Abstract
In this paper, on the basis of analyzing some existing limitations in the operational laws defined for triangular intuitionistic fuzzy numbers (TIFNs), we first proposed some improved operational laws for TIFNs. Then, based on new operational laws, we developed some aggregation operators for TIFNs, such as triangular intuitionistic fuzzy-weighted averaging operator, triangular intuitionistic fuzzy geometric operator, triangular intuitionistic fuzzy-ordered-weighted averaging operator, triangular intuitionistic fuzzy-ordered-weighted geometric operator, triangular intuitionistic fuzzy hybrid averaging operator, and triangular intuitionistic fuzzy hybrid geometric operators, and discussed some desirable properties of these operators. Furthermore, based on these aggregation operators, we developed a multi-criteria decision-making (MCDM) method in which the criteria values were represented by TIFNs. Finally, a numerical example was used to show the practicality and effectiveness of the proposed MCDM method by comparing the proposed method with the existing methods.
Similar content being viewed by others
Explore related subjects
Discover the latest articles, news and stories from top researchers in related subjects.Avoid common mistakes on your manuscript.
1 Introduction
With the arising complexities of decision-making environment, the decision makers felt difficulties to get a decision within appropriate time using imprecise, vague, and uncertain information (Liu and Li 2017a; Liu and Shi 2017; Liu and Tang 2016; Liu and Wang 2017; Liu et al. 2016; Chen and Hong 2014a, b; Qin 2017; Pedrycz and Chen 2015; Chen et al. 2014). Intuitionistic fuzzy sets (IFSs) proposed by Atanassov (1986) are one of the most admissible theories to handle the impreciseness, vagueness, and uncertainties (Garg 2016; Jiang et al. 2017; Xu and Yager 2006; Liu 2017; Liu and Chen 2017; Liu and Li 2017b; Liu et al. 2017; Xu 2007a, b; Xu and Gou 2017; Chen et al. 2016a, b, c; Chen and Chang 2015, 2016; Chen and Randyanto 2013). However, in some circumstances, it is difficult to give the preference of decision makers by a crisp number and it is suitable to manifest the preference of decision makers by TIFNs (Liu and Yuan 2007). The basic feature of TIFNs is that the values of its truth-membership and falsity-membership functions are triangular fuzzy numbers rather than exact numbers. Now, some research achievements for TIFNs have been done. Liang et al. (2014) introduced the TIFWG operator, TIFOWG operator, and the TIFHG operator for TIFNs and applied them to multiple-attribute group decision making (MAGDM) with TIFNs. Wang (2008a, b) introduced the fuzzy number intuitionistic fuzzy set (FNIFS) and proposed some aggregation operators for FNIFSs, such as FNIFOWA operator, FNIFHA operator, FNIFWG operator, FINIFOWG operator, and FNIFHG operator then applied them to MCDM problems with FNIFNs. Recently, Zhou and Chang (2014) proposed some Hamacher aggregation operators for FNIFNs and applied them to MCDM problems with FNIFNs. Wei et al. (2010) proposed some induced ordered weighted geometric (I-TIFOWG) operators for TIFNs to deal with MCDM problems. More and more MCDM methods (Qiu 2011; Wang 2012; Zhao et al. 2015) have been applied under triangular intuitionistic fuzzy (TIF) environment. However, these methods can only deal with decision-making problems with TIFNs and cannot handle trapezoidal intuitionistic fuzzy information. To solve these problems, Ye (2014) proposed the concepts of trapezoidal intuitionistic fuzzy sets (TrIFSs), trapezoidal intuitionistic fuzzy numbers (TrIFNs), and defined some operational laws, score and accuracy functions. Furthermore, they defined some prioritized aggregation operators for TrIFNs, such as TRIFPWA operators and prioritized weighted geometric (TrIFPWG) operator for TrIFNs, and developed an MCDM method based on these prioritized aggregation operators Liu and Su (2010). The proposed trapezoidal fuzzy linguistic sets and defined some aggregation operators applied them to MCDM problems. Moreover, Liu and Qin (2017) proposed Maclaurin symmetric mean (MSM) operators of linguistic intuitionistic fuzzy numbers, and Şahin and Liu (2017) proposed possibility induced aggregation operator for simplified neutrosophic sets and applied them MCDM. Ye (2015) introduced the concept of trapezoidal neutrosophic set and defined some aggregation operators and applied them to MCDM. Ye (2016) proposed projection and bidirectional projection measures for single-valued neutrosophic sets, and based on these, measures presented MCDM method to solve mechanical design scheme problem.
Practically, all the above studies are reasonable under the conditions that truth-membership or falsity-membership degrees are not equal to zero. However, suppose that there are two TIFNs \(\tilde{a}_{1}\) and \(\tilde{a}_{2}\), such that \(T\left( {\tilde{a}_{1} } \right) = 0\) and \(T\left( {\tilde{a}_{2} } \right) \ne 0\), or \(F\left( {\tilde{a}_{1} } \right) = 0\) and \(F\left( {\tilde{a}_{2} } \right) \ne 0\), then based on the aggregation operators defined for TIFNs or for FNIFNs, the overall aggregated value of either truth-membership or falsity-membership values is zero. In other words, there is no effect of other degrees on the overall aggregated values either truth membership or falsity membership. Moreover, it has been pointed out that the overall truth-membership degree (or falsity-membership degree) is independent of their relative falsity-membership degree (or truth-membership degree) in the aggregation process. Therefore, the aggregated results are unacceptable. Hence, there is a need to modify these existing operations by a proper considering the correlation between truth-membership degree and falsity-membership degree.
Thus, the objective of this article is to propose some modified operations for TIFNs. Then, based on these operational laws, we propose some aggregation operators, such as triangular intuitionistic fuzzy-weighted averaging (TIFWA) operator, triangular intuitionistic fuzzy-weighted geometric (TIFWG) operator, triangular intuitionistic fuzzy-ordered-weighted averaging (TIFOWA) operator, triangular intuitionistic fuzzy-ordered-weighted geometric (TIFOWG) operator, triangular intuitionistic fuzzy hybrid averaging (TIFHA) operator, and triangular intuitionistic fuzzy hybrid geometric (TIFHG) operator to overcome the limitations of the existing operators. Then, we apply them to MCDM problems under TIF information. Finally, a comparison has been made with the existing MCDM methods.
The rest of the article is arranged as follows. In Sect. 2, some basic definitions of TIFNs are given. In Sect. 3, we define some improved operational laws for TIFNs. In Sect. 4, based on these operations, some weighted averaging aggregation and weighted geometric operators are developed. In Sect. 5, some ordered weighted averaging aggregation, ordered weighted geometric operators, and hybrid aggregation operators for TIFNs are proposed. In Sect. 6, we develop an MADM method to handle MADM problem under TIF environment. In Sect. 7, a numerical example adapted from Herrera et al. (2000) is used to show the practicality and effectiveness of the proposed method, and comparison and discussion are done. At the end, the concluding remarks and future work are given.
2 Preliminaries
In this section, some basic definitions and operational laws of TIFSs are briefly reviewed.
2.1 Triangular intuitionistic fuzzy set and their operations
Definition 1 (Liu and Yuan 2007)
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{U}\) be a non-empty universe of discourse set. Then, a triangular intuitionistic set \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{M}\) in \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{U}\) is defined and mathematically represented as follows:
where \(T_{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{M} }} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{u} )\) and \(F_{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{M} }} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{u} )\) are two triangular fuzzy numbers \(T_{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{M} }} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{u} ) = \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{t} _{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{M} }}^{1} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{u} ),\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{t} _{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{M} }}^{2} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{u} )\,,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{t} _{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{M} }}^{3} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{u} )} \right)\) and \(F_{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{M} }} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{u} ) = \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{f} _{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{M} }}^{1} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{u} ),\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{f} _{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{M} }}^{2} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{u} )\,,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{f} _{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{M} }}^{3} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{u} )} \right)\) satisfying the condition \(0 \le \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{t} _{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{M} }}^{3} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{u} ) + \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{f} _{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{M} }}^{3} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{u} ) \le 1\). For simplicity, let \(T_{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{M} }} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{u} ) = \left\langle {p,\,q,\,r} \right\rangle\) and \(F_{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{M} }} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{u} ) = \left\langle {t,\,u,\,v} \right\rangle\), then the TIFN is denoted by \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} = \left\langle {\left( {p,\,q,\,r} \right),\,\left( {t,\,u,\,v} \right)} \right\rangle .\)
Definition 2 (Liu and Yuan 2007)
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} = \langle {( {p,\,q,\,r} ),\,( {t,\,u,\,v} )} \rangle ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} = \langle {( {p_{1} ,\,q_{1} ,\,r_{1} } ),\,( {t_{1} ,\,u_{1} ,\,v_{1} } )} \rangle\) and \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} = \langle {( {p_{2} ,\,q_{2} ,\,r_{2} } ),\,( {t_{2} ,\,u_{2} ,\,v_{2} } )} \rangle\) be any three TIFNs and \(\chi \ge 0\). Then, some operational laws for TIFNs are defined as follows:
The above-defined operational laws for TIFNs have some limitations which can be discussed in an example given below.
Example 1
Three groups of professors \(P_{j} \;(j = 1,\,2,\,3)\) want to select a student for research project among three students \(a_{i} \;(i = 1,\,2,\,3)\), and the selection and rejection rating of the three students from the three groups of professors is given by TIFN \(L_{ij} (i,\,j = 1,\,2,\,3).\) Suppose \(L_{11} = \langle {( {0.1,\,0.2,\,0.3} ),\,( {0.3,\,0.4,\,0.5} )} \rangle ,\) \(\,L_{12} = \langle {( {0.3,\,0.4,\,0.5} ),\,( {0.1,\,0.1,\,0.2} )} \rangle\), and \(L_{13} = \langle {( {0.4,\,0.5,\,0.6} ),\,( {0.0,\,0.0,\,0.0} )} \rangle\), and the corresponding importance degree of the groups of three professors is \(\omega = ( {0.3,\,0.2,\,0.4} )^{T} ;\) then, by the operations defined in Definition 2, we get the overall triangular intuitionistic fuzzy information \(L_{1} = \mathop \oplus \nolimits_{i = 1}^{3} \omega_{i} L_{1i} =\) \(\left\langle {\left( {0.2178,\,0.3116,\,0.4072} \right),\,\left( {0.0,\,0.0,\,0.0,\,0.0} \right)} \right\rangle .\) That is to say that \(F_{{L_{1j} }} (j = 1,\,2)\) have no effects on the overall result. This is an undesirable property. Obviously, these operations can handle a situation in which membership and non-membership functions are not equal to zero. In other words, if either of the membership or no-membership degree equals to zero, respectively, then there is no effect of the other membership or no-membership degrees on the overall aggregated result in the aggregation process.
Definition 3 (Wang 2012)
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} = \left\langle {\left( {p,\,q,\,r} \right),\,\left( {t,\,u,\,v} \right)} \right\rangle\) be a TIFN. Then, the score and accuracy functions defined by Wang (2012) are as follows:
Definition 4
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} = \left\langle {\left( {p_{1} ,\,q_{1} ,\,r_{1} } \right),\,\left( {t_{1} ,\,u_{1} ,\,v_{1} } \right)} \right\rangle\) and \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} = \left\langle {\left( {p_{2} ,\,q_{2} ,\,r_{2} } \right),\,\left( {t_{2} ,\,u_{2} ,\,v_{2} } \right)} \right\rangle\) be any two TIFNs. Then, the comparison rules for comparing TIFNs are defined as follows:
-
1.
If \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{E} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ) > \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{E} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} ),\) then \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} > \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} .\)
-
2.
If \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{E} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ) = \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{E} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} )\) and \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{H} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ) > \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{H} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} ),\) then \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} > \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} .\)
-
3.
If \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{E} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ) = \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{E} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} )\) and \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{H} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ) = \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{H} (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} ),\) then \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} = \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} .\)
3 Improved operational laws for TIFNs
In this section, we propose some improved operational laws for TIFNs.
Definition 5
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} = \left\langle {\left( {p,\,q,\,r} \right),\,\left( {t,\,u,\,v} \right)} \right\rangle ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} = \left\langle {\left( {p_{1} ,\,q_{1} ,\,r_{1} } \right),\,\left( {t_{1} ,\,u_{1} ,\,v_{1} } \right)} \right\rangle\) and \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} = \left\langle {\left( {p_{2} ,\,q_{2} ,\,r_{2} } \right),\,\left( {t_{2} ,\,u_{2} ,\,v_{2} } \right)} \right\rangle\) be any three TIFNs and \(\psi \ge 0\). Then, new operational laws for TIFNs are defined as follows:
Example 2
If we re-calculate Example 1 with this new operations of TIFNs in Definition 5, then we get the overall triangular intuitionistic information of the student \(a_{1}\) as
This new operations are more effective and practical in some cases than the other operations defined in Definition 2.
Theorem 1
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} = \left\langle {\left( {p,\,q,\,r} \right),\,\left( {t,\,u,\,v} \right)} \right\rangle ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} _{1} = \left\langle {\left( {p_{1} ,\,q_{1} ,\,r_{1} } \right),\,\left( {t_{1} ,\,u_{1} ,\,v_{1} } \right)} \right\rangle\) and \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} _{2} = \left\langle {\left( {p_{2} ,\,q_{2} ,\,r_{2} } \right),\,\left( {t_{2} ,\,u_{2} ,\,v_{2} } \right)} \right\rangle\) be any three TIFNs and \(\chi ,\,\chi_{1} ,\,\chi_{2} \ge 0\). Then
-
1.
\(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} \oplus \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} = \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} \oplus \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ;\)
-
2.
\(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} \otimes \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} = \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} \otimes \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ;\)
-
3.
\(\,\chi ( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} \oplus \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} } ) = \chi \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} \oplus \chi \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} ;\)
-
4.
\(\,( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} \otimes \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} } )^{\chi } = \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1}^{\chi } \otimes \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2}^{\chi } ;\)
-
5.
\(\,\,\chi_{1} \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} \oplus \chi_{2} \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} = \left( {\chi_{1} + \chi_{2} } \right)\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d};\)
-
6.
\(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}^{{\chi_{1} }} \otimes \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}^{{\chi_{2} }} = \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}^{{\left( {\chi_{1} + \chi_{2} } \right)}} .\)
4 Some aggregation operators for TIFNs
In this section, we defined some aggregation operator for TIFNs based on these improved operational laws.
4.1 TIF-weighted averaging and TIF-weighted geometric operator
In this section, we propose TIF-weighted averaging and TIF-weighted geometric operators to aggregate TIF information.
Definition 6
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \langle {( {p_{z} ,\,q_{z} ,\,r_{z} } ),\,( {t_{z} ,\,u_{z} ,\,v_{z} } )} \rangle (z = 1,\,2, \ldots ,\,m)\) be a family of TIFNs. If the mapping
then, TIFWA is called triangular intuitionistic fuzzy-weighted averaging (TIFWA) operator, where \(\omega = ( {\omega_{1} ,\,\omega_{2} , \ldots ,\,\omega_{m} } )^{T}\) is the importance degree of \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} (z = 1,\,2, \ldots ,\,m)\) with the condition that \(\omega_{z} \in [ {0,\,1} ],\,\sum \nolimits_{z = 1}^{m} \omega_{z} = 1.\) Especially, when the importance degree \(\omega = ( {\tfrac{1}{m},\,\tfrac{1}{m}, \ldots ,\,\tfrac{1}{m}} )^{T}\), then the TIFWA operators becomes TIFA operator:
Theorem 2
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \langle {( {p_{z} ,\,q_{z} ,\,r_{z} } ),\,( {t_{z} ,\,u_{z} ,\,v_{z} } )} \rangle (z = 1,\,2, \ldots ,\,m)\) be a family of TIFNs. Then
where \(\omega = \left( {\omega_{1} ,\,\omega_{2} , \ldots ,\,\omega_{m} } \right)^{T}\) is the importance degree of \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} (z = 1,\,2, \ldots ,\,m)\) with the condition that \(\omega_{z} \in \left[ {0,\,1} \right],\,\sum \nolimits_{z = 1}^{m} \omega_{z} = 1.\)
Proof
We prove Eq. (14) using mathematical induction on \(m.\)
-
1.
When \(m = 1,\,\;\omega_{1} = 1,\) we have
$$\begin{aligned} {\text{TrIFWA}}(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} _{1} ) & = \left\langle {\left( {1 - \left( {1 - p_{1} } \right)^{1} ,\,1 - \left( {1 - q_{1} } \right)^{1} ,\,1 - \left( {1 - r_{1} } \right)^{1} } \right)} \right.,\left( {\left( {1 - p_{1} } \right)^{1} - \left( {1 - \left( {p_{1} + t_{1} } \right)} \right)^{1} ,} \right. \\ & \quad \left. {\left. {\left( {1 - q_{1} } \right)^{1} - \left( {1 - \left( {q_{1} + u_{1} } \right)} \right)^{1} ,\left( {1 - r_{1} } \right)^{1} - \left( {1 - \left( {r_{1} + v_{1} } \right)} \right)^{1} } \right)} \right\rangle . \\ \end{aligned}$$Thus, Eq. (4) is true for \(m = 1.\)
-
2.
For \(m = 2,\), we have
$$\begin{aligned} \omega_{1} \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} &= \left\langle {\left( {1 - \left( {1 - p_{1} } \right)^{{\omega_{1} }} ,1 - \left( {1 - q_{1} } \right)^{{\omega_{1} }} ,1 - \left( {1 - r_{1} } \right)^{{\omega_{1} }} } \right)} \right.,\left( {\left( {1 - p_{1} } \right)^{{\omega_{1} }} - \left( {1 - \left( {p_{1} + t_{1} } \right)} \right)^{{\omega_{1} }} ,} \right. \hfill \\ & \quad \left( {1 - q_{1} } \right)^{{\omega_{1} }} - \left( {1 - \left( {q_{1} + u_{1} } \right)} \right)^{{\omega_{1} }} ,\left( {1 - r_{1} } \right)^{{\omega_{1} }} \left. {\left. { - \left( {1 - \left( {r_{1} + v_{1} } \right)} \right)^{{\omega_{1} }} } \right)} \right\rangle \hfill, \\ \end{aligned}$$$$\begin{aligned} \omega_{2} \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} &= \left\langle {\left( {1 - \left( {1 - p_{2} } \right)^{{\omega_{2} }} ,1 - \left( {1 - q_{2} } \right)^{{\omega_{2} }} ,1 - \left( {1 - r_{2} } \right)^{{\omega_{2} }} } \right)} \right., \hfill \\ & \quad \left. {\left( {\left( {1 - p_{2} } \right)^{{\omega_{2} }} - \left( {1 - \left( {p_{2} + t_{2} } \right)} \right)^{{\omega_{2} }} ,\left( {1 - q_{2} } \right)^{{\omega_{2} }} - \left( {1 - \left( {q_{2} + u_{2} } \right)} \right)^{{\omega_{2} }} ,\left( {1 - r_{2} } \right)^{{\omega_{2} }} - \left( {1 - \left( {r_{2} + v_{2} } \right)} \right)^{{\omega_{2} }} } \right)} \right\rangle \hfill, \\ \end{aligned}$$$$\begin{gathered} {\text{TIFWA}}\left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} _{1} ,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} _{2} } \right) = \left\langle {\left( \begin{gathered} 1 - \left( {1 - \left( {1 - \left( {1 - p_{1} } \right)^{{\omega _{1} }} } \right)} \right) \cdot 1 - \left( {1 - \left( {1 - \left( {1 - p_{2} } \right)^{{\omega _{2} }} } \right)} \right),1 - \left( {1 - \left( {1 - \left( {1 - q_{1} } \right)^{{\omega _{1} }} } \right)} \right) \hfill \\ \cdot 1 - \left( {1 - \left( {1 - \left( {1 - q_{2} } \right)^{{\omega _{2} }} } \right)} \right),1 - \left( {1 - \left( {1 - \left( {1 - r_{1} } \right)^{{\omega _{1} }} } \right)} \right) \cdot 1 - \left( {1 - \left( {1 - \left( {1 - r_{2} } \right)^{{\omega _{2} }} } \right)} \right) \hfill \\ \end{gathered} \right)} \right., \hfill \\ \left( {\left( {1 - \left( {1 - \left( {1 - p_{1} } \right)^{{\omega _{1} }} } \right)} \right) \cdot \left( {1 - \left( {1 - \left( {1 - p_{2} } \right)^{{\omega _{2} }} } \right)} \right) - \left( {1 - \left( {1 - \left( {1 - p_{1} } \right)^{{\omega _{1} }} } \right)} \right) + \left( {\left( {1 - p_{1} } \right)^{{\omega _{1} }} - \left( {1 - \left( {p_{1} + t_{1} } \right)^{{\omega _{1} }} } \right)} \right)} \right. \hfill \\ \cdot \left( {1 - \left( {1 - \left( {1 - p_{2} } \right)^{{\omega _{2} }} } \right)} \right) + \left( {\left( {1 - p_{2} } \right)^{{\omega _{2} }} - \left( {1 - \left( {p_{2} + t_{2} } \right)^{{\omega _{2} }} } \right)} \right),\left( {1 - \left( {1 - \left( {1 - q_{1} } \right)^{{\omega _{1} }} } \right)} \right) \cdot \left( {1 - \left( {1 - \left( {1 - q_{2} } \right)^{{\omega _{2} }} } \right)} \right) \hfill \\ - \left( {1 - \left( {1 - \left( {1 - q_{1} } \right)^{{\omega _{1} }} } \right)} \right) + \left( {\left( {1 - q_{1} } \right)^{{\omega _{1} }} - \left( {1 - \left( {q_{1} + u_{1} } \right)^{{\omega _{1} }} } \right)} \right) \cdot \left( {1 - \left( {1 - \left( {1 - q_{2} } \right)^{{\omega _{2} }} } \right)} \right) + \left( {\left( {1 - q_{2} } \right)^{{\omega _{2} }} - \left( {1 - \left( {q_{2} + u_{2} } \right)^{{\omega _{2} }} } \right)} \right), \hfill \\ \left( {1 - \left( {1 - \left( {1 - r_{1} } \right)^{{\omega _{1} }} } \right)} \right) \cdot \left( {1 - \left( {1 - \left( {1 - r_{2} } \right)^{{\omega _{2} }} } \right)} \right) - \left( {1 - \left( {1 - \left( {1 - r_{1} } \right)^{{\omega _{1} }} } \right)} \right) + \left( {\left( {1 - r_{1} } \right)^{{\omega _{1} }} - \left( {1 - \left( {r_{1} + v_{1} } \right)^{{\omega _{1} }} } \right)} \right) \hfill \\ \left. {\left. { \cdot \left( {1 - \left( {1 - \left( {1 - r_{2} } \right)^{{\omega _{2} }} } \right)} \right) + \left( {\left( {1 - r_{2} } \right)^{{\omega _{2} }} - \left( {1 - \left( {r_{2} + v_{2} } \right)^{{\omega _{2} }} } \right)} \right)} \right)} \right\rangle \hfill \\ = \left\langle {\left( {1 - \left( {1 - p_{1} } \right)^{{\omega _{1} }} \cdot \left( {1 - p_{2} } \right)^{{\omega _{2} }} ,1 - \left( {1 - q_{1} } \right)^{{\omega _{1} }} \cdot \left( {1 - q_{2} } \right)^{{\omega _{2} }} ,1 - \left( {1 - r_{1} } \right)^{{\omega _{1} }} \cdot \left( {1 - r_{2} } \right)^{{\omega _{2} }} } \right)} \right., \hfill \\ \left( {\left( {1 - p_{1} } \right)^{{\omega _{1} }} \cdot \left( {1 - p_{2} } \right)^{{\omega _{2} }} - \left( {\left( {1 - \left( {p_{1} + t_{1} } \right)} \right)^{{\omega _{1} }} \cdot \left( {1 - \left( {p_{2} + t_{2} } \right)} \right)^{{\omega _{2} }} } \right)} \right., \hfill \\ \left( {1 - q_{1} } \right)^{{\omega _{1} }} \cdot \left( {1 - q_{2} } \right)^{{\omega _{2} }} - \left( {\left( {1 - \left( {q_{1} + u_{1} } \right)} \right)^{{\omega _{1} }} \cdot \left( {1 - \left( {q_{2} + u_{2} } \right)} \right)^{{\omega _{2} }} } \right), \hfill \\ \left. {\left. {\left( {1 - r_{1} } \right)^{{\omega _{1} }} \cdot \left( {1 - r_{2} } \right)^{{\omega _{2} }} - \left( {\left( {1 - \left( {r_{1} + v_{1} } \right)} \right)^{{\omega _{1} }} \cdot \left( {1 - \left( {r_{2} + v_{2} } \right)} \right)^{{\omega _{2} }} } \right)} \right)} \right\rangle . \hfill \\ \end{gathered}$$That is, Eq. (14) holds for \(m = 2.\)
-
3.
Let us assume that Eq. (14) is true for \(m = l\), that is
$${\text{TIFWA}}(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} , \ldots ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{l} ) = \left\langle {\left( {1 - \mathop \prod \limits_{z = 1}^{l} \left( {1 - p_{z} } \right)^{{\omega_{z} }} ,\,1 - \mathop \prod \limits_{z = 1}^{l} \left( {1 - q_{z} } \right)^{{\omega_{z} }} ,1 - \mathop \prod \limits_{z = 1}^{l} \left( {1 - r_{z} } \right)^{{\omega_{z} }} } \right)} \right.,\left( {\mathop \prod \limits_{z = 1}^{l} \left( {1 - p_{z} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{l} \left( {1 - \left( {p_{z} + t_{z} } \right)} \right)} \right.^{{\omega_{z} }} ,$$$$\left. {\left. {\mathop \prod \limits_{z = 1}^{l} \left( {1 - q_{z} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{l} \left( {1 - \left( {q_{z} + u_{z} } \right)} \right)^{{\omega_{z} }} ,\mathop \prod \limits_{z = 1}^{l} \left( {1 - r_{z} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{l} \left( {1 - \left( {r_{z} + v_{z} } \right)} \right)^{{\omega_{z} }} } \right)} \right\rangle .$$ -
4.
Now, when \(m = l + 1,\)
$$\begin{gathered} {\text{TIFWA}}( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} _{1} ,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} _{2} , \ldots ,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} _{{l + 1}} } ) = \oplus _{{z = 1}}^{{l + 1}} \omega _{z} \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} _{z} = {\text{TIFWA}} ( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} _{1} ,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} _{2} , \ldots ,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} _{l} } ) + \omega _{{l + 1}} \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d} _{{l + 1}} \hfill \\ = \left\langle {\left( {1 - \mathop \prod \limits_{{z = 1}}^{l} \left( {1 - p_{z} } \right)^{{\omega _{z} }} ,\,1 - \mathop \prod \limits_{{z = 1}}^{l} \left( {1 - q_{z} } \right)^{{\omega _{z} }} ,1 - \mathop \prod \limits_{{z = 1}}^{l} \left( {1 - r_{z} } \right)^{{\omega _{z} }} } \right)} \right.,\left( {\mathop \prod \limits_{{z = 1}}^{l} \left( {1 - p_{z} } \right)^{{\omega _{z} }} - \mathop \prod \limits_{{z = 1}}^{l} \left( {1 - \left( {p_{z} + t_{z} } \right)} \right)^{{\omega _{z} }} ,} \right. \hfill \\ \mathop \prod \limits_{{z = 1}}^{l} \left( {1 - q_{z} } \right)^{{\omega _{z} }} - \mathop \prod \limits_{{z = 1}}^{l} \left( {1 - \left( {q_{z} + u_{z} } \right)} \right)^{{\omega _{z} }} ,\left. {\left. {\mathop \prod \limits_{{z = 1}}^{l} \left( {1 - r_{z} } \right)^{{\omega _{z} }} - \mathop \prod \limits_{{z = 1}}^{l} \left( {1 - \left( {r_{z} + v_{z} } \right)} \right)^{{\omega _{z} }} } \right)} \right\rangle \hfill \\ + \left\langle {\left( {1 - \left( {1 - p_{{l + 1}} } \right)^{{\omega _{{l + 1}} }} ,1 - \left( {1 - q_{{l + 1}} } \right)^{{\omega _{{l + 1}} }} ,1 - \left( {1 - r_{{l + 1}} } \right)^{{\omega _{{l + 1}} }} } \right),\left( {1 - \left( {1 - p_{{l + 1}} } \right)^{{\omega _{{l + 1}} }} - \left( {1 - \left( {p_{{l + 1}} + t_{{l + 1}} } \right)} \right)} \right.^{{\omega _{{l + 1}} }} ,} \right. \hfill \\ \left. {\left. {1 - \left( {1 - q_{{l + 1}} } \right)^{{\omega _{{l + 1}} }} - \left( {1 - \left( {q_{{l + 1}} + u_{{l + 1}} } \right)} \right)^{{\omega _{{l + 1}} }} ,1 - \left( {1 - r_{{l + 1}} } \right)^{{\omega _{{l + 1}} }} - \left( {1 - \left( {r_{{l + 1}} + v_{{l + 1}} } \right)} \right)^{{\omega _{{l + 1}} }} } \right)} \right\rangle \hfill \\ = \left\langle {\left( {1 - \mathop \prod \limits_{{z = 1}}^{{l + 1}} \left( {1 - p_{z} } \right)^{{\omega _{z} }} ,\,1 - \mathop \prod \limits_{{z = 1}}^{{l + 1}} \left( {1 - q_{z} } \right)^{{\omega _{z} }} ,1 - \mathop \prod \limits_{{z = 1}}^{{l + 1}} \left( {1 - r_{z} } \right)^{{\omega _{z} }} } \right)} \right.,\left( {\mathop \prod \limits_{{z = 1}}^{{l + 1}} \left( {1 - p_{z} } \right)^{{\omega _{z} }} - \mathop \prod \limits_{{z = 1}}^{{l + 1}} \left( {1 - \left( {p_{z} + t_{z} } \right)} \right)^{{\omega _{z} }} ,\mathop \prod \limits_{{z = 1}}^{{l + 1}} \left( {1 - q_{z} } \right)^{{\omega _{z} }} - } \right. \hfill \\ \left. {\left. {\mathop \prod \limits_{{z = 1}}^{{l + 1}} \left( {1 - \left( {q_{z} + u_{z} } \right)} \right)^{{\omega _{z} }} ,\mathop \prod \limits_{{z = 1}}^{{l + 1}} \left( {1 - r_{z} } \right)^{{\omega _{z} }} - \mathop \prod \limits_{{z = 1}}^{{l + 1}} \left( {1 - \left( {r_{z} + v_{z} } \right)} \right)^{{\omega _{z} }} } \right)} \right\rangle . \hfill \\ \end{gathered}$$Hence, Eq. (14) is true for \(m = l + 1.\) Therefore, Eq. (14) is true for all \(m = z.\)
Theorem 3
If \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \left\langle {\left( {p_{z} ,\,q_{z} ,\,r_{z} } \right),\,\left( {t_{z} ,\,u_{z} ,\,v_{z} } \right)} \right\rangle (z = 1,\,2, \ldots ,\,m)\) be a family of TIFNs. Then, the aggregated result using TIFWA operator is also a TIFN, i.e., \(TIFWA(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} , \ldots ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{m} ) \in TIFNs.\)
Proof
Since \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \left\langle {\left( {p_{z} ,\,q_{z} ,\,r_{z} } \right),\,\left( {t_{z} ,\,u_{z} ,\,v_{z} } \right)} \right\rangle \in {\text{TIFNs}}\;(z = 1,\,2, \ldots ,\,m),\,\;\) so by Definition 1, we have
Then
and
\({\text{Thus, TIFWA}}(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} , \ldots ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{m} )\,{\text{is}}\;{\text{TIFNs}} .\)Now, we discussed some desirable properties of TIFWA operators.
Theorem 4
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \langle {( {p_{z} ,\,q_{z} ,\,r_{z} } ),\,( {t_{z} ,\,u_{z} ,\,v_{z} } )} \rangle\) and \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{e}_{z} = \left\langle {\left( {p_{z}^{ \circ } ,\,q_{z}^{ \circ } ,\,r_{z}^{ \circ } } \right),\,\left( {t_{z}^{ \circ } ,\,u_{z}^{ \circ } ,\,v_{z}^{ \circ } } \right)} \right\rangle\) be two families of TIFNs and \(\omega = \left( {\omega_{1} ,\,\omega_{2} , \ldots ,\,\omega_{m} } \right)\) is the associated importance degree satisfying \(\omega_{z} \in \left[ {0,\,1} \right]\) and \(\sum \nolimits_{z = 1}^{m} \omega_{z} = 1.\)
-
1.
\(\left( {Idempotency} \right)\). If \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{0} = \left\langle {\left( {p_{0} ,\,q_{0} ,\,r_{0} } \right),\,\left( {t_{0} ,\,u_{0} ,\,v_{0} } \right)} \right\rangle\) for all \(z,\) then \(TIFWA(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} , \ldots ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{m} ) = \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{0} .\)
-
2.
\(( {Boundedness} )\). Let
$$\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}^{ - } = {\langle {\hbox{max} ( {0,\,\hbox{min} (p_{z} + t_{z} ) - \hbox{max} (t_{z} ),\,\hbox{min} (q_{z} + u_{z} ) - \hbox{max} (u_{z} ),\,\hbox{min} (r_{z} + v_{z} ) - \hbox{max} (v_{z} )} )}} \rangle ,\langle {{( {\hbox{max} (t_{z} ),\,\hbox{max} (u_{z} ),\,\hbox{max} (v_{z} )} )} \rangle } .$$and
$$\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}^{ + } = \left\langle {\hbox{max} \left. {\left( {\hbox{max} (p_{z} + t_{z} ) - \hbox{min} (t_{z} ),\,\hbox{max} (q_{z} + u_{z} ) - \hbox{min} (u_{z} ),\,\hbox{max} (r_{z} + v_{z} ) - \hbox{min} (v_{z} )} \right)} \right\rangle ,} \right.\left. {\left\langle {\left( {\hbox{min} (t_{z} ),\,\hbox{min} (u_{z} ),\,\hbox{min} (v_{z} )} \right)} \right.} \right\rangle.$$Then, \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}^{ - } \le TIFWA(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} , \ldots ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{m} ) \le \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}^{ + }.\)
-
3.
\(\left( {Monotonicity} \right)\). When \(p_{z} \le p_{z}^{ \circ } ,\,q_{z} \le q_{z}^{ \circ } ,\,r_{z} \le r_{z}^{ \circ } ,\,t_{z} \ge t_{z}^{ \circ } ,\,u_{z} \ge u_{z}^{ \circ } ,\,v_{z} \ge v_{z}^{ \circ } ,\,\) for all \(z,\) then \(TIFWA(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} , \ldots ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{m} ) \le TIFWA(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{e}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{e}_{2} , \ldots ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{e}_{m} ).\)
Proof
-
1.
Since \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{0} = \left\langle {\left( {p_{0} ,\,q_{0} ,\,r_{0} } \right),\,\left( {t_{0} ,\,u_{0} ,\,v_{0} } \right)} \right\rangle \,\;(z = 1,\,2, \ldots ,\,m)\) and \(\sum \nolimits_{z = 1}^{m} \omega_{z} = 1.\), then by Theorem 2, we have
$$\begin{aligned} &{\text{TIFWA}}(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} ,\ldots,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{m} ) = \left\langle {\left( {1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - p_{z}^{{}} } \right)^{{\omega_{z} }} ,\,1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - q_{z}^{{}} } \right)^{{\omega_{z} }} ,1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - r_{z}^{{}} } \right)^{{\omega_{z} }} } \right),} \right. \\ & \left. {\left( {\mathop \prod \limits_{z = 1}^{m} \left( {1 - p_{z}^{{}} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - \left( {p_{z}^{{}} + t_{z}^{{}} } \right)} \right)^{{\omega_{z} }} ,\mathop \prod \limits_{z = 1}^{m} \left( {1 - q_{z}^{{}} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - \left( {q_{z}^{{}} + u_{z}^{{}} } \right)} \right)^{{\omega_{z} }} ,\mathop \prod \limits_{z = 1}^{m} \left( {1 - r_{z}^{{}} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - \left( {r_{z}^{{}} + v_{z}^{{}} } \right)} \right)^{{\omega_{z} }} } \right)} \right\rangle ,\\ & = \left\langle {\left( {1 - \left( {1 - p_{0} } \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} ,\,1 - \left( {1 - q_{0} } \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} ,1 - \left( {1 - r_{0} } \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} } \right),} \right. \\ & \left. {\left( {\left( {1 - p_{0} } \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} - \left( {1 - \left( {p_{0} + t_{0} } \right)} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} ,\left( {1 - q_{0} } \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} - \left( {1 - \left( {q_{0} + u_{0} } \right)} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} ,\left( {1 - r_{0} } \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} - \left( {1 - \left( {r_{0} + v_{0} } \right)} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} } \right)} \right\rangle ,\\ & = \left\langle {\left( {p_{0} ,\,q_{0} ,\,r_{0} } \right),\,\left( {t_{0} ,\,u_{0} ,\,v_{0} } \right)} \right\rangle = \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{0} . \\ \end{aligned}$$ -
2.
Since
$$\begin{aligned} \mathop {\hbox{max} }\limits (t_{z} ) &= 1 - \left( {1 - \mathop {\hbox{max} }\limits (t_{z} )} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} \ge 1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - t_{z} } \right)^{{\omega_{z} }} ,\mathop {\hbox{max} }\limits (u_{z} ) & = 1 - \left( {1 - \mathop {\hbox{max} }\limits (u_{z} )} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} \ge 1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - u_{z} } \right)^{{\omega_{z} }} , \\ \mathop {\hbox{max} }\limits (v_{z} ) & = 1 - \left( {1 - \mathop {\hbox{max} }\limits (v_{z} )} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} \ge 1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - v_{z} } \right)^{{\omega_{z} }}. \\ \end{aligned}$$and
$$\begin{aligned} 1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - t_{z} } \right)^{{\omega_{z} }} \ge & 1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - \mathop {\hbox{min} }\limits (t_{z} )} \right)^{{\omega_{z} }} = 1 - \left( {1 - \mathop {\hbox{min} }\limits (t_{z} )} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} = \mathop {\hbox{min} }\limits (t_{z} ), \\ 1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - u_{z} } \right)^{{\omega_{z} }} \ge & 1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - \mathop {\hbox{min} }\limits (u_{z} )} \right)^{{\omega_{z} }} = 1 - \left( {1 - \mathop {\hbox{min} }\limits (u_{z} )} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} = \mathop {\hbox{min} }\limits (u_{z} ), \\ 1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - v_{z} } \right)^{{\omega_{z} }} \ge & 1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - \mathop {\hbox{min} }\limits (v_{z} )} \right)^{{\omega_{z} }} = 1 - \left( {1 - \mathop {\hbox{min} }\limits (v_{z} )} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} = \mathop {\hbox{min} }\limits (v_{z} ). \\ \end{aligned}$$thus
$$\begin{aligned} \mathop {\hbox{max} }\limits (p_{z} + t_{z} ) - \mathop {\hbox{min} }\limits (t_{z} ) = & \left( {1 - \mathop {\hbox{min} }\limits \left( {t_{z} } \right)} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} - \left( {1 - \mathop {\hbox{max} }\limits (p_{z} + t_{z} )} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} \ge \mathop \prod \limits_{z = 1}^{m} \left( {1 - t_{z} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - (p_{z} + t_{z} )} \right)^{{\omega_{z} }} \\ \ge & \left( {1 - \mathop {\hbox{max} }\limits \left( {t_{z} } \right)} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} \left( {1 - \mathop {\hbox{min} }\limits (p_{z} + t_{z} )} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }}=&\mathop {\hbox{min} }\limits (p_{z} + t_{z} ) - \mathop {\hbox{max} }\limits (t_{z} ), \\ \end{aligned}$$$$\begin{aligned} \mathop {\hbox{max} }\limits (q_{z} + u_{z} ) - \mathop {\hbox{min} }\limits (u_{z} ) = & \left( {1 - \mathop {\hbox{min} }\limits \left( {u_{z} } \right)} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} - \left( {1 - \mathop {\hbox{max} }\limits (q_{z} + u_{z} )} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} \ge \mathop \prod \limits_{z = 1}^{m} \left( {1 - u_{z} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - (q_{z} + u_{z} } \right))^{{\omega_{z} }} \\ \ge & \left( {1 - \mathop {\hbox{max} }\limits \left( {u_{z} } \right)} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} - \left( {1 - \mathop {\hbox{min} }\limits (q_{z} + u_{z} )} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} = & \mathop {\hbox{min} }\limits (q_{z} + u_{z} ) - \mathop {\hbox{max} }\limits (u_{z} ), \\ \end{aligned}$$$$\begin{aligned} \mathop {\hbox{max} }\limits (r_{z} + v_{z} ) - \mathop {\hbox{min} }\limits (v_{z} ) = & \left( {1 - \mathop {\hbox{min} }\limits \left( {v_{z} } \right)} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} - \left( {1 - \mathop {\hbox{max} }\limits (r_{z} + v_{z} )} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} \ge \mathop \prod \limits_{z = 1}^{m} \left( {1 - v_{z} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - (r_{z} + v_{z} } \right))^{{\omega_{z} }} \\\ge & \left( {1 - \mathop {\hbox{max} }\limits \left( {v_{z} } \right)} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} - \left( {1 - \mathop {\hbox{min} }\limits (r_{z} + v_{z} )} \right)^{{\sum \nolimits_{z = 1}^{m} \omega_{z} }} = & \mathop {\hbox{min} }\limits (r_{z} + v_{z} ) - \mathop {\hbox{max} }\limits (v_{z} ). \\ \end{aligned}$$Then, according to Theorem 3, we have
$$\mathop \prod \limits_{z = 1}^{m} \left( {1 - t_{z} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - (p_{z} + t_{z} )} \right)^{{\omega_{z} }} \ge 0,\,\mathop \prod \limits_{z = 1}^{m} \left( {1 - u_{z} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - (q_{z} + u_{z} )} \right)^{{\omega_{z} }} \ge 0,\mathop \prod \limits_{z = 1}^{m} \left( {1 - v_{z} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - (r_{z} + v_{z} )} \right)^{{\omega_{z} }} \ge 0.$$Therefore,
$$\begin{aligned} & \mathop \prod \limits_{z = 1}^{m} \left( {1 - t_{z} } \right)^{{\omega_{z} }} { - }\mathop \prod \limits_{z = 1}^{m} \left( {1 - (p_{z} + t_{z} )} \right)^{{\omega_{z} }} ,\,\mathop \prod \limits_{z = 1}^{m} \left( {1 - u_{z} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - (q_{z} + u_{z} )} \right)^{{\omega_{z} }} , & \mathop \prod \limits_{z = 1}^{m} \left( {1 - v_{z} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - (r_{z} + v_{z} )} \right)^{{\omega_{z} }}\\\ge & \left\langle {\hbox{max} \left. {\left( {0,\,\hbox{min} (p_{z} + t_{z} ) \hbox{max} (t_{z} ),\,\hbox{min} (q_{z} + u_{z} ) - \hbox{max} (u_{z} ),\,\hbox{min} (r_{z} + v_{z} ) - \hbox{max} (v_{z} )} \right)} \right\rangle ,\left\langle {\left( {\hbox{max} (t_{z} ),\,\hbox{max} (u_{z} ),\,\hbox{max} (v_{z} )} \right)} \right.} \right\rangle . \\ \end{aligned}$$Hence, we can obtain that \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}^{ - } \le {\text{TIFWA}}(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} , \ldots ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{m} ) \le \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}^{ + } .\)
-
3.
Since \(p_{z} \le p_{z}^{ \circ } ,\,q_{z} \le q_{z}^{ \circ } ,\,r_{z} \le r_{z}^{ \circ } ,\), we have \(1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - p_{z} } \right)^{{\omega_{z} }} \le 1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - p_{z}^{ \circ } } \right)^{{\omega_{z} }} ,\,1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - q_{z} } \right)^{{\omega_{z} }} \le 1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - q_{z}^{ \circ } } \right)^{{\omega_{z} }} ,1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - r_{z} } \right)^{{\omega_{z} }} \le 1 - \mathop \prod \limits_{z = 1}^{m} \left( {1 - r_{z}^{ \circ } } \right)^{{\omega_{z} }} .\) Because \(t_{z} \ge t_{z}^{ \circ } ,\,u_{z} \ge u_{z}^{ \circ } ,\,v_{z} \ge v_{z}^{ \circ } ,\,w_{z} \ge w_{z}^{ \circ } ,\,\;\) and \(p_{z} + t_{z} \le p_{z}^{ \circ } + t_{z}^{ \circ } ,\,q_{z} + u_{z} \le q_{z}^{ \circ } + u_{z}^{ \circ } ,r_{z} + v_{z} \le r_{z}^{ \circ } + v_{z}^{ \circ } ,\) \(\,\,s_{z} + w_{z} \le s_{z}^{ \circ } + w_{z}^{ \circ }\) \((z = 1,\,2, \ldots ,\,m)\), we have
$$\begin{aligned} \;\mathop \prod \limits_{z = 1}^{m} \left( {1 - t_{z} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - (p_{z} + t_{z} )} \right) \le \mathop \prod \limits_{z = 1}^{m} \left( {1 - t_{z}^{ \circ } } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - (p_{z}^{ \circ } + t_{z}^{ \circ } )} \right),\mathop \prod \limits_{z = 1}^{m} \left( {1 - u_{z} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - (q_{z} + u_{z} )} \right) \hfill \\ \, \le \mathop \prod \limits_{z = 1}^{m} \left( {1 - u_{z}^{ \circ } } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - (q_{z}^{ \circ } + u_{z}^{ \circ } )} \right),\,\mathop \prod \limits_{z = 1}^{m} \left( {1 - v_{z} } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - (r_{z} + v_{z} )} \right) \le \mathop \prod \limits_{z = 1}^{m} \left( {1 - v_{z}^{ \circ } } \right)^{{\omega_{z} }} - \mathop \prod \limits_{z = 1}^{m} \left( {1 - (r_{z}^{ \circ } + v_{z}^{ \circ } )} \right). \hfill \\ \end{aligned}$$Therefore, according to Definition 5, we have
$${\text{TIFWA}}(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} ,\ldots,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{m} ) \le {\text{TIFWA}}(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{e}_{1} ,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{e}_{2} \,,\ldots,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{e}_{m} \,).$$
Definition 7
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \langle {\left( {p_{z} ,\,q_{z} ,\,r_{z} } \right),\,\left( {t_{z} ,\,u_{z} ,\,v_{z} } \right)} \rangle (z = 1,\,2, \ldots ,\,m)\) be a family of TIFNs. If the mapping
then, TIFWG is called triangular intuitionistic fuzzy-weighted geometric operator (TIFWG), where \(\omega = ( {\omega_{1} ,\,\omega_{2} , \ldots ,\,\omega_{m} } )^{T}\) is the importance degree of \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} (z = 1,\,2, \ldots ,\,m)\) with the condition that \(\omega_{z} \in [ {0,\,1}],\,\sum \nolimits_{z = 1}^{m} \omega_{z} = 1.\) Especially, when the importance degree \(\omega =( {\tfrac{1}{m},\,\tfrac{1}{m}, \ldots ,\,\tfrac{1}{m}} )^{T}\), then the TIFWG operators becomes TIFG operator:
Theorem 5
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \langle {( {p_{z} ,\,q_{z} ,\,r_{z} } ),\,( {t_{z} ,\,u_{z} ,\,v_{z} } )} \rangle (z = 1,\,2,\ldots,\,m)\) be a family of TIFNs. Then
where \(\omega = \left( {\omega_{1} ,\,\omega_{2} , \ldots ,\,\omega_{m} } \right)^{T}\) is the importance degree of \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} (z = 1,\,2, \ldots ,\,m)\) with the condition that \(\omega_{z} \in \left[ {0,\,1} \right],\,\sum \nolimits_{z = 1}^{m} \omega_{z} = 1.\)
Proof
The proof of Theorem (5) is the same as in Theorem (2).
Theorem 6
If \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \langle {( {p_{z} ,\,q_{z} ,\,r_{z} } ),\,( {t_{z} ,\,u_{z} ,\,v_{z} } )} \rangle (z = 1,\,2, \ldots ,\,m)\) be a family of TIFNs. Then, the aggregated result using TIFWG operator is also a TIFN, i.e., \(TIFWG(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} , \ldots ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{m} ) \in TIFNs.\)
Theorem 7
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \left\langle {\left( {p_{z} ,\,q_{z} ,\,r_{z} } \right),\,\left( {t_{z} ,\,u_{z} ,\,v_{z} } \right)} \right\rangle\) and \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{e}_{z} = \left\langle {\left( {p_{z}^{ \circ } ,\,q_{z}^{ \circ } ,\,r_{z}^{ \circ } } \right),\,\left( {t_{z}^{ \circ } ,\,u_{z}^{ \circ } ,\,v_{z}^{ \circ } } \right)} \right\rangle\) be two families of TIFNs and \(\omega = \left( {\omega_{1} ,\,\omega_{2} , \ldots ,\,\omega_{m} } \right)^{T}\) is the associated importance degree satisfying \(\omega_{z} \in \left[ {0,\,1} \right]\) and \(\sum \nolimits_{z = 1}^{m} \omega_{z} = 1.\)
-
1.
\(\left( {Idempotency} \right)\). If \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{0} = \left\langle {\left( {p_{0} ,\,q_{0} ,\,r_{0} } \right),\,\left( {t_{0} ,\,u_{0} ,\,v_{0} } \right)} \right\rangle\) for all \(z,\) then \(TIFWG(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} , \ldots ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{m} ) = \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{0} .\)
-
2.
\(\left( {Boundedness} \right)\). Let
$$\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}^{ - } = \left\langle {\hbox{max} \left( {0,\,\hbox{min} (p_{z} + t_{z} ) - \hbox{max} (t_{z} ),\,\hbox{min} (q_{z} + u_{z} ) - \hbox{max} (u_{z} ),\,\hbox{min} (r_{z} + v_{z} ) - \hbox{max} (v_{z} )} \right),\left( {\hbox{max} (t_{z} ),\,\hbox{max} (u_{z} ),\,\hbox{max} (v_{z} )} \right)} \right\rangle,$$and
$$\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}^{ + } = \left\langle {\hbox{max} \left( {\hbox{max} (p_{z} + t_{z} ) - \hbox{min} (t_{z} ),\,\hbox{max} (q_{z} + u_{z} ) - \hbox{min} (u_{z} ),\,\hbox{max} (r_{z} + v_{z} ) - \hbox{min} (v_{z} )} \right),\left( {\hbox{min} (t_{z} ),\,\hbox{min} (u_{z} ),\,\hbox{min} (v_{z} )} \right)} \right\rangle.$$Then, \(d^{ - } \le TIFWG(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} , \ldots ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{m} ) \le d^{ + }.\)
-
3.
\(( {{{Monotonicity}}} )\) When \(p_{z} \le p_{z}^{ \circ } ,\,q_{z} \le q_{z}^{ \circ } ,\,r_{z} \le r_{z}^{ \circ } ,\,t_{z} \ge t_{z}^{ \circ } ,\,u_{z} \ge u_{z}^{ \circ } ,\,v_{z} \ge v_{z}^{ \circ } ,\,\) for all \(z,\) then \(TIFWG(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} , \ldots ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{m} ) \le TIFWG(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{e}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{e}_{2} , \ldots ,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{e}_{m} \,).\)
Proof
Same as Theorem 4.
5 Some hybrid aggregation operators for TIFNs
In this section, some ordered weighted averaging, ordered weighted geometric operators and hybrid weighted averaging, hybrid weighted geometric operators are developed.
Definition 8
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \langle {( {p_{z} ,\,q_{z} ,\,r_{z} } ),\,( {t_{z} ,\,u_{z} ,\,v_{z} } )} \rangle (z = 1,\,2, \ldots ,\,m)\) be a family of TIFNs. If the mapping
Then, TIFOWA is called triangular intuitionistic fuzzy-ordered-weighted averaging operator, where \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (z)}\) is the \(z\)th largest value of \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{(z)} (z = 1,\,2, \ldots ,\,m),\,\omega = ( {\omega_{1} ,\,\omega_{2} , \ldots ,\,\omega_{m} } )^{T} ,\) is the importance degree of TIFOWA operator with \(\omega_{z} \in [ {0,\,1}],\) and \(\sum \nolimits_{z = 1}^{m} \omega_{z} = 1.\) In a special case, when \(\omega = ( {\tfrac{1}{m},\,\tfrac{1}{m}, \ldots ,\,\tfrac{1}{m}} )^{T} ,\) then the TIFOWA operator reduce to TIFA operator.
Definition 9
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \langle {( {p_{z} ,\,q_{z} ,\,r_{z} } ),\,( {t_{z} ,\,u_{z} ,\,v_{z} } )} \rangle (z = 1,\,2, \ldots ,\,m)\) be a family of TIFNs. If the mapping
then TIFOWG is called triangular intuitionistic fuzzy-ordered-weighted geometric operator, where \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (z)}\) is the \(z\) th largest value of \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{(z)} (z = 1,\,2,\ldots,\,m),\,\omega = ( {\omega_{1} ,\,\omega_{2} ,\ldots,\,\omega_{m} } )^{T}\) is the importance degree of TIFOWG operator with \(\omega_{z} \in [ {0,\,1} ],\) and \(\sum \nolimits_{z = 1}^{m} \omega_{z} = 1.\) In a special case, when \(\omega = ( {\tfrac{1}{m},\,\tfrac{1}{m}, \ldots ,\,\tfrac{1}{m}} )^{T} ,\) then the TIFOWG operator reduced to TIFA operator.
Theorem 8
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \langle {( {p_{z} ,\,q_{z} ,\,r_{z} }),\,( {t_{z} ,\,u_{z} ,\,v_{z} } )} \rangle (z = 1,\,2, \ldots ,\,m)\) be a family of TIFNs. Then
where \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (z)}\) is the \(zth\) largest value of \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{(z)} (z = 1,\,2, \ldots ,\,m),\,\omega = \left( {\omega_{1} ,\,\omega_{2} , \ldots ,\,\omega_{z} } \right)^{T}\) is the importance degree of TIFOWG operator with \(\omega_{z} \in [ {0,\,1} ],\) and \(\sum \nolimits_{z = 1}^{m} \omega_{z} = 1.\)
Proof
The proof is similar to Theorem 2.
Theorem 9
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \langle {( {p_{z} ,\,q_{z} ,\,r_{z} } ),\,( {t_{z} ,\,u_{z} ,\,v_{z} })} \rangle (z = 1,\,2, \ldots ,\,m)\) be a family of TIFNs. Then
where \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (z)}\) is the \(zth\) largest value of \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{(z)} (z = 1,\,2, \ldots ,\,m),\,\omega = ( {\omega_{1} ,\,\omega_{2} , \ldots ,\,\omega_{m} } )^{T}\) is the importance degree of TIFOWG operator with \(\omega_{z} \in [ {0,\,1} ],\) and \(\sum \nolimits_{z = 1}^{m} \omega_{z} = 1.\)
Proof
Same as Theorem 2.The TIFOWA and the TIFWG operators have the same properties than the TIFWA and TIFWG operators. However, TIFOWA and TIFOWG operators also satisfy the property of commutativity.
Theorem 10
Assume that \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z}^{ \cdot } = \langle {( {p_{z}^{ \cdot } ,\,q_{z}^{ \cdot } ,\,r_{z}^{ \cdot } } ),\,( {t_{z}^{ \cdot } ,\,u_{z}^{ \cdot } ,\,v_{z}^{ \cdot } } )} \rangle (z = 1,\,2, \ldots ,\,m)\) is any permutation of \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \langle {( {p_{z} ,\,q_{z} ,\,r_{z} } ),\,( {t_{z} ,\,u_{z} ,\,v_{z} } )} \rangle (z = 1,\,2, \ldots ,\,m),\) then
Proof
Since
\(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z}^{ \cdot } = \left\langle {\left( {p_{z}^{ \cdot } ,\,q_{z}^{ \cdot } ,\,r_{z}^{ \cdot } } \right),\,\left( {t_{z}^{ \cdot } ,\,u_{z}^{ \cdot } ,\,v_{z}^{ \cdot } } \right)} \right\rangle (z = 1,\,2,\ldots,\,m)\) is any permutation of \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = \langle {( {p_{z} ,\,q_{z} ,\,r_{z} } ),\,( {t_{z} ,\,u_{z} ,\,v_{z} })} \rangle (z = 1,\,2,\ldots,\,m),\) then \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (1)}^{ \cdot } = \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (1)}\)
Since the TIFWA and TIFWG operators only consider the importance degree of TIFNs and cannot consider the importance degree of the position, the TIFOWA and TIFOWG operators only consider the importance degree of the position and cannot consider the importance degree of the TIFNs. To overcome this limitation, we proposed TIF hybrid aggregation operators.
Definition 10
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (z)} = \langle {( {p_{\sigma (z)} ,\,q_{\sigma (z)} ,\,r_{\sigma (z)} } ),\,( {t_{\sigma (z)} ,\,u_{\sigma (z)} ,\,v_{\sigma (z)} } )} \rangle (z = 1,\,2,\ldots,\,m)\) be a family of TIFNs. Then, if the mapping
then \({\text{TIFHA}}_{\omega ,\,\varpi }\) is said to be triangular intuitionistic fuzzy hybrid averaging operator, where \(\varpi = ( {\varpi_{1} ,\,\varpi_{2} ,\ldots,\,\varpi_{z} } )^{T}\) is the importance degree of the TIFHA operator with \(\varpi_{z} \in [ {0,\,1}],\,\sum \nolimits_{z = 1}^{m} \varpi_{z} = 1.\) and is the \(zth\;\) largest of the TIF values \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = m\omega_{z} \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} \;(z = 1,\,2,\ldots,\,m),\) \(\;( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (1)} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (2)} ,\ldots.,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (z)} } )\) is any permutation of the weighted TIFNs \(( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} ,\ldots.,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} })\) which satisfies \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (z - 1)} \ge \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (z)} \;(z = 1,\,2,\ldots,\,m),\) \(\omega = ( {\omega_{1} ,\,\omega_{2} ,\ldots,\,\omega_{z} })^{T}\) is the importance degree of \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} \;(z = 1,\,2,\ldots,\,m)\), such that \(\omega_{z} \in [ {0,\,1} ],\,\sum \nolimits_{z = 1}^{m} \omega_{z} = 1,\) and \(m\) is the balancing coefficient.
Assume that \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (z)} = \langle {( {p_{\sigma (z)} ,\,q_{\sigma (z)} ,\,r_{\sigma (z)} } ),\,( {t_{\sigma (z)} ,\,u_{\sigma (z)} ,\,v_{\sigma (z)} } )} \rangle (z = 1,\,2,\ldots,\,m)\) be a family of TIFNs. Then, we have
Definition 11
Let \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (z)} = \langle {( {p_{\sigma (z)} ,\,q_{\sigma (z)} ,\,r_{\sigma (z)} } ),\,( {t_{\sigma (z)} ,\,u_{\sigma (z)} ,\,v_{\sigma (z)} } )} \rangle (z = 1,\,2,\ldots,\,m)\) be a family of TIFNs. Then, if the mapping
then \({\text{TIFHG}}_{\omega ,\,\varpi }\) is called triangular intuitionistic fuzzy hybrid geometric operator, where \(\varpi = \left( {\varpi_{1} ,\,\varpi_{2} ,\ldots,\,\varpi_{z} } \right)^{T}\) is the importance degree of the TIFHG operator with \(\varpi_{z} \in \left[ {0,\,1} \right],\,\sum \nolimits_{z = 1}^{m} \varpi_{z} = 1.\) and \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (z)}^{ \cdot \cdot }\) is the \(z - th\;\) largest of the TIF values \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} = m\omega_{z} \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} \;(z = 1,\,2,\ldots,\,m),\,\;( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (1)} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (2)} ,\ldots.,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (z)} } )\) is any permutation of the weighted TIFNs \(( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{2} ,\ldots.,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} } )\) which satisfies \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (z - 1)} \ge \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (z)} \;(z = 1,\,2,\ldots,\,m),\) \(\omega = \left( {\omega_{1} ,\,\omega_{2} ,\ldots,\,\omega_{z} } \right)^{T}\) is the importance degree of \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{z} \;(z = 1,\,2,\ldots,\,m),\) such that \(\omega_{z} \in {0,\,1}],\,\sum \nolimits_{z = 1}^{m} \omega_{z} = 1\) and \(m\) is the balancing coefficient.
Assume that \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{\sigma (z)} = \langle {( {p_{\sigma (z)} ,\,q_{\sigma (z)} ,\,r_{\sigma (z)} } ),\,( {t_{\sigma (z)} ,\,u_{\sigma (z)} ,\,v_{\sigma (z)} } )} \rangle (z = 1,\,2,\ldots,\,m)\) be a family of TIFNs. Then, we have
The proof is similar to the above theorems.
6 The MCDM approach with TIFNs
For MCDM problem with TIF information, assume that there is a set of the alternatives \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A} = \{ {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A} _{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A} _{2} ,...,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A} _{n} } \}\), and there is a set of criteria \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{C} = \{ {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{C}_{1} ,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{C}_{2} ,\ldots,\,\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{C}_{m} } \},\) with associated importance degree \(\omega = \left( {\omega_{1} ,\,\omega_{2} ,\ldots,\,\omega_{m} } \right)^{T} ,\) satisfying \(\omega_{z} \in \left[ {0,\,1} \right]\) and \(\sum \nolimits_{z = 1}^{m} \omega_{z} = 1.\) The decision maker can give the criteria value \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{C}_{z} (z = 1,2, \ldots ,m)\) of alternative \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{y} (y = 1,\,2,\ldots,\,n)\) by the form of TIFNs \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{yz} = \left\langle {\left( {p_{yz} ,\,q_{yz} ,\,r_{yz} } \right),\,\left( {t_{yz} ,\,u_{yz} ,\,v_{yz} } \right)} \right\rangle (y = 1,\,2,\ldots,\,n,\,z = 1,\,2,\ldots,\,m),\) where \(\left( {p_{yz} ,\,q_{yz} ,\,r_{yz} } \right)\) indicates the degree of membership of the alternative \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{y}\) with respect to the criteria \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{C}_{z} ,\,\left( {t_{yz} ,\,u_{z} ,\,v_{z} } \right)\) indicates the degree of non-membership of the alternative \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{y}\) with respect to the criteria \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{C}_{z} \;(y = 1,\,2,\ldots,\,n,\,z = 1,\,2,\ldots,\,m).\) Assume that \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{D} = (\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{yz} )_{n \times m}\) is the decision matrix. Consequently, a ranking of alternatives is required.
In general, the decision steps of this MCDM problem are shown as follows.
Step 1: Calculate the overall TIFN \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{y}\) \((y = 1,\,2,\ldots,\,n)\) for the alternative \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{y} (y = 1,\,2,\ldots,\,n)\) which is shown by
or
where \(y = 1,\,2,\ldots,\,n\).
Step 2: Calculate the score and accuracy values of the final TIFNs \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{y} (y = 1,\,2,\ldots,\,n)\) by the following formulas:
Step 3: Rank all the alternatives and select the best one.
Use the comparison method defined in Definition 4 to rank the alternatives and select the best one(s). First, we can compare with score values of all alternatives, the bigger the score value for one alternative is, the better this alternative is. If the score values for some alternatives are equal, then we can compare with them by accuracy values of all alternatives. The bigger the accuracy value for one alternative is, the better this alternative is.
7 Illustrated example
The following example is adapted on the basis of the case used by Herrera et al. (2000).
Example 3
Let us assume that an investment company wants to invest a sum of money in the best option. There are four possible companies are taken into consideration, which are described as follows:
-
1.
\(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{1}\) is a car company.
-
2.
\(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{2}\) is a food company.
-
3.
\(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{3}\) is a computer company.
-
4.
\(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{4}\) is an arm company.
The investment company must take a decision according to the following four criteria:
-
(a)
\(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{C}_{1}\) is the risk analysis.
-
(b)
\(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{C}_{2}\) is the growth analysis.
-
(c)
\(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{C}_{3}\) is the social–political impact analysis.
-
(d)
\(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{C}_{4}\) is the environmental impact.
The importance degree of the criteria is \(\omega = (0.2,\,0.2,\,0.3,\,0.3)^{T}\), and the assessment values of each alternative given by the decision makers are in the form of TIFNs and are shown in Table 1.The goal of this decision problem is to select one best company for investing.
7.1 Decision steps of the proposed method
Step 1: The overall evaluation value of each alternative \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{y}\) \((y = 1,\,2,3,4)\) is obtained using Eq. (27) and is given as follows:
Step 2: Using Eqs. (29) and (30), to calculate the score and accuracy values of the overall evaluation values, we can obtain
Step 3: By the comparison rules defined in Definition 4, we can get the ranking order of alternatives as follows:
Therefore, the best alternative is \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{2}\) and the worst is \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{4} .\)
In a similar way, the following procedure can be done by TIFWG operator.
Step 1: The overall evaluation value of each alternative is obtained using Eq. (28) which is given as follows:
Step 2: Using Eqs. (29) and (30), to calculate the score and accuracy values of the overall evaluation values, we can obtain
Step 3: By the comparison rules defined in Definition 4, we can get the ranking order of alternatives as follows:
Therefore, the best alternative is \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{2}\) and the worst is \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{4} .\)
Obviously, the ranking results produced by the TIFWA operator and TIFWG operator are the same, i.e., the best alternative is \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{2}\) and the worst is \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{4} .\)
7.2 The validity of the proposed method compared with some existing methods
To show the validity of the proposed method in this paper, now, we calculate Example 3 using aggregation operators proposed by Wang (2008a, b), Zhou and Chang (2014) and Wang et al. (2015), and then compare with the ranking results.
(1) Compared with the method proposed by Wang (2008a, b)
The steps are shown as follows.
Step 1: The overall evaluation value of each alternative \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{y}\) \((y = 1,\,2,3,4)\) is obtained using FNIFWA operator defined by Wang (2008a, b), and we have
Step 2: Using Eqs. (29) and (30), to calculate the score and accuracy values of the overall evaluation values, we can obtain
Step 3: Therefore, the ranking order according to their score values is
Therefore, the best alternative is \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{2}\) and the worst is \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{4} .\)
Obviously, this ranking result is the same as one produced by the proposed method in this paper.
In a similar way, when we use the FNIFWG operator proposed by Wang (2008a, b) to solve this problem, we can get the following steps.
Step 1: The overall evaluation value of each alternative is obtained using FNIFWG operator, and we have
Step 2: Using Eqs. (29) and (30), to calculate the score and accuracy values of the overall evaluation values, we can obtain
Step 3: Therefore, the ranking order according to their score values is
Therefore, the best alternative is \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{2}\) and the worst is \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{4} .\)
This result is the same as the above all ranking results. Therefore, it can show the validity of the proposed method in this paper.
(2) Compared with the method proposed by Zhou and Chang (2014)
In the part, we can compare the proposed method with the method proposed by Wang (2008a, b), and the steps are shown as follows.
Step 1: The overall evaluation value of each alternative \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{d}_{y}\) \((y = 1,\,2,3,4)\) is obtained by the FNIFHWA operator defined by Zhou and Chang (2014), and we have
Step 2: Using Eqs. (29) and (30), to calculate the score and accuracy values of the overall evaluation values, we can obtain
Step 3: Therefore, the ranking order according to their score values is
Obviously, this ranking result is the same as ones produced by the proposed method in this paper and by Wang (2008a, b)’ method.
(3) Compared with the method proposed by Wang et al. (2015)
In a similar way, we can compare the proposed method with the method proposed by Wang et al. (2015), and the steps are shown as follows.
Step 1: The overall evaluation value of each alternative is obtained by the FNIFHWG operator, and we have
Step 2: Using Eqs. (29) and (30), to calculate the score and accuracy values of the overall evaluation values, we can obtain
Step 3: Therefore, the ranking order according to their score values is
Obviously, this ranking result is the same as ones produced by the proposed method in this paper, by Wang (2008a, b)’ method and by Wang et al. (2015)’ method.
This show that our proposed method is valid based on these improved operational laws.
7.3 Comparison and discussion
Because all ranking results are all same compared with above three methods, it is difficult to show the advantage of the proposed method. In this part, we give some revised data from Example 3. Since the triangular intuitionistic fuzzy sets are a generalization of the IFSs and are a better mathematical tool to handle uncertain and inconsistent information then IFS.
For comparison, we take Example 3 with the TIF information values, as given in Table 2.
(1) Ranking by the method in Wang (2008a, b)
Now, the steps based on the FNIFWA operator defined by Wang (2008a, b) are shown as follows.
Step 1: The overall evaluation values obtained using FNIFWA defined by Wang (2008a, b), and we have
Step 2: Using Eqs. (29) and (30), to calculate the score and accuracy values of the overall evaluation values, we can obtain
Step 3: Therefore, the ranking order according to their score values is \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{1} > \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{2} > \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{3} > \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{4} .\)
Then, the steps based on the FNIFWG operator defined by Wang (2008a, b) are shown as follows.
Step 1: The overall evaluation values is obtained using the FNIFWG operator defined Wang (2008a, b), and we have
Step 2: Using Eqs. (29) and (30), to calculate the score and accuracy values of the overall evaluation values, we can obtain
Step 3: Therefore, the ranking order according to their score values is
Obviously, this ranking result is different from the one produced by the FNIFWA operator defined by Wang (2008a, b).
(2) Ranking by the method in Zhou and Chang (2014)
We use the method in Zhou and Chang (2014) to solve this problem and the steps are shown as follows.
Step 1: The overall evaluation values obtained by the FNIFHWA defined by Zhou and Chang (2014), and we have
Step 2: Using Eqs. (29) and (30), to calculate the score and accuracy values of the overall evaluation values, we can obtain
Step 3: Therefore, the ranking order according to their score values is
Obviously, this ranking result is different from the ones produced by the method defined by Wang (2008a, b).
(3) Ranking by the method in Wang et al. (2015)
The steps are shown as follows.
Step 1: The overall evaluation values is obtained by the the FNIFHWG operator defined Wang et al. (2015), and we have
Step 2: Using Eqs. (29) and (30), to calculate the score and accuracy values of the overall evaluation values, we can obtain
Step 3: Therefore, the ranking order according to their score values is
This result is the same as one produced by the FNIFWG operator defined by Wang (2008a, b), and is different from the other ranking results.
(4) Ranking by the proposed method in this paper
Now, we use the proposed method in this paper to solve this problem, and steps are shown as follows.
Step 1: The overall evaluation value is obtained using Eq. (27) as follows:
Step 2: Using Eqs. (29) and (30), we get
Step 3: Therefore, the ranking order according to their score values is
Therefore, the best alternative is \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{2}\) and the worst is \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{4}\).
Now, by the TIFWG operator defined in this article, we have
Step 1: The overall evaluation values is obtained using Eq. (28), and is given below:
Step 2: Using Eqs. (29) and (30), we can get
Step 3: Therefore, the ranking order according to their score values is \(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{2} > \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{3} > \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{4} > \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{A}_{1} .\).
Obviously, the ranking results by the proposed method are different from the ones by the other methods.
The reason produced these results is that these methods adopt the different operational laws of FNIFNs. In Example 3, because all membership and non-membership degrees are not zero, thus, all methods in Wang (2008a, b), Zhou and Chang (2014), Wang et al. (2015) and our method in this paper can produce right ranking results. However, in the revised example, we revise some data to zero, including some membership degree and some non-membership degrees, so the methods in Wang (2008a, b), Zhou and Chang (2014) and Wang et al. (2015) will not give the right results, because the operational laws of FNIFNs used in these methods may result in the unreasonable results which are explained in Example 1, while the proposed method adopts improved operational laws of FNIFNs which can overcome these shortcomings, and it can give a reasonable ranking results. Therefore, the proposed aggregation operators and the method in this article are more practical and effective in the decision-making process.
8 Conclusion
In this paper, we pointed out some existing limitations in the operations of TIFNs and proposed some improved operational laws for TIFNs. Then, based on these improved operational laws, we propose some aggregation operators such as TIFWA operator, TIFWG operator, TIFOWG operator, TIFOWG operator, and TIFHA operator and TIFHG operators, and discussed some desirable properties of these operators. Furthermore, based on these aggregation operators, we define an MCDM method in which the alternatives values with respect to criteria are represented in the form of TIFNs. A numerical example is illustrated to show the practicality and effectiveness of the proposed MCDM method. Finally, a comparison has been made with the existing method to show that the proposed aggregation operators and MCDM method in this paper are more practical and effective in the decision-making process, because they have solved the existing limitations in the operations of TIFNs.
In the future, we shall define Bonferroni mean, Heronian mean, and scaled prioritized aggregation operators for TIFSs based on these new operational laws and applied them to MCGDM problems.
References
Atanassov KT (1986) Intuitionistic fuzzy sets. Fuzzy Sets Syst 20(1):87–96
Chen SM, Chang CH (2015) A novel similarity measure between Atanassov’s intuitionistic fuzzy sets based on transformation techniques with applications to pattern recognition. Inf Sci 291:96–114
Chen SM, Chang CH (2016) Fuzzy multiattribute decision making based on transformation techniques of intuitionistic fuzzy values and intuitionistic fuzzy geometric averaging operators. Inf Sci 352–353:133–149
Chen SM, Hong JA (2014a) Fuzzy multiple attributes group decision making based on ranking interval type-2 fuzzy sets and the TOPSIS method. IEEE Trans Syst Man Cybern Syst 44(12):1665–1673
Chen SM, Hong JA (2014b) Multicriteria linguistic decision making based on hesitant fuzzy linguistic term sets and the aggregation of fuzzy sets. Inf Sci 286:63–74
Chen SM, Randyanto Y (2013) A novel similarity measure between intuitionistic fuzzy sets and its applications. Int J Pattern Recognit Artif Intell 27(7):1350021-1–1350021-34
Chen SM, Lin TE, Lee LW (2014) Group decision making using incomplete fuzzy preference relations based on the additive consistency and the order consistency. Inf Sci 259:1–15
Chen SM, Cheng SH, Chiou CH (2016a) Fuzzy multiattribute group decision making based on intuitionistic fuzzy sets and evidential reasoning methodology. Inf Fusion 27:215–227
Chen SM, Cheng SH, Lan TC (2016b) A novel similarity measure between intuitionistic fuzzy sets based on the centroid points of transformed fuzzy numbers with applications to pattern recognition. Inf Sci 343–344:15–40
Chen SM, Cheng SH, Lan TC (2016c) Multicriteria decision making based on the TOPSIS method and similarity measures between intuitionistic fuzzy values. Inf Sci 367–368:279–295
Garg H (2016) A novel approach for analyzing the reliability of series-parallel system using credibility theory and different types of intuitionistic fuzzy numbers. J Brazilian Soc Mech Sci Eng 38(3):1021–1035
Herrera F, Herrera-Viedma E, Martínez L (2000) A fusion approach for managing multi-granularity linguistic term sets in decision making. Fuzzy Sets Syst 114(1):43–58
Jiang Y, Xu Z, Shu Y (2017) Interval-valued intuitionistic multiplicative aggregation in group decision making. Granul Comput. https://doi.org/10.1007/s41066-017-0050-9
Liang CY, Zhao SP, Zhang JL (2014) Aggregation operators on triangular intuitionistic fuzzy numbers and its application to multi-criteria decision making problems. Found Comput Decis Sci 39(3):189–208
Liu P (2017) Multiple attribute group decision making method based on interval-valued intuitionistic fuzzy power Heronian aggregation operators. Comput Ind Eng 108:199–212
Liu P, Chen SM (2017) Group decision making based on Heronian aggregation operators of intuitionistic fuzzy numbers. IEEE Trans Cybern 47(9):2514–2530
Liu P, Li H (2017a) Multiple attribute decision-making method based on some normal neutrosophic Bonferroni mean operators. Neural Comput Appl 28(1):179–194
Liu P, Li H (2017b) Interval-valued intuitionistic fuzzy power Bonferroni aggregation operators and their application to group decision making. Cogn Comput 9(4):494–512
Liu P, Qin X (2017) Maclaurin symmetric mean operators of linguistic intuitionistic fuzzy numbers and their application to multiple-attribute decision-making. J Exp Theor Artif Intell 1–30
Liu P, Shi L (2017) Some Neutrosophic uncertain linguistic number Heronian mean operators and their application to multi-attribute group decision making. Neural Comput Appl 28(5):1079–1093
Liu P, Su Y (2010) The multiple-attribute decision making method based on the TFLHOWA operator. Comput Math Appl 60(9):2609–2615
Liu P, Tang G (2016) Multi-criteria group decision-making based on interval neutrosophic uncertain linguistic variables and Choquet integral. Cogn Comput 8(6):1036–1056
Liu P, Wang P (2017) Some improved linguistic intuitionistic fuzzy aggregation operators and their applications to multiple-attribute decision making. Int J Inf Technol Decis Mak 16(3):817–850
Liu F, Yuan XH (2007) Fuzzy number intuitionistic fuzzy set. Fuzzy Syst Math 21(1):88–91
Liu P, Zhang L, Liu X, Wang P (2016) Multi-valued Neutrosophic number Bonferroni mean operators and their application in multiple attribute group decision making. Int J Inf Technol Decis Mak 15(5):1181–1210
Liu P, Chen SM, Liu J (2017) Some intuitionistic fuzzy interaction partitioned Bonferroni mean operators and their application to multi-attribute group decision making. Inf Sci 411:98–121
Pedrycz W, Chen SM (2015) Granular computing and decision-making: interactive and iterative approaches. Springer, Heidelberg
Qin J (2017) Interval type-2 fuzzy Hamy mean operators and their application in multiple criteria decision making. Granul Comput. https://doi.org/10.1007/s41066-017-0041-x
Qiu T (2011) Evaluating model of mechanical automation with fuzzy number intuitionistic fuzzy information. IJACT Int J Adv Comput Technol 3(6):42–47
Şahin R, Liu P (2017) Possibility-induced simplified neutrosophic aggregation operators and their application to multi-criteria group decision-making. J Exp Theor Artif Intell 29(4):769–785
Wang X (2008a) Fuzzy number intuitionistic fuzzy arithmetic aggregation operators. Int J Fuzzy Syst 10(2):104–126
Wang X (2008b) Fuzzy number intuitionistic fuzzy geometric aggregation operators and their application to decision making. Control Decis 23(6):607–612
Wang Y (2012) An approach to software selection with triangular intuitionistic fuzzy information. Int J Adv Comput Technol 4(2):284–290
Wang Y, Peng X, Bian J (2015) Study on the security of information system authentication scheme based on the fuzzy number intuitionistic fuzzy information. J Intell Fuzzy Syst 28(5):2225–2232
Wei G, Zhao X, Lin R (2010) Some induced aggregating operators with fuzzy number intuitionistic fuzzy information and their applications to group decision making. Int J Comput Intell Syst 3(1):84–95
Xu Z (2007a) Intuitionistic fuzzy aggregation operators. IEEE Trans Fuzzy Syst 15(6):1179–1187
Xu Z (2007b) Intuitionistic preference relations and their application in group decision making. Inf Sci 177(11):2363–2379
Xu Z, Gou X (2017) An overview of interval-valued intuitionistic fuzzy information aggregations and applications. Granul Comput 2(1):13–39
Xu Z, Yager RR (2006) Some geometric aggregation operators based on intuitionistic fuzzy sets. Int J Gen Syst 35(4):417–433
Ye J (2014) Prioritized aggregation operators of trapezoidal intuitionistic fuzzy sets and their application to multicriteria decision-making. Neural Comput Appl 25(6):1447–1454
Ye J (2015) Trapezoidal neutrosophic set and its application to multiple attribute decision-making. Neural Comput Appl 26(5):1157–1166
Ye J (2016) Projection and bidirectional projection measures of single-valued neutrosophic sets and their decision-making method for mechanical design schemes. J Exp Theor Artif Intell 29(4):731–740
Zhao Q, Chen H, Zhou L, Tao Z, Liu X (2015) The properties of fuzzy number intuitionistic fuzzy prioritized operators and their applications to multi-criteria group decision making. J Intell Fuzzy Syst 28(4):1835–1848
Zhou S, Chang W (2014) Approach to multiple attribute decision making based on the Hamacher operation with fuzzy number intuitionistic fuzzy information and their application. J Intell Fuzzy Syst 27(3):1087–1094
Acknowledgements
This paper is supported by the National Natural Science Foundation of China (Nos. 71771140, 71471172, and 71271124), the Special Funds of Taishan Scholars Project of Shandong Province (No. ts201511045), Shandong Provincial Social Science Planning Project (Nos. 16CGLJ31 and 16CKJJ27), the Natural Science Foundation of Shandong Province (No. ZR2017MG007), the Teaching Reform Research Project of Undergraduate Colleges and Universities in Shandong Province (No. 2015Z057), and Key research and development program of Shandong Province (No. 2016GNC110016).
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Mahmood, T., Liu, P., Ye, J. et al. Several hybrid aggregation operators for triangular intuitionistic fuzzy set and their application in multi-criteria decision making. Granul. Comput. 3, 153–168 (2018). https://doi.org/10.1007/s41066-017-0061-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s41066-017-0061-6