INFORMATICA Informatica 1822-88440868-49520868-4952 Vilnius University INFOR570 10.15388/24-INFOR570 Research Article Prioritized Aggregation Operators of GTHFNs MADM Approach for the Evaluation of Renewable Energy Sources UluçayVakkasvakkas.ulucay@kilis.edu.tr1

V. Uluc̣ay is an associate professor at the Department of Mathematic at Kilis 7 ARALIK University, Turkey. Received a MS degree of the Gaziantep University in 2008-2010 and PhD degree of the Gaziantep University in 2013-2017. His research intersts include extensions of fuzzy sets, extensions of soft sets, extensions of neutrosophic sets and decision-making.

 DeliIrfanirfandeli@kilis.edu.tr1

İ. Deli was born in 1986 Denizli, Turkey. He received his PhD degree from Department of Mathematics, Faculty of Arts and Sciences, Gaziosmanpaṣa University, Tokat, Turkey in 2013. Currently, he is working as a professor at the University of 7 ARALIK, Kilis, Turkey. His main interest areas include soft sets, fuzzy sets, intuitionistic fuzzy sets, neutrosophic sets, game theory, decision-making, optimization, and so on.

 EdalatpanahSeyyed Ahmadsaedalatpanah@gmail.com2∗

S.A. Edalatpanah is an associate professor at the University of Ayandegan, Tonekabon, Iran. He received his PhD in applied mathematics from the University of Guilan, Rasht, Iran. Since 2021, Dr. Edalatpanah has served as a Vice President of Research and Development Department at the University of Ayandegan. He is also an academic member of Guilan University and the Islamic Azad University of Iran. Additionally, he is an Honourary Adjunct at various institutions worldwide, including Maryam Abacha American University in Nigeria (https://maaun.edu.ng/international-faculty/).

Dr. Edalatpanah’s research focuses on soft computing, operational research, uncertainty theories, fuzzy sets and their extensions, numerical computations, numerical linear algebra, and optimization. He has authored over 250 publications in respected journals and conferences. He serves as an editor and lead guest editor for numerous prominent journals, including Economics, Management Decision, Soft Computing, Buildings, Computer Modeling in Engineering and Sciences, Mathematics, Intelligent Automation and Soft Computing, PeerJ Computer Science, Neutrosophic Sets and Systems, Decision Making Applications in Management and Engineering, and Advances in Mechanical Engineering.

Dr. Edalatpanah’s contributions have gained international recognition, and he has been listed among the Top 2 percent of scientists in the world by Stanford University since 2021.

 Department of Mathematics, Kilis 7 Aralık University, Turkey Department of Applied Mathematics, Ayandegan Institute of Higher Education, Iran Corresponding author. 202411020243548598821202492024 © 2024 Vilnius University2024 Open access article under the CC BY license.

In this paper, firstly, we propose two new GTHFNs-prioritized aggregation operators called generalized trapezoidal hesitant fuzzy number prioritized weighted average operator and generalized trapezoidal hesitant fuzzy number prioritized weighted geometric operator. Secondly, we investigate the fundamental properties of the operators in detail such as idempotency, boundedness and monotonicity. Thirdly, we propose a method based on the developed GTHF-numbers prioritized aggregation operators for solving an MADM problem with GTHF-numbers. Fourthly, we give a numerical example of the developed method. Finally, a comparative analysis is given with some existing methods in solving an MADM problem with GTHF-numbers.

 hesitant fuzzy set generalized trapezoidal hesitant fuzzy numbers prioritized aggregation operator multiple attribute decision making renewable energy sources This work was not supported by any funding agency. Introduction

Nowadays, as the society continues developing, practical problems and actual scenarios of both human nature and real-world situations, as well as uncertainty, vagueness, inconsistency and imprecision, seem to be prevalent. Therefore, fuzzy set theory (Zadeh, 1965) has been utilized in various fields with imprecise information. Although the fuzzy set theory is a useful tool for modelling problems, including uncertainty information, it can be too difficult in some cases. To avoid this difficulty, recently, intuitionistic fuzzy sets (Atanassov, 1986), type-2 fuzzy sets (Mizumoto and Tanaka, 1976), type-n fuzzy sets (Rickard et al., 2008) and extensions of fuzzy sets, which express information in different ways, have been defined and researched widely. Also, as a generalization of fuzzy sets, Torra and Narukawa (2009) and Torra (2010) introduced the concept of hesitant fuzzy sets which allows the membership of an element of a set to be represented by several possible values. The relationships among hesitant fuzzy sets were also discussed. Many studies on hesitant fuzzy sets have been conducted: Xu and Xia (2011a) and Li et al. (2015) developed some distance measures for hesitant fuzzy sets under similarity measures, Xu and Xia (2011b) introduced some distance and correlation measures on hesitant fuzzy sets, including desired properties in detail, Wei (2012) developed a few prioritized aggregation operators for hesitant fuzzy sets, and then applied decision making problems in which the attributes are in a different priority level and so on. Since these studies still cannot provide all original data information for the decision making problems, Deli and Karaaslan (2021) introduced the generalized trapezoidal hesitant fuzzy numbers on R. As a consequence, a large number of studies have been conducted by the following authors: Ali et al. (2023), Atanassov (2000), Yager (2008), Xia and Xu (2011), Deli (2021, 2020), Anusha et al. (2023), Liao et al. (2014), Wei (2012).

As the classical sets, fuzzy sets and the generalization of the collected information for the values of the alternatives based on criteria of aggregation operators are useful to convert the whole data into a single value. Therefore, the aggregation operators have great importance and significance in solving MADM problems in the whole set theory. For example, Wan (2013) developed a new decision method based on power average operators of fuzzy numbers. Aydemir and Yilmaz Gündüz (2020) defined some operational laws of q-rung orthopair fuzzy sets and then proposed Dombi prioritized weighted aggregation operators. Zhao and Wei (2013) proposed the intuitionistic fuzzy Einstein hybrid averaging operator and intuitionistic fuzzy Einstein hybrid geometric operator. Verma and Sharma (2014) introduced the trapezoid fuzzy linguistic prioritized weighted average operators and developed an approach to multiple attribute group decision-making trapezoid fuzzy linguistic information. Liu et al. (2016) developed the intuitionistic trapezoidal fuzzy prioritized ordered weighted aggregation operator, then proposed the prioritized multi-criteria decision-making problems under intuitionistic trapezoidal fuzzy information. Jiang (2018) developed some models for interval intuitionistic trapezoidal fuzzy multiple attribute decision making problems in which the attributes are in different priority level and with some prioritized aggregation operators. Fahmi et al. (2019, 2021) defined some new operation laws for trapezoidal cubic hesitant fuzzy numbers and then developed some new aggregation operators. Liang et al. (2017) proposed some new prioritized aggregation operator, and then some desired properties of the new aggregation operators were studied. Liu et al. (2017) proposed two prioritized aggregation operators for hesitant intuitionistic fuzzy linguistic sets, and then, based on these aggregation operators, an approach for multi-attribute decision-making was developed under the hesitant intuitionistic fuzzy linguistic sets. Verma (2017) combined the idea of generalized mean and prioritized weighted average operators. Recently, some authors introduced several prioritized aggregation operators within different mathematical structures (Akram et al., 2020; Garg and Rani, 2023; Jana et al., 2020; Kumar and Chen, 2022; Liu and Gao, 2020; Wang et al., 2022).

Novelty

Some methods have been using generalized trapezoidal hesitant fuzzy (GTHF) numbers such as proposed by Deli and Karaaslan (2021). However, each of these methods can work well in a specific situation when the attributes have the same priority, but can also generate undesirable decision-making results when the attributes have the same priority. Therefore, inspired by the ideal of prioritized aggregation operators (Yager, 2008), we developed an approach to solve a multi-attribute decision-making (MADM) problems with GTHF-numbers in which the attributes are in a different priority level.

 Motivation and Contribution

The motivation and contributions of the study are as follows:

Prioritized operators of GTHF-numbers are introduced to propose a new method for solving MADM problems with GTHF-numbers in which the attributes are in a different priority level.

The main aim is to develop GTHF-numbers-prioritized weighted average operator and GTHF-prioritized weighted geometric operator for MADM problems with GTHF-numbers.

A method is constructed with an algorithm for MADM problems with GTHF-numbers.

An example for application is presented to demonstrate the effectiveness and advantage of the proposed method.

Paper

Structure

The remainder of the paper is organized as follows:

↓ In Section 2, we give a brief introduction on some basic definitions and propositions.

↓ In Section 3, we introduce some GTHF-numbers prioritized operators and discuss their desirable properties.

↓ In Section 4, we develop an MADM method and then initiate an example in which the attributes are in a different priority level.

↓ In Section 5, we give a comparison with some existing methods.

↪ In Section 6, we propose a conclusion.

Preliminaries

In the following, we briefly describe some basic concepts and basic operational laws related to hesitant fuzzy sets and generalized hesitant trapezoidal fuzzy numbers.

(Zadeh, <xref ref-type="bibr" rid="j_infor570_ref_040">1965</xref>)<italic>.</italic>

Let X be a universe. Then, a fuzzy set is defined as follows: A = { μ A ( x ) / x : x ∈ X } , where μ A : X → [ 0 , 1 ] such that 0 ⩽ μ A ( x ) ⩽ 1 for all x ∈ X.

 (Wang, <xref ref-type="bibr" rid="j_infor570_ref_032">2009</xref>)<italic>.</italic>

Let η a ˜ ∈ [ 0 , 1 ] and a , b , c , d ∈ R such that a ⩽ b ⩽ c ⩽ d. Then, a trapezoidal fuzzy number (TF-number) a ˜ = ⟨ ( a , b , c , d ) ; η a ˜ ⟩ is a special fuzzy set on the real number set R, whose membership function is defined as μ a ˜ ( x ) = ( x − a ) η a ˜ / ( b − a ) , a ⩽ x ⩽ b , η a ˜ , b ⩽ x ⩽ c , ( d − x ) η a ˜ / ( d − c ) , c ⩽ x ⩽ d , 0 , otherwise .

(Wang <italic>et al.</italic>, <xref ref-type="bibr" rid="j_infor570_ref_033">2006</xref>)<italic>.</italic>

Let a ˜ = ⟨ ( a , b , c , d ) ; η a ˜ ⟩ be a TF-number with it’s membership function μ a ˜ ( x ). Centroid point of a ˜, denoted by a ˜ ∗ , is computed as: a ˜ ∗ = ∫ x . μ a ˜ ( x ) d x ∫ μ a ˜ ( x ) d x = η a ˜ ( d 2 − 2 c 2 + 2 b 2 − a 2 + d c − a b ) + 3 ( c 2 − b 2 ) 3 η a ˜ ( d − c + b − a ) + 6 ( c − b ) .

(Torra, <xref ref-type="bibr" rid="j_infor570_ref_025">2010</xref>)<italic>.</italic>

Let X be a universe. Then, a hesitant fuzzy set (HFS), denoted by H, is defined as: H = { ⟨ x , ξ ( x ) ) ⟩ : x ∈ X } , where ξ ( x ) is a set of some values in [ 0 , 1 ] and ξ = ξ ( x ) is called a hesitant fuzzy element (HFE).

 (Yager, <xref ref-type="bibr" rid="j_infor570_ref_038">2008</xref>)<italic>.</italic>

Let G = { G j : j ∈ { 1 , 2 , … , n } } be a set of attributes and let a prioritization between the attribute expressed by the linear order G 1 ≻ G 2 ≻ G 3 ≻ ⋯ ≻ G n indicate that attribute G j has a higher priority than G k , if j < k. The value G j ( x ) is the performance of any alternative x under attribute G j , and satisfies G j ( x ) ∈ [ 0 , 1 ]. Then P A ( G 1 ( x ) , G 2 ( x ) , … , G n ( x ) ), called the prioritized average operator, is defined as: P A ( G 1 ( x ) , G 2 ( x ) , … , G n ( x ) ) = ∑ j = 1 n w j G j ( x ) , where w j = T j ∑ i = 1 n T i , T i = ∏ i = 1 i − 1 G i ( x ), T j = ∏ j = 1 j − 1 G j ( x ) such that T 1 = 1.

 (Wei, <xref ref-type="bibr" rid="j_infor570_ref_034">2012</xref>)<italic>.</italic>

Let ξ j ( j ∈ { 1 , 2 , … , n } ) be a collection of HFEs. Then,

The hesitant fuzzy prioritized weighted average operator of the set ξ j ( j ∈ { 1 , 2 , … , n } ) is defined as: HFPWAO ( ξ 1 , ξ 2 , … , ξ n ) = ⨁ j = 1 n ( T j ∑ i = 1 n T i ξ j ) = ⋃ h 1 1 ∈ ξ 1 , h 1 2 ∈ ξ 2 , … , h 1 n ∈ ξ n { 1 − ∏ j = 1 n ( 1 − h 1 j ) T j ∑ i = 1 n T i } , where T i = ∏ k = 1 i − 1 ∑ h 1 k ∈ ξ k h 1 k , T j = ∏ k = 1 j − 1 ∑ h 1 k ∈ ξ k h 1 k ( i , j = 2 , … , n ), T 1 = 1.

The hesitant fuzzy prioritized weighted geometric operator of the set ξ j ( j ∈ { 1 , 2 , … , n } ) is defined as: HFPWGO ( ξ 1 , ξ 2 , … , ξ n ) = ⨂ j = 1 n ( ξ j ) T j ∑ i = 1 n T i = ⋃ h 1 1 ∈ ξ 1 , h 1 2 ∈ ξ 2 , … , h 1 n ∈ ξ n { ∏ j = 1 n ( ξ n ) T j ∑ j = 1 n T j } , where T i = ∏ k = 1 i − 1 ∑ h 1 k ∈ ξ k h 1 k , T j = ∏ k = 1 j − 1 ∑ h 1 k ∈ ξ k h 1 k ( i , j = 2 , … , n ), T 1 = 1.

(Wei, <xref ref-type="bibr" rid="j_infor570_ref_034">2012</xref>).

Let ξ j and ξ ´ j ( j ∈ { 1 , 2 , … , n } ) be two sets of HFEs.

If all ξ j ( j = 1 , 2 , … , n ) are equal, i.e. ξ j = ξ for all j ∈ { 1 , 2 , … , n } then, HFPWGO ( ξ 1 , ξ 2 , … , ξ n ) = HFPWAO ( ξ 1 , ξ 2 , … , ξ n ) = ξ .

Let ξ − = min j ∈ { 1 , 2 , … , n } { ξ j }, ξ + = max j ∈ { 1 , 2 , … , n } { ξ j } then ξ − ⩽ HFPWAO ( ξ 1 , ξ 2 , … , ξ n ) , HFPWGO ( ξ 1 , ξ 2 , … , ξ n ) ⩽ ξ + .

If ξ i ⩽ ξ ´ j for all j ∈ { 1 , 2 , … , n }, then HFPWA ( ξ 1 , ξ 2 , … , ξ n ) ⩽ HFPWA ( ξ ´ 1 , ξ ´ 2 , … , ξ ´ n ) , HFPWA ( ξ 1 , ξ 2 , … , ξ n ) ⩽ HFPWA ( ξ ´ 1 , ξ ´ 2 , … , ξ ´ n ) .

(Deli and Karaaslan, <xref ref-type="bibr" rid="j_infor570_ref_010">2021</xref>)<italic>.</italic>

Let R be a set of real numbers such that a ⩽ b ⩽ c ⩽ d. Then, a generalized hesitant trapezoidal fuzzy number (GTHF-number), denoted by ℏ, is defined as: ℏ = ⟨ ( a , b , c , d ) ; ξ = { h i : h i ∈ [ 0 , 1 ] } ⟩ ( h i is set of some values in [ 0 , 1 ], i ∈ { 1 , 2 , … , n }) which is a special hesitant fuzzy set on the real number set R, whose membership functions are defined as μ i ( x ) = ( x − a ) h i / ( b − a ) , a ⩽ x < b , h i , b ⩽ x ⩽ c , ( d − x ) h i / ( d − c ) , c < x ⩽ d , 0 , otherwise .

In the paper, for purposes of focusing on GTHF- numbers, note that the set of all GTHF-numbers on R + will be denoted by Θ. (Deli and Karaaslan, <xref ref-type="bibr" rid="j_infor570_ref_010">2021</xref>)<italic>.</italic>

Let ℏ = ⟨ ( a , b , c , d ) ; ξ ⟩, ℏ 1 = ⟨ ( a 1 , b 1 , c 1 , d 1 ) ; ξ 1 ⟩, ℏ 2 = ⟨ ( a 2 , b 2 , c 2 , d 2 ) ; ξ 2 ⟩ ∈ Θ and γ ≠ 0 be any real number. Then,

ℏ 1 ⊕ ℏ 2 = ⟨ ( a 1 + a 2 , b 1 + b 2 , c 1 + c 2 , d 1 + d 2 ) ; ⋃ h 1 1 ∈ ξ 1 , h 1 2 ∈ ξ 2 { ξ 1 1 + h 1 2 − h 1 1 . h 1 2 } ⟩;

ℏ 1 ⊙ ℏ 2 = ⟨ ( a 1 a 2 , b 1 b 2 , c 1 c 2 , d 1 d 2 ) ; ⋃ h 1 1 ∈ ξ 1 , h 1 2 ∈ ξ 2 { h 1 1 . h 1 2 } ⟩;

γ ℏ = ⟨ ( γ a , γ b , γ c , γ d ) ; ⋃ h ∈ ξ { 1 − ( 1 − h ) γ } ⟩ ( γ ⩾ 0 );

( ℏ ) γ = ⟨ ( a γ , b γ , c γ , d γ ) ; ⋃ h ∈ ξ { h γ } ⟩ ( γ ⩾ 0 ).

Deli and Karaaslan (2021) Let ℏ j = ⟨ ( a j , b j , c j , d j ) ; ξ j ⟩, j ∈ I n be a set of GTHF-numbers. Then,

The hesitant fuzzy weighted geometric operator of the set ℏ j ( j ∈ { 1 , 2 , … , n } ) is defined as: H w G ( ℏ 1 , ℏ 2 , … , ℏ n ) = ⨂ j = 1 n ℏ j w j = ⟨ ( ∏ j = 1 n a j w j , ∏ j = 1 n b j w j , ∏ j = 1 n c j w j , ∏ j = 1 n d j w j ) ; ⋃ h 1 1 ∈ ξ 1 , h 1 2 ∈ ξ 2 , … , h 1 n ∈ ξ n { ∏ j = 1 n h 1 j w j } ⟩ ;

The hesitant fuzzy weighted average operator of collection ℏ j ( j ∈ { 1 , 2 , … , n } ) is defined as: H w A ( ℏ 1 , ℏ 2 , … , ℏ n ) = ⨁ j = 1 n w j . ℏ j = ⟨ ( ∑ j = 1 n w j . a j , ∑ j = 1 n w j . b j , ∑ j = 1 n w j . c j , ∑ j = 1 n w j . d j ) ; ⋃ h 1 1 ∈ ξ 1 , h 1 2 ∈ ξ 2 , … , h 1 n ∈ ξ n { 1 − ∏ j = 1 n ( 1 − h 1 j ) w j } ⟩ ,

where w = ( w 1 , w 2 , … , w n ) T is the weight vector of ℏ j , j ∈ I n such that w j ∈ [ 0 , 1 ] and ∑ j = 1 n w j = 1.

 GTHFN-Prioritized Aggregation Operators

In this section, we developed GTHFN-prioritized average operator and GTHF-prioritized geometric operator and examined some desired properties such as idempotency, boundedness and monotonicity in detail.

Let ℏ j = ⟨ ( a j , b j , c j , d j ) ; ξ j ⟩ ( j = 1 , 2 , … , n ) be a set of GTHF-numbers on [ 0 , 1 ] ⊆ R, G = { G j : j ∈ { 1 , 2 , … , n } } be a set of attributes such that there is a prioritization as in the Definition 5. Then, the GTHFN-prioritized average operator, denoted by Ϝ A ( ℏ 1 , ℏ 2 , … , ℏ n ), is defined by Ϝ A ( ℏ 1 , ℏ 2 , … , ℏ n ) = ⨁ j = 1 n ( T j ∑ i = 1 n T i ℏ j ) , where T 1 = 1 , T j = ∏ k = 1 j − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( j = 2 , … , n ) and T 1 = 1 , T i = ∏ k = 1 i − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( i = 2 , … , n ) .

Let ℏ j = ⟨ ( a j , b j , c j , d j ) ; ξ j ⟩ ( j = 1 , 2 , … , n ) be a set of GTHF-numbers on [ 0 , 1 ] ⊆ R, G = { G j : j ∈ { 1 , 2 , … , n } } be a set of attributes such that there is a prioritization as in the Definition 5. Then, their aggregated value by using the Ϝ A ( ℏ 1 , ℏ 2 , … , ℏ n ) operator is also a GTHF-number, and Ϝ A ( ℏ 1 , ℏ 2 , … , ℏ n ) = ⨁ j = 1 n ( T j ℏ j ∑ i = 1 n T i ) = ⟨ ( ∑ j = 1 n T j ∑ i = 1 n T i a j , ∑ j = 1 n T j ∑ i = 1 n T i b j , ∑ j = 1 n T j ∑ i = 1 n T i c j , ∑ j = 1 n T j ∑ i = 1 n T i d j ) ; ⋃ h 1 1 ∈ ξ 1 , h 1 2 ∈ ξ 2 , … , h 1 n ∈ ξ n { 1 − ∏ j = 1 n ( 1 − h 1 j ) T j ∑ i = 1 n T i } ⟩ , where T 1 = 1 , T j = ∏ k = 1 j − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( j = 2 , … , n ) and T 1 = 1 , T i = ∏ k = 1 i − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( i = 2 , … , n ) .

It can be easily proved that the Ϝ A operator has the following properties. (Idempotency).

Let ℏ j = ⟨ ( a j , b j , c j , d j ) ; ξ j ⟩ ( j = 1 , 2 , … , n ) be a set of GTHF-numbers on [ 0 , 1 ] ⊆ R, G = { G j : j ∈ { 1 , 2 , … , n } } be a set of attributes such that there is a prioritization as in the Definition 5. If ℏ j = ℏ ( ℏ = ⟨ ( a , b , c , d ) ; ξ j ⟩ ( j = 1 , 2 , … , n ) ) for all j ∈ { 1 , 2 , … , n }, then Ϝ A ( ℏ 1 , ℏ 2 , … , ℏ n ) = ℏ , where T 1 = 1 , T j = ∏ k = 1 j − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( j = 2 , … , n ) and T 1 = 1 , T i = ∏ k = 1 i − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( i = 2 , … , n ) .

For ℏ j = ℏ for all j ∈ { 1 , 2 , … , n }, and by definition of Ϝ A operator, we have Ϝ A ( ℏ 1 , ℏ 2 , … , ℏ n ) = ⨁ j = 1 n ( T j ℏ j ∑ i = 1 n T i ) = ⟨ ( ∑ j = 1 n T j ∑ i = 1 n T i a j , ∑ j = 1 n T j ∑ i = 1 n T i b j , ∑ j = 1 n T j ∑ i = 1 n T i c j , ∑ j = 1 n T j ∑ i = 1 n T i d j ) ; ⋃ h 1 1 ∈ ξ 1 , h 1 2 ∈ ξ 2 , … , h 1 n ∈ ξ n { 1 − ∏ j = 1 n ( 1 − h 1 j ) T j ∑ i = 1 n T i } ⟩ = ⟨ ( ∑ j = 1 n T j ∑ i = 1 n T i a , ∑ j = 1 n T j ∑ i = 1 n T i b , ∑ j = 1 n T j ∑ i = 1 n T i c , ∑ j = 1 n T j ∑ i = 1 n T i d ) ; ⋃ h 1 1 ∈ ξ 1 , h 1 2 ∈ ξ 2 , … , h 1 n ∈ ξ n { 1 − ∏ j = 1 n ( 1 − h 1 j ) T j ∑ i = 1 n T i } ⟩ , ( ∑ j = 1 n T j ∑ i = 1 n T i = 1 ) = ⟨ ( a , b , c , d ) ; ξ j ⟩ ( j = 1 , 2 , … , n ) } ⟩ = ℏ , where T 1 = 1 , T j = ∏ k = 1 j − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( j = 2 , … , n ) and T 1 = 1 , T i = ∏ k = 1 i − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( i = 2 , … , n ) . This completes the proof.  □

 (Boundedness).

Let ℏ j = ⟨ ( a j , b j , c j , d j ) ; ξ j ⟩ ( j = 1 , 2 , … , n ) be a set of GTHF-numbers on [ 0 , 1 ] ⊆ R, G = { G j : j ∈ { 1 , 2 , … , n } } be a set of attributes such that there is a prioritization as in the Definition 5. Let ℏ − = ⟨ ( min j ∈ { 1 , 2 , … , n } { a j } , min j ∈ { 1 , 2 , … , n } { b j } , min j ∈ { 1 , 2 , … , n } { c j } , min j ∈ { 1 , 2 , … , n } { d j } ) ; min j ∈ { 1 , 2 , … , n } { min h 1 j ∈ ξ j { h 1 j } } ⟩ and ℏ + = ⟨ ( max j ∈ { 1 , 2 , … , n } { a j } , max j ∈ { 1 , 2 , … , n } { b j } , max j ∈ { 1 , 2 , … , n } { c j } , max j ∈ { 1 , 2 , … , n } { d j } ) ; max j ∈ { 1 , 2 , … , n } { max h 1 j ∈ ξ j { h 1 j } } ⟩ . If “ a j ⩽ a i , b j ⩽ b i , c j ⩽ c i , d j ⩽ d i and h 1 j ⩽ h 1 i for all h 1 j ∈ ξ j , h 1 i ∈ ξ i ” ⇒ “ ℏ j = ⟨ ( a j , b j , c j , d j ) ; ξ j ⟩ ⩽ ℏ i = ⟨ ( a i , b i , c i , d i ) ; ξ i ⟩ ” then ℏ − ⩽ ℏ j ⩽ ℏ + .

We have ∑ j = 1 n T j ∑ i = 1 n T i = 1, since min j ∈ { 1 , 2 , … , n } { a j } ⩽ { a j } ⩽ max j ∈ { 1 , 2 , … , n } { a j } , min j ∈ { 1 , 2 , … , n } { b j } ⩽ { b j } ⩽ max j ∈ { 1 , 2 , … , n } { b j } , min j ∈ { 1 , 2 , … , n } { c j } ⩽ { c j } ⩽ max j ∈ { 1 , 2 , … , n } { c j } , min j ∈ { 1 , 2 , … , n } { d j } ⩽ { d j } ⩽ max j ∈ { 1 , 2 , … , n } { d j } , and min h 1 j ∈ ξ j { h 1 j } ⩽ h 1 j ⩽ max h 1 j ∈ ξ j { h 1 j } . Let Ϝ A be an operator. Then, by using the Theorem 3.2, it yields that ℏ − ⩽ ℏ j ⩽ ℏ + .  □

 (Monotonicity).

Let ℏ j = ⟨ ( a j , b j , c j , d j ) ; ξ j ⟩ ( j = 1 , 2 , … , n ) and ℏ ´ j = ⟨ ( a ´ j , b ´ j , c ´ j , d ´ j ) ; x i ´ j ⟩ ( j = 1 , 2 , … , n ) be two sets of GTHF-numbers on [ 0 , 1 ] ⊆ R, G = { G j : j ∈ { 1 , 2 , … , n } } be a set of attributes such that there is a prioritization as in the Definition 5. If ℏ j ⩽ ℏ ´ j for all j ∈ { 1 , 2 , … , n } based on equation (13), then Ϝ A ( ℏ 1 , ℏ 2 , … , ℏ n ) ⩽ Ϝ A ( ℏ ´ 1 , ℏ ´ 2 , … , ℏ ´ n ) . where T 1 = 1 , T j = ∏ k = 1 j − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( j = 2 , … , n ) , T 1 = 1 , T i = ∏ k = 1 i − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( i = 2 , … , n ) , T 1 = 1 , T j = ∏ k = 1 j − 1 ∑ h ´ 1 k ∈ x i ´ k h ´ 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h ´ 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( j = 2 , … , n ) and T 1 = 1 , T i = ∏ k = 1 i − 1 ∑ h ´ 1 k ∈ x i ´ k h ´ 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h ´ 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( i = 2 , … , n ) .

Let ℏ j = ⟨ ( a j , b j , c j , d j ) ; ξ j ⟩ ( j = 1 , 2 , … , n ) be a collection of GTHF-numbers on [ 0 , 1 ] ⊆ R, G = { G j : j ∈ { 1 , 2 , … , n } } be a set of attributes such that there is a prioritization as in the Definition 5. Then, the GTHFN-prioritized geometric operator, denoted by Ϝ G ( ℏ 1 , ℏ 2 , … , ℏ n ), is defined by Ϝ G ( ℏ 1 , ℏ 2 , … , ℏ n ) = ℏ 1 T 1 ∑ i = 1 n T i ⊗ ℏ 2 T 2 ∑ i = 1 n T i ⊗ ⋯ ⊗ ℏ 1 T 2 ∑ i = 1 n T i , where T 1 = 1 , T j = ∏ k = 1 j − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( j = 2 , … , n ) and T 1 = 1 , T i = ∏ k = 1 i − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( i = 2 , … , n ) .

Let ℏ j = ⟨ ( a j , b j , c j , d j ) ; ξ j ⟩ ( j = 1 , 2 , … , n ) be a set of GTHF-numbers on [ 0 , 1 ] ⊆ R, G = { G j : j ∈ { 1 , 2 , … , n } } be a set of attributes such that there is a prioritization as in the Definition 5. Then, their aggregated value by using the Ϝ G ( ℏ 1 , ℏ 2 , … , ℏ n ) operator is also a GTHF-number, and Ϝ G ( ℏ 1 , ℏ 2 , … , ℏ n ) = ℏ 1 T 1 ∑ j = 1 n T j ⨂ ℏ 2 T 2 ∑ j = 1 n T j ⨂ … ⨂ ℏ n T n ∑ j = 1 n T j = ⟨ ( a j T j ∑ j = 1 n T j , b j T j ∑ j = 1 n T j , c j T j ∑ j = 1 n T j , d j T j ∑ j = 1 n T j ) ; ⋃ γ 1 ∈ ℏ 1 , γ 2 ∈ ℏ 2 , … , γ n ∈ ℏ n { ∏ j = 1 n ( γ j ) T j ∑ j = 1 n T j } ⟩ , where T 1 = 1 , T j = ∏ k = 1 j − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( j = 2 , … , n ) and T 1 = 1 , T i = ∏ k = 1 i − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( i = 2 , … , n ) .

It can be easily proved that the Ϝ G operator has the following properties.

 (Idempotency).

Let ℏ j = ⟨ ( a j , b j , c j , d j ) ; ξ j ⟩ ( j = 1 , 2 , … , n ) be a set of GTHF-numbers on [ 0 , 1 ] ⊆ R, G = { G j : j ∈ { 1 , 2 , … , n } } be a set of attributes such that there is a prioritization as in the Definition 5. If ℏ j = ℏ for all j ∈ { 1 , 2 , … , n }, then Ϝ G ( ℏ 1 , ℏ 2 , … , ℏ n ) = ℏ where T 1 = 1 , T j = ∏ k = 1 j − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( j = 2 , … , n ) and T 1 = 1 , T i = ∏ k = 1 i − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( i = 2 , … , n ) .

(Boundedness).

Let ℏ j = ⟨ ( a j , b j , c j , d j ) ; ξ j ⟩ ( j = 1 , 2 , … , n ) be a set of GTHF-numbers on [ 0 , 1 ] ⊆ R, G = { G j : j ∈ { 1 , 2 , … , n } } be a set of attributes such that there is a prioritization as in the Definition 5. Let ℏ − = ⟨ ( min j ∈ { 1 , 2 , … , n } { a j } , min j ∈ { 1 , 2 , … , n } { b j } , min j ∈ { 1 , 2 , … , n } { c j } , min j ∈ { 1 , 2 , … , n } { d j } ) ; min j ∈ { 1 , 2 , … , n } { min h 1 j ∈ ξ j { h 1 j } } ⟩ and ℏ + = ⟨ ( max j ∈ { 1 , 2 , … , n } { a j } , max j ∈ { 1 , 2 , … , n } { b j } , max j ∈ { 1 , 2 , … , n } { c j } , max j ∈ { 1 , 2 , … , n } { d j } ) ; max j ∈ { 1 , 2 , … , n } { max h 1 j ∈ ξ j { h 1 j } } ⟩ . If “ a j ⩽ a i , b j ⩽ b i , c j ⩽ c i , d j ⩽ d i and h 1 j ⩽ h 1 i for all h 1 j ∈ ξ j , h 1 i ∈ ξ i ” ⇒ ” ℏ j = ⟨ ( a j , b j , c j , d j ) ; ξ j ⟩ ⩽ ℏ i = ⟨ ( a i , b i , c i , d i ) ; ξ i ⟩ ” then ℏ − ⩽ Ϝ G ( ℏ 1 , ℏ 2 , … , ℏ n ) ⩽ ℏ + , where T 1 = 1 , T j = ∏ k = 1 j − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( j = 2 , … , n ) and T 1 = 1 , T i = ∏ k = 1 i − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( i = 2 , … , n ) .

(Monotonicity).

Let ℏ j = ⟨ ( a j , b j , c j , d j ) ; ξ j ⟩ ( j = 1 , 2 , … , n ) and ℏ ´ j = ⟨ ( a ´ j , b ´ j , c ´ j , d ´ j ) ; x i ´ j ⟩ ( j = 1 , 2 , … , n ) be two sets of GTHF-numbers on [ 0 , 1 ] ⊆ R, G = { G j : j ∈ { 1 , 2 , … , n } } be a set of attributes such that there is a prioritization as in the Definition 5. If ℏ j ⩽ ℏ ´ j for all j ∈ { 1 , 2 , … , n } based on equation (20), then Ϝ G ( ℏ 1 , ℏ 2 , … , ℏ n ) ⩽ Ϝ G ( ℏ ´ 1 , ℏ ´ 2 , … , ℏ ´ n ) , where T 1 = 1 , T j = ∏ k = 1 j − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( j = 2 , … , n ) , T 1 = 1 , T i = ∏ k = 1 i − 1 ∑ h 1 k ∈ ξ k h 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( i = 2 , … , n ) , T 1 = 1 , T j = ∏ k = 1 j − 1 ∑ h ´ 1 k ∈ x i ´ k h ´ 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h ´ 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( j = 2 , … , n ) and T 1 = 1 , T i = ∏ k = 1 i − 1 ∑ h ´ 1 k ∈ x i ´ k h ´ 1 k ( d k 2 − 2 c k 2 + 2 b k 2 − a k 2 + d k c k − a k b k ) + 3 ( c k 2 − b k 2 ) 3 h ´ 1 k ( d k − c k + b k − a k ) + 6 ( c k − b k ) ( i = 2 , … , n ) .

An Approach for GTHFPA Operator to MADM Decision-Making Steps

Frame diagram for proposed work.

In this section, we shall utilize the GTHFPA operators to MADM. To do this, we develop an algorithm which is presented in Fig. 1.

Let U = { U 1 , U 2 , … , U m } be a set of alternatives, E = { E 1 = ⋃ r = 1 k 1 { e 1 r }, E 2 = ⋃ r = 1 k 2 { e 1 r } , … , E n = ⋃ r = 1 k n { e 1 r } : k 1 , k 2 , …, k n ∈ Z + } be a set of attributes. Here, there is a prioritization between the attributes expressed by the linear order E 1 ≻ E 2 ≻ ⋯ ≻ E n that indicates attribute E j has a higher priority than E k , if j < k. Suppose that GTHF sub-decision matrix for sub-attribute set E j = ⋃ r = 1 k j { e 1 r } be H j = ( x i r j ) m × k j ( i = 1 , 2 , … , m, r = 1 , 2 , … , k j ), where x i r j is in the form of GTHF-number based on the alternative U i and sub-attribute e j r .

Now, we develop an approach for MADM problems under the GTHFPA operator (or GTHFPG operator) with GTHF-numbers as the following algorithm.

Algorithm:

Step 1.

Construct the GTHF sub-decision matrix for sub-attribute set E j = ⋃ r = 1 k j { e j r } as H j = ( x i r j ) m × k j = ( ⟨ ( a i r j , b i r j , c i r j , d i r j ) ; ξ i r j ⟩ ) m × k j ( i = 1 , 2 , … , m , r = 1 , 2 , … , k j ) for decision.

Step 2.

Compute overall values matrix A i j = ⟨ ( a i j , b i j , c i j , d i j ) ; ξ i j ⟩ = GTHFPA ( x i 1 j , x i 2 j , … , x i k j j ) = T i 1 J ∑ r = 1 k j T i r J x i 1 j ⨁ T i 2 J ∑ r = 1 k j T i r J x i 2 j ⨁ ⋯ ⨁ T i k j J ∑ r = 1 k j T i r J x i k j j for i ∈ { 1 , 2 , … , m } and j ∈ { 1 , 2 , … , n } where A i j denotes evaluation of the alternative U i with respect to the attribute E j ,

where T i r j = ∏ p = 1 r − 1 ∑ h i p ∈ ξ i r j h i p ( d i p 2 − 2 c i p 2 + 2 b i p 2 − a i p 2 + d i p c i p − a i p b i p ) + 3 ( c i p 2 − b i p 2 ) 3 h i p ( d i p − c i p + b i p − a i p ) + 6 ( c i p − b i p ) ( i = 1 , 2 , … , m , r = 2 , … , k j ) and T i 1 = 1 , ( i = 1 , 2 , … , m ) .

Step 3.

Find A i ( i = 1 , 2 , … , m ) by agregating all GTHF-numbers under A i j ( j = 1 , 2 , … , n ) using the GTHFPA operator as follows: A i = GTHFPA ( A i 1 , A i 2 , … A i n ) = ⟨ ( a i , b i , c i , d i ) ; ξ i ⟩ = T i 1 ∑ j = 1 n T i j A i 1 ⨁ T i 2 ∑ j = 1 n T i j A i 2 ⨁ … ⨁ T i n ∑ j = 1 n T i j A i n , where T i j = ∏ p = 1 j − 1 ∑ h i p ∈ ξ i j h i p ( d i p 2 − 2 c i p 2 + 2 b i p 2 − a i p 2 + d i p c i p − a i p b i p ) + 3 ( c i p 2 − b i p 2 ) 3 h i p ( d i p − c i p + b i p − a i p ) + 6 ( c i p − b i p ) ( i = 1 , 2 , … , m , j = 2 , … , n ) and T i 1 = 1 , ( i = 1 , 2 , … , m ) .

Step 4.

Calculate the centroid values s ( A i ) of A i ( i = 1 , 2 , … , m ) using the Definition 3: s ( A i ) = ∑ h i ∈ ξ i h i ( d i 2 − 2 c i 2 + 2 b i 2 − a i 2 + d i c i − a i b i ) + 3 ( c i 2 − b i 2 ) 3 h i ( d i − c i + b i − a i ) + 6 ( c i − b i ) .

Step 5.

Rank all the alternatives U i ( i = 1 , 2 , , … , m ) and select the best one(s). Here if s ( A s ) > s ( A t ) ⇒ U s > U t ( s , t ∈ { 1 , 2 , … , m } ).

 An Application Case for Turkey

In this section, we present a MADM problem which is a software selection problem adapted/inspirated from Çelikbilek and Tüysüz (2016), Yuan et al. (2018). The application of the proposed approach is for evaluating renewable energy under subjective perspective with linguistic scales.

Case Study

According to the plan of renewable energy development in Turkey, the Turkish government aims to reduce the country’s dependence on imported energy. Having insufficient quantities of domestic oil and natural gas resources to support demand, the best guarantee of security of energy supply is clearly to maintain a diversity of energy sources and to focus on renewable energy sources, which the country has abundantly. There are five alternatives which are given in Table 1: Geothermal energy U 1 , Solar energy U 2 , Biomass energy U 3 and Wind energy U 4 . The most common attributes for renewable energy evaluation involve E 1 economic, E 2 technical and E 3 environmental. While many national and international companies invest in these resources for a new source of income, states encourage these investments both to produce clean energy and to utilize their own resources in energy production. Therefore, our priority is environment E 3 . Then we evaluate wind power plants as they will have a significant contribution to the country’s economy since it is a new source of income E 1 , and finally, E 2 technical effect. That is, in this case, there is a strict prioritization of parameters E 3 > E 1 > E 2 , here > indicates preference.

Renewable energy resource alternatives.

Symbol The renewable energy resource
U 1 Geothermal energy
U 2 Solar energy
U 3 Biomass energy
U 4 Wind energy

GTHF-numbers for linguistic terms.

Linguistic terms Linguistic values of GTHF-numbers Score
Absolutely Low (AL) ⟨ ( 0.1 , 0.2 , 0.3 , 0.5 ) ; { 0.1 , 0.2 , 0.4 } ⟩ 0, 0338
Very very Low (L) ⟨ ( 0.2 , 0.3 , 0.4 , 0.6 ) ; { 0.1 , 0.3 , 0.5 } ⟩ 0, 0585
Very Low (VL) ⟨ ( 0.3 , 0.4 , 0.5 , 0.6 ) ; { 0.2 , 0.4 , 0.7 } ⟩ 0, 0780
Fairly Low (FL) ⟨ ( 0.1 , 0.4 , 0.5 , 0.6 ) ; { 0.2 , 0.4 , 0.6 } ⟩ 0, 0880
Low (L) ⟨ ( 0.4 , 0.5 , 0.7 , 0.8 ) ; { 0.2 , 0.4 , 0.7 } ⟩ 0, 1560
Medium (M) ⟨ ( 0.5 , 0.7 , 0.8 , 0.9 ) ; { 0.3 , 0.5 , 0.6 } ⟩ 0, 1657
Fairly High (FH) ⟨ ( 0.3 , 0.5 , 0.6 , 0.8 ) ; { 0.2 , 0.5 , 0.9 } ⟩ 0, 1760
High (H) ⟨ ( 0.1 , 0.4 , 0.8 , 0.9 ) ; { 0.1 , 0.3 , 0.5 } ⟩ 0, 1920
Very High (VH) ⟨ ( 0.2 , 0.5 , 0.6 , 0.7 ) ; { 0.7 , 0.8 , 0.9 } ⟩ 0, 2240
Very Very High (VVH) ⟨ ( 0.6 , 0.7 , 0.8 , 0.9 ) ; { 0.6 , 0.8 , 0.9 } ⟩ 0, 2300
Absolutely High (AH) ⟨ ( 0.4 , 0.5 , 0.7 , 0.9 ) ; { 0.3 , 0.7 , 0.9 } ⟩ 0, 2818

Linguistic assessment of the renewable energy alternatives based on GTHF-number.

Primary criteria Secondary criteria Alternatives Evaluation values
E 1 -Economical criterion e 11 -Service life U 1 {(L)}
U 2 {(FL)}
U 3 {(H)}
U 4 {(AH)}
e 12 -Investment cost U 1 {(H)}
U 2 {(FH)}
U 3 {(VH)}
U 4 {(AH)}
e 13 -Operation and maintenance cost U 1 {(AH)}
U 2 {(L)}
U 3 {(VVH)}
U 4 {(FL)}
E 2 -Technical criterion e 21 -Availability U 1 {(M)}
U 2 {(VVH)}
U 3 {(H)}
U 4 {(L)}
e 22 -Capacity U 1 {(M)}
U 2 {(L)}
U 3 {(H)}
U 4 {(VH)}
e 23 -Resource density U 1 {(FH)}
U 2 {(H)}
U 3 {(VH)}
U 4 {(VVH)}
e 24 -Efficiency U 1 {(AL) }
U 2 {(VL)}
U 3 {(FH)}
U 4 {(VVH)}
E 3 -Environmental criterion e 31 -Air pollution U 1 {(AL)}
U 2 {(AH)}
U 3 {(H)}
U 4 {(VH)}
e 32 -Noise pollution U 1 {(M)}
U 2 {(VL)}
U 3 {(AL)}
U 4 {(VVH)}

Also, description of subattributes for attributes E 1 , E 2 and E 3 is presented in Table 3 as linguistic assessment of the renewable energy alternatives based on GTHF-numbers, is given as:

Economical attribute contains e 11 = ervice life, e 12 = investment cost and e 13 = operation and maintenance cost. There is a strict prioritization between parameters e 11 > e 12 > e 13 , here > indicates preferred to.

Technical attribute contains e 21 = availability, e 22 = capacity, e 23 = resource density and e 24 = efficiency. There is a strict prioritization between parameters e 21 > e 22 > e 23 > e 24 , here > indicates preference.

Environmental attribute contains e 31 = air pollution and e 32 = noise pollution. There is a strict prioritization between parameters e 31 > e 32 , here > indicates preference.

Moreover, experts are selected from a variety of departments of faculty of engineering to increase the objectivity of the results as much as possible. Experts give their evaluation information by GTHF-numbers for linguistic terms shown in Table 2.

Then, in order to select/rank a renewable energy resource, we utilize the GTHFPA operator to develop an approach to multiple-attribute decision-making problems with GTHF information, which can be expressed by using the following algorithm:

Algorithm:

Step 1.

We constructed three GTHF sub-decision matrices for sub-attribute set E j = ⋃ r = 1 k j { e j r } as H j = ( x i r j ) 4 × k j ( i = 1 , 2 , 3 , 4 ; j = 1 , 2 , 3 ; r = k 1 , k 2 , k 3 ; k 1 = 1 , 2 , 3 ; k 2 = 1 , 2 , 3 , 4 ; k 3 = 1 , 2 ) for decision in Table 4–6.

GTHF sub-decision matrices ( x i r 1 ) 5 × 3 .

Economical Service life Investment cost
u 1 ⟨ ( 0.2 , 0.3 , 0.4 , 0.6 ) ; { 0.1 , 0.3 , 0.5 } ⟩ ⟨ ( 0.1 , 0.4 , 0.8 , 0.9 ) ; { 0.1 , 0.3 , 0.5 } ⟩
u 2 ⟨ ( 0.1 , 0.4 , 0.5 , 0.6 ) ; { 0.2 , 0.4 , 0.6 } ⟩ ⟨ ( 0.3 , 0.5 , 0.6 , 0.8 ) ; { 0.2 , 0.5 , 0.9 } ⟩
u 3 ⟨ ( 0.1 , 0.4 , 0.8 , 0.9 ) ; { 0.1 , 0.3 , 0.5 } ⟩ ⟨ ( 0.2 , 0.5 , 0.6 , 0.7 ) ; { 0.7 , 0.8 , 0.9 } ⟩
u 4 ⟨ ( 0.4 , 0.5 , 0.7 , 0.9 ) ; { 0.3 , 0.7 , 0.9 } ⟩ ⟨ ( 0.4 , 0.5 , 0.7 , 0.9 ) ; { 0.3 , 0.7 , 0.9 } ⟩
Maintenance cost
u 1 ⟨ ( 0.4 , 0.5 , 0.7 , 0.9 ) ; { 0.3 , 0.7 , 0.9 } ⟩
u 2 ⟨ ( 0.4 , 0.5 , 0.7 , 0.8 ) ; { 0.2 , 0.4 , 0.7 } ⟩
u 3 ⟨ ( 0.6 , 0.7 , 0.8 , 0.9 ) ; { 0.6 , 0.8 , 0.9 } ⟩
u 4 ⟨ ( 0.1 , 0.4 , 0.5 , 0.6 ) ; { 0.2 , 0.4 , 0.6 } ⟩

GTHF sub-decision matrices ( x i r 2 ) 5 × 4 .

Technical Availability Capacity
u 1 ⟨ ( 0.5 , 0.7 , 0.8 , 0.9 ) ; { 0.3 , 0.5 , 0.6 } ⟩ ⟨ ( 0.5 , 0.7 , 0.8 , 0.9 ) ; { 0.3 , 0.5 , 0.6 } ⟩
u 2 ⟨ ( 0.6 , 0.7 , 0.8 , 0.9 ) ; { 0.6 , 0.8 , 0.9 } ⟩ ⟨ ( 0.4 , 0.5 , 0.7 , 0.8 ) ; { 0.2 , 0.4 , 0.7 } ⟩
u 3 ⟨ ( 0.1 , 0.4 , 0.8 , 0.9 ) ; { 0.1 , 0.3 , 0.5 } ⟩ ⟨ ( 0.1 , 0.4 , 0.8 , 0.9 ) ; { 0.1 , 0.3 , 0.5 } ⟩
u 4 ⟨ ( 0.4 , 0.5 , 0.7 , 0.8 ) ; { 0.2 , 0.4 , 0.7 } ⟩ ⟨ ( 0.2 , 0.5 , 0.6 , 0.7 ) ; { 0.7 , 0.8 , 0.9 } ⟩
Resource density Efficiency
u 1 ⟨ ( 0.3 , 0.5 , 0.6 , 0.8 ) ; { 0.2 , 0.5 , 0.9 } ⟩ ⟨ ( 0.1 , 0.2 , 0.3 , 0.5 ) ; { 0.1 , 0.2 , 0.4 } ⟩
u 2 ⟨ ( 0.1 , 0.4 , 0.8 , 0.9 ) ; { 0.1 , 0.3 , 0.5 } ⟩ ⟨ ( 0.3 , 0.4 , 0.5 , 0.6 ) ; { 0.2 , 0.4 , 0.7 } ⟩
u 3 ⟨ ( 0.2 , 0.5 , 0.6 , 0.7 ) ; { 0.7 , 0.8 , 0.9 } ⟩ ⟨ ( 0.3 , 0.5 , 0.6 , 0.8 ) ; { 0.2 , 0.5 , 0.9 } ⟩
u 4 ⟨ ( 0.6 , 0.7 , 0.8 , 0.9 ) ; { 0.6 , 0.8 , 0.9 } ⟩ ⟨ ( 0.6 , 0.7 , 0.8 , 0.9 ) ; { 0.6 , 0.8 , 0.9 } ⟩

GTHF sub-decision matrices ( x i r 3 ) 5 × 2 .

Environmental Air pollution Noise pollution
u 1 ⟨ ( 0.1 , 0.2 , 0.3 , 0.5 ) ; { 0.1 , 0.2 , 0.4 } ⟩ ⟨ ( 0.5 , 0.7 , 0.8 , 0.9 ) ; { 0.3 , 0.5 , 0.6 } ⟩
u 2 ⟨ ( 0.4 , 0.5 , 0.7 , 0.9 ) ; { 0.3 , 0.7 , 0.9 } ⟩ ⟨ ( 0.3 , 0.4 , 0.5 , 0.6 ) ; { 0.2 , 0.4 , 0.7 } ⟩
u 3 ⟨ ( 0.1 , 0.4 , 0.8 , 0.9 ) ; { 0.1 , 0.3 , 0.5 } ⟩ ⟨ ( 0.1 , 0.2 , 0.3 , 0.5 ) ; { 0.1 , 0.2 , 0.4 } ⟩
u 4 ⟨ ( 0.2 , 0.5 , 0.6 , 0.7 ) ; { 0.7 , 0.8 , 0.9 } ⟩ ⟨ ( 0.6 , 0.7 , 0.8 , 0.9 ) ; { 0.6 , 0.8 , 0.9 } ⟩

The decision makers’ evaluation of the alternatives with respect to criteria.

Economical Technical
u 1 ⟨ ( 0.191 , 0.321 , 0.469 , 0.654 ) ; { 0.107 , 0.318 , 0.524 } ⟩ ⟨ ( 0.459 , 0.662 , 0.770 , 0.906 ) ; { 0.1 , 0.3 , 0.5 } ⟩
u 2 ⟨ ( 0.159 , 0.426 , 0.533 , 0.651 ) ; { 0.2 , 0.420 , 0.697 } ⟩ ⟨ ( 0.497 , 0.627 , 0.801 , 0.910 ) ; { 0.2 , 0.5 , 0.9 } ⟩
u 3 ⟨ ( 0.152 , 0.437 , 0.769 , 0.868 ) ; { 0.286 , 0.475 , 0.655 } ⟩ ⟨ ( 0.112 , 0.414 , 0.801 , 0.907 ) ; { 0.7 , 0.8 , 0.9 } ⟩
u 4 ⟨ ( 0.363 , 0.488 , 0.675 , 0.863 ) ; { 0.288 , 0.673 , 0.881 } ⟩ ⟨ ( 0.409 , 0.552 , 0.720 , 0.824 ) ; { 0.3 , 0.7 , 0.9 } ⟩
Environmental
u 1 ⟨ ( 0.141 , 0.251 , 0.351 , 0.541 ) ; { 0.123 , 0.238 , 0.9 } ⟩
u 2 ⟨ ( 0.369 , 0.469 , 0.637 , 0.806 ) ; { 0.270 , 0.627 , 0.859 } ⟩
u 3 ⟨ ( 0.1 , 0.366 , 0.715 , 0.832 ) ; { 0.1 , 0.284 , 0.484 } ⟩
u 4 ⟨ ( 0.324 , 0.562 , 0.662 , 0.762 ) ; { 0.672 , 0.8 , 0.9 } ⟩

Step 2.

We computed overall values matrix based on Eqs. (23)–(24) in Table 7.

Step 3.

We found A i ( i = 1 , 2 , … , m ) by aggregated all GTHF-numbers under A i j ( j = 1 , 2 , … , n ) by using Eqs. (26)–(28) in Table 8.

A i ( i = 1 , 2 , 3 , 4 ) by aggregated all GTHF-numbers.

A 1 ⟨ ( 0.2271 , 0.3665 , 0.5053 , 0.6835 ) ; { 0.1468 , 0.3409 , 0.5417 } ⟩
A 2 ⟨ ( 0.2470 , 0.4698 , 0.5966 , 0.7183 ) ; { 0.2610 , 0.5001 , 0.7479 } ⟩
A 3 ⟨ ( 0.1398 , 0.4282 , 0.7730 , 0.8751 ) ; { 0.2450 , 0.4363 , 0.6229 } ⟩
A 4 ⟨ ( 0.3703 , 0.5135 , 0.6855 , 0.8405 ) ; { 0.3812 , 0.6727 , 0.8656 } ⟩

Step 4.

We calculated the centroid values s ( A i ) of A i by using Eq. (29) in Table 9.

Centroid values s ( A i ) of A i .

s ( A 1 ) = 1.3245
s ( A 2 ) = 1.5383
s ( A 3 ) = 1.7179
s ( A 4 ) = 1.8062

Step 5.

All the alternatives U i ( i = 1 , 2 , 3 , 4 ) are ranked and the best one(s) is selected. Here, if s ( A 4 ) > s ( A 3 ) > s ( A 2 ) > s ( A 1 ) ⇒ U 4 > U 3 > U 2 > U 1 . Thus, the most environmentally friendly renewable energy source is U 4 .

 Comparison and Analysis Discussion

When a decision analyst collects the data or information using GTHF-numbers, no prioritization operator can handle this information or data. The above defined GTHFN-prioritized aggregation operators are the only tool to solve this kind of information and help the decision analyst to make a decision. To demonstrate the effectiveness of the proposed method, a comparison of the decision-making results of the MADM methods based on prioritized aggregation operators of existing presented by Wei (2012), Wan et al. (2015) and Liang et al. (2017) is carried out, as shown in Table 10.

The results from the different operators.

Methods Ranking of alternatives The optimal alternative
The method in Wei (2012) U 4 > U 2 > U 3 > U 1 , U 4
The method in Wan et al. (2015) U 4 > U 3 > U 1 > U 2 , U 4
The method in Liang et al. (2017) U 4 > U 3 > U 2 > U 1 , U 4
Proposed (GTHFNPA)Method U 4 > U 3 > U 2 > U 1 , U 4
Proposed (GTHFNPG) Method U 4 > U 3 > U 2 > U 1 , U 4
Conclusion

GTHF-numbers are very useful for expressing ill-known quantities. Since aggregation operators play a vital role in decision-making, this paper investigates the prioritized MADM problems in which the attribute values are in the form of GTHF-numbers. Firstly, we introduced two aggregation techniques called generalized trapezoidal hesitant fuzzy prioritized weighted average operator and generalized trapezoidal hesitant fuzzy prioritized weighted geometric operator for aggregating the generalized trapezoidal hesitant fuzzy information. Next, we discussed some basic properties of the developed operators, namely idempotency, boundary and monotonicity. In addition, two approaches for multiple-attribute decision-making under the generalized trapezoidal hesitant fuzzy environments are developed. Finally, a practical study has been conducted to demonstrate the proposed MADM method in more practicality and effectiveness, since it considers prioritization relationships among attributes. Meanwhile, the prioritized operators of GTHF-numbers provide a new tool of information fusion for solving decision problems under hesitant environments. Further, the extensions of hesitant prioritized aggregation operators model to the MADM problems under other fuzzy environments would also be studied in the near future. In future, we plan to extend our research work to TOPSIS, ARAS, ELECTRE, WASPAS, MABAC, EDAS, QUALIFLEX, and so on. Obtaining the data and writing in this study as GTHF-numbers is the most common problem. Therefore, we will try to solve this problem in future.

 References Akram, M., Alsulami, S., Khan, A., Karaaslan, F. (2020). Multi-criteria group decision-making using spherical fuzzy prioritized weighted aggregation operators. International Journal of Computational Intelligence Systems, 13(1), 1429–1446. Ali, W., Shaheen, T., Haq, I.U., Toor, H., Akram, F., Garg, H., Uddin, M.Z., Hassan, M.M. (2023). Aczel-Alsina-based aggregation operators for intuitionistic hesitant fuzzy set environment and their application to multiple attribute decision-making process. AIMS Mathematics, 8(8), 18021–18039. Anusha, G., Ramana, P.V., Sarkar, R. (2023). Hybridizations of Archimedean copula and generalized MSM operators and their applications in interactive decision-making with q-rung probabilistic dual hesitant fuzzy environment. Decision Making: Applications in Management and Engineering, 6(1), 646–678. Atanassov, K. (1986). Intuitionistic Fuzzy Sets, Vol. 20. Elsevier, pp. 87–96. Atanassov, K. (2000). Two theorems for intuitionistic fuzzy sets. Fuzzy Sets and Systems, 110, 267–269. Aydemir, S.B., Yilmaz Gündüz, S. (2020). Extension of multi-Moora method with some q-rung orthopair fuzzy Dombi prioritized weighted aggregation operators for multi-attribute decision making. Soft Computing, 24(24), 18545–18563. Çelikbilek, Y., Tüysüz, F. (2016). An integrated grey based multi-criteria decision making approach for the evaluation of renewable energy sources. Energy, 115, 1246–1258. Deli, I. (2020). A TOPSIS method by using generalized trapezoidal hesitant fuzzy numbers and application to a robot selection problem. Journal of Intelligent & Fuzzy Systems, 38(1), 779–793. Deli, I. (2021). Bonferroni mean operators of generalized trapezoidal hesitant fuzzy numbers and their application to decision-making problems. Soft Computing, 25(6), 4925–4949. Deli, I., Karaaslan, F. (2021). Generalized trapezoidal hesitant fuzzy numbers and their applications to multi criteria decision-making problems. Soft Computing, 25(2), 1017–1032. Fahmi, A., Abdullah, S., Amin, F. (2021). Aggregation operators on cubic linguistic hesitant fuzzy numbers and their application in group decision-making. Granular Computing, 6, 303–320. Fahmi, A., Abdullah, S., Amin, F., Ali, A., Ahmed, R., Shakeel, M. (2019). Trapezoidal cubic hesitant fuzzy aggregation operators and their application in group decision-making. Journal of Intelligent & Fuzzy Systems, 36(4), 3619–3635. Garg, H., Rani, D. (2023). New prioritized aggregation operators with priority degrees among priority orders for complex intuitionistic fuzzy information. Journal of Ambient Intelligence and Humanized Computing, 14(3), 1373–1399. Jana, C., Pal, M., Wang, J.-q. (2020). Bipolar fuzzy Dombi prioritized aggregation operators in multiple attribute decision making. Soft Computing, 24, 3631–3646. Jiang, X.-P. (2018). Interval intuitionistic trapezoidal fuzzy prioritized aggregating operators and their application to multiple attribute decision making. Journal of Intelligent Computing Volume, 9(3), 102. Kumar, K., Chen, S.-M. (2022). Multiple attribute group decision making based on advanced linguistic intuitionistic fuzzy weighted averaging aggregation operator of linguistic intuitionistic fuzzy numbers. Information Sciences, 587, 813–824. Li, D., Zeng, W., Zhao, Y. (2015). Note on distance measure of hesitant fuzzy sets. Information Sciences, 321, 103–115. Liang, C., Zhao, S., Zhang, J. (2017). Multi-criteria group decision making method based on generalized intuitionistic trapezoidal fuzzy prioritized aggregation operators. International Journal of Machine Learning and Cybernetics, 8, 597–610. Liao, H., Xu, Z., Zeng, X.-J. (2014). Distance and similarity measures for hesitant fuzzy linguistic term sets and their application in multi-criteria decision making. Information Sciences, 271, 125–142. Liu, P., Gao, H. (2020). A novel green supplier selection method based on the interval type-2 fuzzy prioritized choquet bonferroni means. IEEE/CAA Journal of Automatica Sinica, 8(9), 1549–1566. Liu, P., Li, Y., Antuchevičienė, J. (2016). Multi-criteria decision-making method based on intuitionistic trapezoidal fuzzy prioritised OWA operator. Technological and Economic Development of Economy, 22(3), 453–469. Liu, P., Mahmood, T., Khan, Q. (2017). Multi-attribute decision-making based on prioritized aggregation operator under hesitant intuitionistic fuzzy linguistic environment. Symmetry, 9(11), 270. Mizumoto, M., Tanaka, K. (1976). Some properties of fuzzy sets of type 2. Information and Control, 31(4), 312–340. Rickard, J.T., Aisbett, J., Gibbon, G. (2008). Fuzzy subsethood for fuzzy sets of type-2 and generalized type-n. IEEE Transactions on Fuzzy Systems, 17(1), 50–60. Torra, V. (2010). Hesitant fuzzy sets. International Journal of Intelligent Systems, 25(6), 529–539. Torra, V., Narukawa, Y. (2009). On hesitant fuzzy sets and decision. In: 2009 IEEE International Conference on Fuzzy Systems, pp. 1378–1382. Verma, R. (2017). Multiple attribute group decision making based on generalized trapezoid fuzzy linguistic prioritized weighted average operator. International Journal of Machine Learning and Cybernetics, 8(6), 1993–2007. Verma, R., Sharma, B.D. (2014). Trapezoid fuzzy linguistic prioritized weighted average operators and their application to multiple attribute group decision-making. Journal of Uncertainty Analysis and Applications, 2, 1–19. Wan, S., Dong, J., Yang, D. (2015). Trapezoidal intuitionistic fuzzy prioritized aggregation operators and application to multi-attribute decision making. Iranian Journal of Fuzzy Systems, 12(4), 1–32. Wan, S.-P. (2013). Power average operators of trapezoidal intuitionistic fuzzy numbers and application to multi-attribute group decision making. Applied Mathematical Modelling, 37(6), 4112–4126. Wang, W., Jiang, W., Han, X., Liu, S. (2022). An extended gained and lost dominance score method based risk prioritization for Fine-Kinney model with interval type-2 fuzzy information. Human and Ecological Risk Assessment: An International Journal, 28(1), 154–183. Wang, Y.-M. (2009). Centroid defuzzification and the maximizing set and minimizing set ranking based on alpha level sets. Computers & Industrial Engineering, 57(1), 228–236. Wang, Y.-M., Yang, J.-B., Xu, D.-L., Chin, K.-S. (2006). On the centroids of fuzzy numbers. Fuzzy Sets and Systems, 157(7), 919–926. Wei, G. (2012). Hesitant fuzzy prioritized operators and their application to multiple attribute decision making. Knowledge-Based Systems, 31, 176–182. Xia, M., Xu, Z. (2011). Hesitant fuzzy information aggregation in decision making. International Journal of Approximate Reasoning, 52(3), 395–407. Xu, Z., Xia, M. (2011a). Distance and similarity measures for hesitant fuzzy sets. Information Sciences, 181(11), 2128–2138. Xu, Z., Xia, M. (2011b). On distance and correlation measures of hesitant fuzzy information. International Journal of Intelligent Systems, 26(5), 410–425. Yager, R.R. (2008). Prioritized aggregation operators. International Journal of Approximate Reasoning, 48(1), 263–274. Yuan, J., Li, C., Li, W., Liu, D., Li, X. (2018). Linguistic hesitant fuzzy multi-criterion decision-making for renewable energy: a case study in Jilin. Journal of Cleaner Production, 172, 3201–3214. Zadeh, L.A. (1965). Fuzzy sets. Information and Control, 8(3), 338–353. Zhao, X., Wei, G. (2013). Some intuitionistic fuzzy Einstein hybrid aggregation operators and their application to multiple attribute decision making. Knowledge-Based Systems, 37, 472–479.