽘 xn yn
,e principal eigenvector v � (v1 , . . . , vj , . . . , vn )T can n�1 n�1 n�1 n�1
be determined from the following equation:
(A.6)
Av � λmax v, (A.1) ,is system will have a unique solution a0, a1, . . ., am−1,
am.
For evaluating the performance of the approximate poly-
where λmax is the maximum eigenvalue of A and v is the
nomial, the mean absolute error (MAE), mean relative error
corresponding eigenvector.
(MRE), and mean squared error (MSE) are commonly used.
Step 4. Test the consistency of A using the consistency ,e MAE, MRE, and MSE of the approximate poly-
ratio (CR): nomial are as follows:
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
20 Mathematical Problems in Engineering

p
MAEm
Si � 􏽘 wk zik ; i � 1, n, (A.9)
􏼌􏼌 􏼌
1 􏼌􏼌y − Pm xn 􏼁􏼌􏼌􏼌
N k�1
MREm � 􏽘 􏼌􏼌􏼌 n 􏼌􏼌 × 100%,
􏼌􏼌
N n�1􏼌 yn (A.7) where wk denotes the k-th attribute weight.
p
(wk ≥ 0, k � 1, p, 􏽐k�1 wk � 1.)
1 N 2 Step 3. Rank the alternatives in the descending order
MSEm � 􏽘 􏼂 y − Pm x n 􏼁 􏼃 .
N n�1 n based on S1, . . ., Si, . . ., Sn, and select the alternative
with the maximum value as the best one.
,e error may be reduced by increasing the polynomial
order. However, for the high-order polynomial, the curve
shows oscillatory behaviour. ,erefore, the low-order A.3.2. TOPSIS. ,e main steps of the TOPSIS method are as
polynomial with acceptable error is used commonly follows [32, 34, 35]:
Step 1. Constitute a normalized decision-matrix
A.3. Some well-known multiattribute decision-making Z � (zik)n×p from the decision-matrix X � (xik)n×p using
methods. Multiattribute decision-making (MADM) method equation (A.10).
comprehensively evaluates the comprehensive scores of the Step 2. Constitute the weighted normalized decision-
alternatives based on multiple evaluating attributes and matrix V � (vik )n×p as follows:
select the best one to have good performance from the al- vik � wk × zik , i � 1, n, k � 1, p. (A.10)
ternatives. Simple additive weighting (SAW) method,
technique for order preference by similarity to ideal solution Step 3. Determine the positive ideal solution (PIS) V+ �
(TOPSIS) method, and grey relational analysis (GRA) (v+1 , · · · , v+k , · · · , v+p ) and the negative ideal solution
method are well-known MADMs [31–34]. (NIS) V− � (v−1 , · · · , v−k , · · · , v−p ) as follows:
Let the considered alternatives be A1, A2, . . ., An (n ≥ 2)
and the evaluating attributes be x1, . . ., xk, . . ., xp, where n v+k � max vik , v−k � min vik . (A.11)
1≤i≤n 1≤i≤n
and p are, respectively, the numbers of alternatives and
attributes. ,e alternatives are evaluated on the basis of p Step 4. Calculate the distances from the alternatives to
attributes and their values constitute a decision-matrix the PIS and NIS as follows:
X � (xik)n×p, where xik is the performance value of k-th at- 􏽶������������ 􏽶�����������
􏽴 􏽴
tribute for i-th alternative (i � 1, n, k � 1, p). p p
2 2
D+i � 􏽘 v+k − vik 􏼁 , D−i � 􏽘 v−k − vik 􏼁 ; i � 1, n.
k�1 j�1
A.3.1. SAW. ,e main steps of the SAW method are as
follows [33, 34]: (A.12)

Step 1. Constitute a normalized decision-matrix Step 5. Calculate the relative closeness values of the
Z � (zik)n×p from the decision-matrix X � (xik)n×p. alternatives as follows:
,ere are some available normalization formulas such D−i
as vector normalization formula, linear sum-based Ci � ; i � 1, n. (A.13)
D+i + D−i 􏼁
normalization formula, linear ratio-based normaliza-
tion formula, and linear min-max normalization for- Step 6. Rank the alternatives in the descending order
mula [31, 34]. In this paper, the following linear min- based on C1, . . ., Ci, . . ., Cn, and select the alternative
max normalization formula is applied. with the maximum value as the best one.

⎧
⎪ xik − xk min 􏼁
⎪
⎪
⎪ ; k ∈ K+ ,
⎪
⎪ xk max − xk min 􏼁 A.3.3. GRA. ,e main steps of the GRA method are as
⎨
zik � ⎪ (A.8) follows [34]:
⎪
⎪
⎪
⎪ xk max − xik 􏼁 Step 1. Constitute a normalized decision-matrix
⎪
⎩ ; k ∈ K− ,
xk max − xk min 􏼁 Z � (zik)n × p from X � (xij)n × p using equation (A.10).
Step 2. Set {(zi1, . . ., zik,. . ., zip); i � 1, n} as n com-
where K+ and K− are the sets of the indices for the parative sequences.
benefit and cost attributes and xkmin and xkmax are the Step 3. Determine the PIS Z+ � (z+1 , · · · , z+k , · · · , z+p ) and
minimum and maximum values of k-th attribute, set it as a reference sequence, where z+k � max1≤i≤n zik .
respectively. Step 4. Calculate the maximum and minimum values of
Step 2. Calculate the following simple weighted sums of the absolute deviations between the PIS (reference
all the alternatives: sequence) Z+ � (z+1 , · · · , z+k , · · · , z+p ) and n alternatives
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Mathematical Problems in Engineering 21

(comparative sequences) {(zi1,. . ., zik,. . .,rip); i � 1, n} as [6] A. Soni, S. Chakraborty, P. Kumar Das, and A. Kumar Saha,
follows: “Materials selection of reinforced sustainable composites by
recycling waste plastics and agro-waste: an integrated multi-
Δmin � min min 􏼈Δik 􏼉, Δmax � min min 􏼈Δik 􏼉, (A.14) criteria decision making approach,” Construction and
1≤i≤n 1≤k≤p 1≤i≤n 1≤k≤p
Building Materials, vol. 348, no. 9, Article ID 128608, 2022.
[7] C. L. Zhong, M. Q. Zhang, X. Cui, and Z. Liu, “Compre-
where
􏼌􏼌 􏼌􏼌 hensive evaluation of China’s prefabricated decoration cost
Δik � 􏼌􏼌z+k − zik 􏼌􏼌; i � 1, n, k � 1, p. (A.15) based on analytic hierarchy process,” Advances in Civil En-
gineering, vol. 2020, pp. 1–10, Article ID 1583748, 2020.
Step 5. Calculate the grey relational coefficients between [8] C. Z. Radulescu, M. Radulescu, and B. Radu, “A multi-criteria
j-th attributes of i-th alternatives and PIS as follows: deision support and application to the evaluation of the fourth
wave of COVID-19 pandemic,” Entropy, vol. 24, no. 642,
Δmin + ρ · Δmax 2022.
ξ ik � , i � 1, n, k � 1, p, (A.16)
Δik + ρ · Δmax [9] D. X. Peng and H. L. Wu, “Research on the impact of the
cancellation of state subsidies on stakeholders of offshore
where ρ is the distinguishing coefficient and it is wind power,” Mathematical Problems in Engineering,
normally selected as 0.5. vol. 2022, pp. 1–13, Article ID 6804943, 2022.
[10] D. Y. Wei, C. F. Du, Y. F. Lin, B. M. Chang, and Y. Wang,
Step 6. Calculate the grey relational grades of the al- “,ermal environment assessment of deep mine based on
ternatives as follows: analytic hierarchy process and fuzzy comprehensive evalua-
p tion,” Case Studies in Cermal Engineering, vol. 19, Article ID
ci � 􏽘 wk ξ ik . (A.17) 100618, 2020.
k�1 [11] M. Mathew, R. K. Chakrabortty, and M. J. Ryan, “A novel
approach integrating AHP and TOPSIS under spherical fuzzy
Step 7. Rank the alternatives in the descending order sets foradvanced manufacturing system selection,” Engi-
based on c1, . . ., ci, . . ., cn, and select the alternative neering Applications of Artificial Intelligence, vol. 96, Article
with the maximum value as the best one. ID 103988, 2020.
[12] M. Rawa, A. Abusorrah, H. Bassi et al., “Economical-tech-
Data Availability nical-environmental operation of power networks with wind-
solar-hydropower generation using analytic hierarchy process
Data used to support the findings of this work are included and improved grey wolf algorithm,” Ain Shams Engineering
within the article. Journal, vol. 12, no. 3, pp. 2717–2734, 2021.
[13] O. Okudan and C. Budayan, “Assessment of project char-
acteristics affecting risk occurrences in construction projects
Conflicts of Interest using fuzzy AHP,” Sigma Journal of Engineering and Natural
,e authors declare that they have no conflicts of interest. Sciences, vol. 38, no. 3, pp. 1447–1462, 2020.
[14] U. L. Dano, “Flash flood impact assessment in jeddah city: an
analytic hierarchy process approach,” Hydrology, vol. 7, no. 1,
Acknowledgments pp. 10–15, 2020.
[15] Y. G. Chen, C. M. Chen, J. Ma et al., “Multi-objective opti-
,is research received no specific grant from any funding
mization strategy of multi-sources power system operation
agency in the public, commercial, or not for-profit sectors.
based on fuzzy chance constraint programming and improved
analytic hierarchy process,” Energy Reports, vol. 7, pp. 268–
References 274, 2021.
[16] D. Rani and H. Garg, “Multiple attributes group decision-
[1] T. L. Saaty, Ce Analytic Hierarchy Process, McGraw-Hill,
making based on trigonometric operators, particle swarm
New York, 1980.
[2] T. L. Saaty, Fundamentals of Decision Making and Priority optimization and complex intuitionistic fuzzy values,” Arti-
Ceory with the Analytic Hierarchy Process, RWS Publications, ficial Intelligence Review, pp. 1–45, 2022.
Pittsburgh, 1994. [17] F. Jin, H. Garg, L. Pei, J. Liu, and H. Chen, “Multiplicative
[3] A. Ishizaka and A. Labib, “Review of the main developments consistency adjustment model and data envelopment analy-
in the analytic hierarchy process,” Expert Systems with Ap- sis-driven decision-making process with probabilistic hesitant
plications, vol. 38, pp. 14336–14345, 2011. fuzzy preference relations,” International Journal of Fuzzy
[4] F. Dweiri, S. Kumar, S. A. Khan, and V. Jain, “Designing an Systems, vol. 22, no. 7, pp. 2319–2332, 2020.
integrated AHP based decision support system for supplier [18] X. Liu, Z. Wang, S. Zhang, and H. Garg, “An approach to
selection in automotive industry,” Expert Systems with Ap- probabilistic hesitant fuzzy risky multiattribute decision
plications, vol. 62, pp. 273–283, 2016. making with unknown probability information,” Interna-
[5] W. C. Yang, J. B. Ri, J. Y. Yang, and J. S. Kim, “Materials tional Journal of Intelligent Systems, vol. 36, no. 10,
selection criteria weighting method using analytic hierarchy pp. 5714–5740, 2021.
process (AHP) with simplest questionnaire and modifying [19] R. Khan, K. Ullah, D. Pamucar, and M. Bari, “Performance
method of inconsistent pairwise comparison matrix,” Pro- measure using a multi-attribute decision making approach
ceedings of the Institution of Mechanical Engineers - Part L: based on Complex T-spherical fuzzy power aggregation op-
Journal of Materials: Design and Applications, vol. 236, no. 1, erators,” Journal of Computational and Cognitive Engineering,
pp. 69–85, 2022. 2022.
 2629, 2022, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2022/1463006 by National Institutes Of Health Malaysia, Wiley Online Library on [11/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
22 Mathematical Problems in Engineering

[20] X. Liu, Z. Wang, S. Zhang, and H. Garg, “Novel correlation

coefficient between hesitant fuzzy sets with application to
medical diagnosis,” Expert Systems with Applications, vol. 183,
Article ID 115393, 2021.
[21] A. S. Girsang, C. W. Tsai, and C. S. Yang, “Ant algorithm for
modifying an inconsistent pairwise weighting matrix in an
analytic hierarchy process,” Neural Computing & Applica-
tions, vol. 26, no. 2, pp. 313–327, 2015.
[22] S. H. Wu, J. Xie, X. D. Liu, B. He, M. X. Yang, and Z. X. Li,
“Marginal optimization method to improve the inconsistent
comparison matrix in the analytic hierarchy process,” Journal
of Systems Engineering and Electronics, vol. 28, no. 6,
pp. 1141–1151, 2017.
[23] X. Zeshui and W. Cuiping, “A consistency improving method
in the analytic hierarchy process,” European Journal of Op-
erational Research, vol. 116, no. 2, pp. 443–449, 1999.
[24] D. Cao, L. C. Leung, and J. S. Law, “Modifying inconsistent
comparison matrix in analytic hierarchy process: a heuristic
approach,” Decision Support Systems, vol. 44, no. 4,
pp. 944–953, 2008.
[25] I. T. Yang, W. C. Wang, and T. I. Yang, “Automatic repair of
inconsistent pairwise weighting matrices in analytic hierarchy
process,” Automation in Construction, vol. 22, pp. 290–297,
2012.
[26] J. Benı́tez, X. Delgado-Galván, J. Izquierdo, and R. Pérez-
Garcı́a, “Achieving matrix consistency in AHP through lin-
earization,” Applied Mathematical Modelling, vol. 35, no. 9,
pp. 4449–4457, 2011.
[27] A. Jahan, F. Mustapha, M. Y. Ismail, S. M. Sapuan, and
M. Bahraminasab, “A comprehensive VIKOR method for
material selection,” Materials & Design, vol. 32, no. 3,
pp. 1215–1221, 2011.
[28] T. L. Saaty, “Decision making-,e analytic hierarchy and
network processes (AHP/ANP),” Journal of Systems Science
and Systems Engineering, vol. 13, pp. 1–35, 2004.
[29] M. Ali and V. Zalizniak, Practical Scientific Computing,
Woodhead Publishing Limited, Sawston, Cambridge, 2011.
[30] S. Makridakis and S. C. Wheelwright, Forecasting Methods
and Applications, John Wiley & Sons, New York, 1998.
[31] W. C. Yang, S. H. Chon, C. M. Choe, and U. H. Kim,
“Materials selection method combined with different MADM
methods,” Journal of Artificial Intelligence, vol. 1, no. 2,
pp. 89–100, 2019.
[32] W. C. Yang, C. M. Choe, J. S. Kim, M. S. Om, and U. H. Kim,
“Materials selection method using improved TOPSIS without
rank reversal based on linear max-min normalization with
absolute maximum and minimum values,” Materials Research
Express, vol. 9, no. 6, Article ID 065503, 2022.
[33] A. Jahan A, K. L. Edwards, and M. Bahraminasab, Multi-
criteria Decision Analysis for Supporting the Selection of En-
gineering Materials in Product Design, Butterworth-Heine-
mann, Oxford, 2016.
[34] V. M. Athawale and S. Chakraborty, “Material selection using
multi-criteria decision-making methods: a comparative
study,” Proceedings of the Institution of Mechanical Engineers -
Part L: Journal of Materials: Design and Applications, vol. 226,
no. 4, pp. 266–285, 2012.
[35] W. C. Yang, S. H. Chon, C. M. Choe, and J. Y. Yang, “Ma-
terials selection method using TOPSIS with some popular
normalization methods,” Engineering Research Express, vol. 3,
no. 1, Article ID 015020, 2021.