mathematics

Article
Analysis of Robot Selection Based on 2-Tuple Picture
Fuzzy Linguistic Aggregation Operators
Arshad Ahmad Khan 1 , Muhammad Qiyas 2 , Saleem Abdullah 2, *, Jianchao Luo 1, * and
Mahwish Bano 3
1 College of Economics and Management, Northwest A&F University, Yangling 712100, China;
arshadkhan@[Link]
2 Department of Mathematics, Abdul Wali Khan University, Mardan 23200, Pakistan;
muhammadqiyas@[Link]
3 Department of Mathematics, Air University, Islamabad 44000, Pakistan; mahwish@[Link]
* Correspondence: jchluo@[Link] (J.L); saleemabdullah@[Link] (S.A.)




Received: 12 September 2019; Accepted: 12 October 2019; Published: 21 October 2019


Abstract: The aim of this article is to propose the 2-tuple picture fuzzy linguistic aggregation
operators and a decision-making model to deal with uncertainties in the form of 2-tuple picture fuzzy
linguistic sets; 2-tuple picture fuzzy linguistic operators have more flexibility than general fuzzy
set. We proposed a number of aggregation operators, namely, 2-TPFLWA, 2-TPFLOWA, 2-TPFLHA,
2-TPFLWG, 2-TPFLOWG, and 2-TPFLHG operators. The distinguished feature of the developed
operators are studied. At that point, we used these operators to design a model to deal with multiple
attribute decision-making issues under the 2-tuple picture fuzzy linguistic information. Then, a
practical application of robot selection by manufacturing unit is given to prove the introduced
technique and to show its practicability and effectiveness. Besides this, a systematic comparison
analysis with other existent approaches is conducted to reveal the advantage of our developed method.
Results indicate that the proposed method is suitable and effective for decision-making problems.

Keywords: picture fuzzy set; 2-tuple linguistic model; 2-tulpe picture fuzzy linguistic set

1. Introduction
Generally, there exists some uncertainty in the representation of data information during the
decision-making (DM) process. To overcome this drawback, first, Zadeh [1] defined fuzzy set (FS).
In fuzzy set, Zadeh only showed the positive membership grade of a number in the defined set,
and applied it to many other fields, i.e., fuzzy decision-making problems [2,3]. However, there was no
discussion for the negative membership grade. Therefore, due to this non-membership grade, the fuzzy
set theory failed to solve the complete uncertainty in real-life problem. For this purpose, Atanassov [4]
defined the concept of intuitionistic fuzzy set (IFS), which consists of positive and negative membership
grades; IFS has the advantage of two memberships, which diminish the fuzziness. Garg [5,6] displayed
generalized improved interactive aggregation operators to solve a decision-making problem under
the IF set condition. Kaur and Garg [7] introduced cubic intuitionistic fuzzy set aggregation operator.
Later, Kaur & Garg [8] developed generalized aggregation operators with the cubic IF set information
using the t-norm operations. Aside from them, different scholars (Garg & Arora [9]; Garg & Rani [10];
Shen & Wang [11]; Peng, & Wang [12]; Hoskova & Maturo [13]; Collan & Kacprzyk [14]) incorporated
the idea of aggregation process in different applications and also given their related decision-making
algorithm with the IF set and their expansion.
Atanassov discussed only two categories of responses i.e., “yes” and “no”, but we have three
types of responses in case of selection, for example, “yes”, “no”, and “refusal”, and the complicated

Mathematics 2019, 7, 1000; doi:10.3390/math7101000 [Link]/journal/mathematics

Mathematics 2019, 7, 1000 2 of 19

answer is “refusal”. Therefore, to overcome this drawback, Cuong [15,16] introduced a novel notion of
picture fuzzy set (PFS), which dignified in three different functions, presenting the positive, neutral,
and negative membership grades. Cuong [17] discussed some characteristic of PF sets and also
approved their distance measures. Cuong and Hai [18] introduced fuzzy logic operators and defined
fundamental operations for fuzzy derivation forms in the PF logic. Cuong, Kreinovich, and Ngan [19]
examined the characteristic of picture fuzzy t-norm and t-conorm. Phong et al. [20] explored certain
configuration of picture fuzzy relations. Wei et al. [21–23] defined many procedures to compute the
closeness between PFSs. Presently, many researchers have developed more models in the PF sets
condition: Correlation coefficients of PFS are proposed by Singh [24] and apply it to clustering analysis.
Son et al. and Thong [25,26] provided time arrangement calculation and temperature estimation on the
basis of PFSs domain. Son [27,28] defined PF as separation, distance and association measures, also
combined them under the condition of PF sets. Van et al. [29] defined novel fuzzy derivation structure
on PF set and improved classical fuzzy inference system. In [30,31], Thong et al. utilized the PF
clustering approach for complex and particle clump optimization. Using the PF weighted cross-entropy
concept, Wei [32] studied basic leadership technique and used this technique to rank the alternative.
Based on PF sets, Yang et al. [33] defined flexible soft discernibly matrix of decision-making. In [34],
Garg design aggregation of PF sets for MCDM problems. In [35], Peng et al. proposed picture fuzzy set
approach and tested it in decision-making, also see [36–38] for the PF set. Ashraf et al. [39] extended
the structure of cubic sets to the picture fuzzy sets. They also defined the notion of positive–internal,
neutral–internal, negative–internal and positive–external, and neutral–external and negative-external
cubic picture fuzzy sets.
Herrera & Martinez [40] developed the concept of 2-tuple linguistic processing model based on
the symbolic translation model, and also showed that the 2-tuple linguistic information processing
manner can effectively avoid the loss and distortion of information. A group decision-making model
was proposed by Herrera et al. [41], to manage a nonhomogeneous information. For choosing the
appropriate agile manufacturing system, Wang [42] developed a 2-tuple fuzzy linguistic processing
model. The TOPSIS technique is extended by Wei [43] for MAGDM problem with 2-tuple linguistic
information. Chang & Wen [44] proposed the efficient algorithm for DFMEA, combining the 2-tuple
model and OWA operator. Bonferroni mean operators are extended by Jiang & Wei [45] for the 2-tuple
linguistic information. The dependent interval 2-tuple linguistic aggregation operators are developed
by Li et al. [46] for MAGDM. Wang et al. [47] proposed an algorithm for the MAGDM problems, using
the interval 2-tuple linguistic information and Choquet integral aggregation operators. To study the
application of MAGDM on the supplier selection, Liu [48] defined the 2-tuple linguistic Muirhead
mean operators. A consensus reaching model for the 2-tuple linguistic MAGDM was proposed by
Zhang et al. [49], where the weight information are incomplete.
The remainder of the article is arranged as follows. In Section 2, we briefly discuss the basic
knowledge about the picture fuzzy set and 2-tuple linguistic processing model. In Section 3, we define
2-TPFLS and their operational laws. In Section 4, we present some 2-tuple picture fuzzy linguistic
averaging aggregation operators and discuss their basic properties. In Section 5, we present geometric
aggregation operators of 2-tuple picture fuzzy linguistic number and their properties. Utilizing
the 2TPFLWA and 2TPFLWG operators, we developed a model for multiattribute decision-making
(MADM) problem and discussed the application of the developed approach and their comparison
with existing approaches in Section 6. Finally, we write the conclusion in Section 7.

2. Preliminaries
In this section, we have presented some basic knowledge of the 2-tuple linguistic term set, picture
fuzzy set, and 2-tuple picture fuzzy linguistic set.

Definition 1. Let Ś = (s1 , ..., sτ ) be the collection of linguistic terms set and τ denote the odd cardinality [50,51],
such that sτ , and τ represent the possible value of the linguistic variable and positive integer, respectively.
Mathematics 2019, 7, 1000 3 of 19

Generally, τ is taken as 3, 5,, etc., i.e., when τ = 5, then the linguistic term set Ś is defined as {s1 = Poor, s2 =
Slightly poor, s3 = Fair, s4 = Slightly good , s5 = Good}.
If sκ , st ∈ Ś , then the following characteristic must be satisfied;

(1). The ordered of set as: sκ ≺ st , ⇔ κ ≺ t;

(2). The operator of negation as: Neg (sκ ) = sτ −κ ;
(3). Maximum (sκ , st ) = sκ , iff sκ ≥ st ;
(4). Minimum (sκ , st ) = sκ , iff sκ ≤ st .

Using the concept of symbolic translation, Herrera & Martinez [50,51] introduced the 2-tuple fuzzy
linguistic model. The model is used to denote the linguistic assessment information by means of a 2-tuple (sı , αı ),
and sı and αı denote the linguistic label and symbolic translation, respectively, from the linguistic term set Ś and
α ∈ [−0.5, 0.5].

Definition 2. Let β be the result of an aggregation of the indices of a set of labels assessed in a linguistic term
set Ś [50,51] , for example, the result of a symbolic aggregation operation, β ∈ [1, τ ], where τ denotes the odd
cardinality of Ś. Let ı = round( β) and α = β − ı be two values, such that, ı ∈ [1, τ ] and α ∈ [−0.5, 0.5], then
α is said to be the symbolic translation.

Definition 3. Let Ś = (s1 , ..., sτ ) be the finite linguistic term set and β ∈ [1, τ ] be the number value of the
aggregation result of linguistic symbolic [50,51]. Then, the mapping Λ are utilized to get the 2-tuple linguistic
information equivalent to β, and defined as

Λ : [1, τ ] → Ś × [−0.5, 0.5), (1)



 sı , ı = round( β)
Λ( β) = (2)
 α = β − ı, α ∈ [−0.5, 0.5),


where round (.), sı , α denotes the usual round operation, closest index label to β and the value of the symbolic
translation, respectively.

Definition 4. Let Ś = (s1 , ..., sτ ) be the finite linguistic term set and (sı , αı ) be a 2-tuple [50,51]. Then, there
exists a mapping Λ−1 ,, such that from a 2-tuple (sı , αı ) it returns its equivalent numerical value β ∈ [1, τ ] ⊂ R,
which is
Λ−1 : Ś × [−0.5, 0.5) → [1, τ ], (3)

Λ−1 (sı , α) = ı + α = β, (4)

We observe from Definitions 2 and 3 that the conversion of a linguistic term into a linguistic 2-tuple
consists of adding a value 0 as symbolic translation:

Λ ( s ı ) = ( s ı , 0). (5)

Definition 5. Let R 6= φ, be a set [52]. Then, = is called picture fuzzy set, defined as

= = {h ä= (r ), ë= (r ), ü= (r )| r ∈ Ri} , (6)

where ä= (r ), ë= (r ), ü= (r ) : R → [0, 1] be the positive, neutral, and negative membership grades of each
r ∈ R, correspondingly. ä= (r ), ë= (r ), and ü= (r ) satisfy that 0 ≤ ä= (r ) + ë= (r ) + ü= (r ) ≤ 1 ∀ r ∈ R.
π= (r ) = 1 − ä= (r ) + ë= (r ) + ü= (r ) is the refusal grade of r in R, and a triple-component h ä= , ë= , ü= i is
called the picture fuzzy number, and each picture fuzzy number is in the form of E = h ä= , ë= , ü= i, where ä= , ë= ,
and ü= ∈ [0, 1], having the condition
0 ≤ ä= + ë= + ü= ≤ 1. (7)
Mathematics 2019, 7, 1000 4 of 19

Definition 6. Let =1 = ä=1 (r ), ë=1 (r ), ü=1 (r ) and =2 = ä=2 (r ), ë=2 (r ), ü=2 (r ) [52], which are two
picture fuzzy numbers defined on the of discourse R 6= φ, some operations on picture fuzzy numbers are defined
as follows.

(1). =1 ⊆ =2 if
ä=1 (r ) ≤ ä=2 (r ), ë=1 (r ) ≤ ë=2 (r ) and ü=1 (r ) ≥ ü=2 (r ), ∀r ∈ R,
(2). Union 


(r, max ä=1 (r ), ä=2 (r ) , min ë=1 (r ), ë=2 (r ) ,
=1 ∪ =2 =  ;
 


min ü=1 (r ), ü=2 (r )|r ∈ R
(3). Intersection 


(r, min ä=1 (r ), ä=2 (r ) , min ë=1 (r ), ë=2 (r ) ,
=1 ∩ =2 =  ;
 


max ü=1 (r ), ü=2 (r )|r ∈ R
(4). Compliment
=1c = (r, ü=1 (r ), ë=1 (r ), ä=1 (r )|r ∈ R .


Definition 7. Let =1 = ä=1 , ë=1 , ü=1 and =2 = ä=2 , ë=2 , ü=2 [53], which are two picture fuzzy numbers
defined on the universe of discourse R 6= φ; some operations on picture fuzzy numbers are defined as follows,
with λ ≥ 0.
n o
1. =1 ⊕ =2 = ä2= + ä2=2 − ä2= .ä2=2 , ë=1 .ë=2 , ü=1 .ü=2 ;
n 1 1 o
2. =1 ⊗ =2 = ä=1 .ä=2 , ë= + ë2=2 − ë2= .ë2=2 , , ü2= + ü2=2 − ü2= .ü2=2 ;
2
n 1 1 o 1 1

3. λ · =1 = 1 − (1 − ä2= )λ , (ë=1 )λ , (ü=1 )λ ;

n 1 o
4. =1 = ( ä=1 ) , 1 − (1 − ë2= )λ , 1 − (1 − ü2= )λ ;
λ λ
1 1

3. 2-Tuple Picture Fuzzy Linguistic Sets

In the following, we introduced the concept of 2-tuple picture fuzzy linguistic sets and their basic
operations based on the picture fuzzy set and 2-tuple linguistic information.

Definition 8. A 2-tuple picture fuzzy linguistic set = in R 6= φ is defined as

nD Eo
== s ä(r) , së(r) , sü(r) | r ∈ R , (8)

where s ä(r) , së(r) , sü(r) ∈ Ś, with the condition 3 ≤ ä(r ) + ë(r ) + ü(r ) ≤ τ + 1, ∀ r ∈ R. The numbers
s ä(r) , së(r) and sü(r) denote positive, neutral, and negative membership grades of the number r to linguistic
variable =. The term sχ(r) is said to be the refusal grade of the element r to linguistic variable =, and is defined
as sχ(r) = sτ +1− ä(r)−ë(r)−ü(r) .
D     E
For convenience, we said = e = s ä(r) , α , së(r) , β , sü(r) , γ , a 2-tuple picture fuzzy linguistic
     
number (2TPFLN), where s ä(r) , së(r) , sü(r) ∈ Ś, 3 ≤ Λ−1 s ä(r) , α + Λ−1 së(r) , β + Λ−1 sü(r) , γ ≤
τ + 1, and α, β, γ ∈ [−0.5, 0.5).

Definition 9. Let φ, ϕ, and χ be the results of the aggregation of the indices of a set of labels assessed in a
linguistic term set, Ś; for example, the result of a symbolic aggregation operation, φ, ϕ, χ ∈ [1, τ ], 3 ≤ φ + ϕ +
χ ≤ τ + 1, (where τ be the cardinality of Ś). Assume that ä = round(φ), ë = round( ϕ), ü = round(χ) and
that α = φ − ä, β = ϕ − ë, γ = χ − ü are the six values, such that φ, ϕ, χ ∈ [1, τ ] and α, β, γ ∈ [−0.5, 0.5),
then α, β, and γ are called symbolic translations.
Mathematics 2019, 7, 1000 5 of 19

Definition 10. Let Ś = (s1 , ..., sτ ) be the linguistic set and φ, ϕ, χ ∈ [1, τ ] be the three-number value,
denoting the aggregation result of linguistic symbolic. Then, a mapping Λ is utilized to get the 2-tuple linguistic
information equivalent to φ, ϕ, and χ, and is defined as

Λ : [1, τ ] → Ś × [−0.5, 0.5), (9)


s ä , ä = round(φ)
Λ(φ) = (10)
α = φ − ä, α ∈ [−0.5, 0.5),

së , ë = round( ϕ)
Λ( ϕ) = (11)
β = ϕ − ë, β ∈ [−0.5, 0.5),

sü , ü = round(χ)
Λ(γ) = (12)
γ = χ − ü, γ ∈ [−0.5, 0.5),

where round (.), s ä , së , and sü and α, β, and γ denoted the usual round operation; closest index label to φ, ϕ, and
χ; and the value of the symbolic translation, respectively.

Definition 11. Let Ś = (s1 , ..., sτ ) is the finite linguistic term set and = e = h(s ä , α) , (së , β) , (sü , γ)i, be a
2-tuple picture fuzzy linguistic number (2TPFLN). Then, there exists a mapping Λ−1 , such that from a 2-tuple
picture fuzzy numbers h(s ä , α) , (së , β) , (sü , γ)i and it returns its equivalent numerical value φ, ϕ, χ ∈ [1, τ ] ⊂
R,, which is
Λ−1 : Ś × [−0.5, 0.5) → [1, τ ], (13)

Λ−1 (s ä , α) = ä + α = φ, (14)

Λ−1 (së , β) = ë + β = ϕ, (15)

Λ−1 (sü , γ) = ü + γ = χ. (16)

From Definitions 2 and 3, we observe that the conversion of a linguistic term into a linguistic 2-tuple
consists of adding a value 0 as symbolic translation:
D E D     E
Λ s ä(r) , së(r) , sü(r) = s ä(r) , 0 , së(r) , 0 , sü(r) , 0 . (17)

Definition 12. Let = e = h(s ä , α) , (së , β) , (sü , γ)i, a 2-tuple picture fuzzy linguistic number (2TPFLN). Then,
the score function of 2TPFLN is as follows,

τ + Λ−1 (s ä , α) − Λ−1 (së , β) − Λ−1 (sü , γ)

 
e =Λ
Sc(=) , Λ−1 (Sc(=))
e ∈ [1, τ ] (18)
3

Definition 13. Let =e = h(s ä , α) , (së , β) , (sü , γ)i, a 2-tuple picture fuzzy linguistic number (2TPFLN). Then,
the accuracy function of 2TPFLN is as follows,

Λ−1 (s ä , α) + Λ−1 (së , β) + Λ−1 (sü , γ)

 
e =Λ
Hc(=) , Λ−1 ( Hc(=))
e ∈ [1, τ ] (19)
3

Definition 14. Let = e 1 = h(s ä , α1 ) , (së , β 1 ) , (sü , γ1 )i and =

1 1 1
e 2 = h(s ä , α2 ) , (së , β 2 ) , (sü , γ2 )i be the
2 2 2
2-tuple picture fuzzy linguistic numbers (2TPFLNs). Then, using Definitions 12 and 13, the equating technique
can be described as

(1) If Sc(=
e 1 ) > Sc(=
e 2 ), then =e 1 is grater than =
e 2,
(2) If Sc(=1 ) = Sc(=2 ), then
e e
(3) If Hc(=e 1 ) > Hc(=e 2 ), then =
e 1 is grater than =
e 2,
(4) If Hc(=1 ) = Hc(=2 ), then =1 and =2 have the same information.
e e e e
Mathematics 2019, 7, 1000 6 of 19

Definition 15. Let = e 1 = h(s ä , α1 ) , (së , β 1 ) , (sü , γ1 )i and =

1 1 1
e 2 = h(s ä , α2 ) , (së , β 2 ) , (sü , γ2 )i be the two
2 2 2
2-tuple picture fuzzy linguistic number (2TPFLN). Then,
  −1 
Λ (s ä1 ,α1 ) Λ−1 (s ä2 ,α2 ) Λ−1 (s ä1 ,α1 ) Λ−1 (s ä2 ,α2 )
 
 Λ τ τ + τ − τ  τ ,
 
=
e1 ⊕ =
e2 = 
   −1    −1   ;

Λ (së1 ,β 1 ) Λ−1 (së2 ,β 2 ) Λ (sü2 ,γ1 ) Λ−1 (sü2 ,γ2 )
Λ τ , Λ τ
 
τ τ τ τ
  −1 
Λ (s ä1 ,α1 ) Λ−1 (s ä2 ,α2 )
 
Λ τ τ τ ,
 
 
 
    
−1 −1 −1 −1
 Λ τ Λ (së1 ,α1 ) + Λ (së2 ,α2 ) − Λ (së1 ,α1 ) Λ (së2 ,α2 )
 
=
e1 ⊗ =
e2 =  τ τ τ τ , ;

 
 
   −1  
 Λ (sü1 ,α1 ) Λ−1 (sü2 ,α2 ) Λ−1 (sü1 ,α1 ) Λ−1 (sü2 ,α2 ) 
Λ τ τ + τ − τ τ

  λ !!  λ ! 
Λ−1 (s ä1 ,α1 ) Λ−1 (së1 ,α1 )
 Λ τ 1− 1− τ ,Λ τ τ , 
 
 
λ=
e1 = 

;

 λ !
Λ−1 (sü1 ,α1 )
 
Λ
 
τ τ

  λ !  λ !! 
Λ−1 (s ä1 ,α1 ) Λ−1 (së1 ,α1 )
 Λ τ τ ,Λ τ 1− 1− τ 
 λ  
 
=
e1 = 

.

 λ !!
Λ−1 (sü1 ,α1 )
 
Λ 1− 1−
 
τ τ

4. Averaging Aggregation Operators on 2-Tuple Picture Fuzzy Linguistic Numbers

In this section, we used the information of 2-tuple picture fuzzy linguistic numbers, and developed
some averaging aggregation operators and discussed their properties.

Definition 16. Let = e = h(s ä , α) , (së , β) , (sü , γ)i (  = 1, ..., n) be the set of 2-tuple picture fuzzy linguistic
numbers. Then, the function Ωn → Ω is a 2-tuple picture fuzzy linguistic weighted average (2TPFLWA)
operator, defined as
n  
Θ =
M
2TPFLWAΘ (= e 1 , ..., =
e n) = e , (20)
 =1

n
where Θ = (Θ1 , ..., Θn ) T is the weighting vector of =
e  , and Θ  > 0, ∑ Θ  = 1.
 =1
Mathematics 2019, 7, 1000 7 of 19

Theorem 1. The aggregated value by using the 2TPFLWA operator is also a 2-tuple picture fuzzy linguistic
numbers, such that
n  
2TPFLWAΘ (=
e 1 , ..., = Θ =
e n) L
= e
 =1
  Θ  !! 
n Λ−1 (s ä  ,α  )
 Λ τ 1− ∏ 1− τ , 

  =1 
 (21)
= ,
  Θ  !  −1 Θ  
!
 n Λ−1 (së  ,β  ) n Λ (sü  ,γ  )

Λ τ∏ ,Λ τ∏
 
τ τ
 =1  =1

n
where Θ = (Θ1 , ..., Θn ) T is the weights of =
e  , and Θ  > 0, ∑ Θ  = 1.
 =1

Proof. We used the mathematical induction principle to prove Equation (21).

(1). When n = 2, we get

  Θ1 !! 
Λ−1 (s ä1 ,α1 )
 Λ τ 1− 1− τ , 
 
Θ1 =
 
e1 =  .
   Θ1 !  −1  Θ1 ! 
Λ−1 (së1 ,β 1 ) Λ (sü1 ,γ1 )
 
Λ ,Λ
 
τ τ τ τ

  Θ2 !! 
Λ−1 (s ä2 ,α2 )
 Λ τ 1− 1− τ , 
 
Θ2 =
 
e2 =  .
   Θ2 !  −1  Θ2 
!
Λ−1 (së1 ,β 2 ) Λ (sü2 ,γ2 )
 
Λ ,Λ
 
τ τ τ τ

Then,

2TPFLWAΘ (= e 2 ) = ( Θ1 =
e 1, = e 1 ⊕ Θ2 =
e 2)

   Θ1  Θ2 !! 
Λ−1 (s ä1 ,α1 ) Λ−1 (s ä2 ,α2 )
 Λ τ 1− 1− τ 1− ,  τ
 
 
 
  −1  Θ 1  −1  Θ2 ! 
Λ (së1 ,β 1 ) Λ (së2 ,β 2 )
 
= Λ τ , .
 

 τ τ 
 
 
  −1  Θ 1  −1  Θ2 ! 
Λ (sü1 ,γ1 ) Λ (sü2 ,γ2 )
 
Λ τ
 
τ τ
Mathematics 2019, 7, 1000 8 of 19

(2). Let Equation (21) is true for n = κ; that is,

κ  
Θ =
M
2TPFLWAΘ (=
e 1 , ..., =
eκ) = e
 =1
  Θ  !! 
κ Λ−1 (s ä  ,α  )
 Λ τ 1− ∏ 1− τ , 
  =1 
 
=  
  Θ  !  −1 Θ  
!
Λ−1 (së  ,β  ) Λ (sü  ,γ  )
 κ κ 
Λ τ∏ ,Λ τ∏
 
τ τ
 =1  =1

and we prove Equation (21), for n = κ + 1. Then,

2TPFLWAΘ (=
e 1 , ..., = e 1 ⊕ Θ2 =
e κ +1 ) = Θ 1 = e 2 ⊕ ... ⊕ Θκ =
e κ ⊕ Θ κ +1 =
e κ +1

  Θ  !! 
κ Λ−1 (s ä  ,α  )
 Λ τ 1− ∏ 1− τ , 
  =1 
 
 
  Θ  !  −1 Θ  
!
Λ−1 (së  ,β  ) Λ (sü  ,γ  )
 κ κ 
Λ τ∏ ,Λ τ∏
 
τ τ
 =1  =1

=      ! Θ κ +1   
Λ−1 s äκ +1 ,ακ +1

 Λ  τ 1 − 1− τ
 , 

 
 
⊕ 
 ! Θ κ +1   ! Θ κ +1
      
 
 Λ−1 sëκ +1 ,β κ +1 Λ−1 süκ +1 ,γκ +1 
 Λ τ  , Λ τ  
τ τ

 Θ  !! 
κ +1

Λ−1 (s ä  ,α  )
 Λ τ 1− ∏ 1− τ ,
  =1 
 
=  
 Θ  ! Θ  
!
κ +1

κ +1 Λ (sü ,γ  )
 −1
Λ−1 (së  ,β  )
 
Λ τ ∏ ,Λ τ ∏ 
 
τ τ
 =1  =1

which shows that the aggregated value is also a 2TPFLN. Therefore, Equation (21) holds for all n.

Property 1 (Idempotency). If =
e = =
e for all , then

2TPFLWAΘ (=
e 1 , ..., =
e n) = =
e. (22)
Mathematics 2019, 7, 1000 9 of 19

Proof. As =
e  = =(
e  = 1, ..., n), the WA aggregation result of 2TPFLNs can be calculated using

n  
Θ =
M
2TPFLWAΘ (=
e 1 , ..., =
e n) = e
 =1
  Θ  !! 
n Λ−1 (s ä  ,α  )
 Λ τ 1− ∏ 1− τ , 
  =1 
 
=  ,
  Θ  !  −1 Θ  ! 
 n Λ−1 (së  ,β  ) n Λ (sü  ,γ  )

Λ τ∏ ,Λ τ∏
 
τ τ
 =1  =1

n
 ∑n
      
∑ Θ  Θ
Λ −1 ( s Λ −1 (së  ,β  )

 Λ  τ 1 − 1 − ä ,α ) =1  , Λ τ  =1 , 
 τ τ 
 
 
= 



 ∑n Θ 
   
 −1
Λ ( ü   ) =1 
s ,γ
 
 Λ τ τ


  −1
Λ−1 (s ä ,α) Λ (së ,β)
      
Λ τ 1− 1− τ ,Λ τ τ ,
 
= 



  −1
Λ (sü ,γ)

Λ τ τ

= h(s ä , α) , (së , β) , (sü , γ)i = =

e

Property 2 (Boundedness). Let = e + = max=

e  (  = 1, ..., n) be a set of 2TPFLNs, = e − = min=
e  , and = e ,
 
then
e − ≤ 2TPFLWAΘ (=
= e 1 , ..., = e +.
e n) ≤ = (23)

e − are the minimum 2TPFLNs and =

Proof. As = e− ≤ =
e + are the maximum 2TPFLNs, there exists = e ≤
+
= . Thus, we have
e
n n n
∑ Θ  =e − ≤ ∑ Θ  =e ≤ ∑ Θ  =e + .
 =1  =1  =1

n n
e− = =
Corresponding to the property (1), there are ∑ Θ  = e − and ∑ Θ  =
e+ = =
e + . Therefore,
 =1  =1

e − ≤ 2TPFLWAΘ (=
= e 1 , ..., = e+
e n) ≤ =

Property 3 (Monotonicity). Let =

e  (  = 1, ..., n) and = e / (  = 1, ..., n) be a collection of 2TPFLNs, if =
e ≤
e / , ∀ , then
=
2TPFLWAΘ (= e 1 , ..., =
e n ) ≤ 2TPFLWAΘ (= e / , ..., =
e / ). (24)
1 n
Mathematics 2019, 7, 1000 10 of 19

n n
Proof. As = e / (  = 1, ..., n), there exist ∑ Θ  =
e ≤ = e ≤ ∑ Θ =
e / . Therefore,
 =1  =1

2TPFLWAΘ (=
e 1 , ..., = e / , ..., =
e n ) ≤ 2TPFLWAΘ (= e / ).
1 n

Definition 17. Let =

e = h(s ä , α) , (së , β) , (sü , γ)i (  = 1, ..., n) be the set of 2-tuple picture fuzzy linguistic
numbers. Then, the 2-tuple picture fuzzy linguistic ordered weighted averaging (2TPFLOWA) operator is a
n
function Ωn → Ω, with Θ = (Θ1 , ..., Θn ) T as the associated weights and Θ  > 0, ∑ Θ  = 1. Then,
 =1

n  
2TPFLOWAΘ (=
e 1 , ..., = Θ =
e n) L
= e
σ( )
 =1
     !Θ   
n Λ−1 s äσ( ) ,ασ( )
 Λ  τ 1 − ∏ 1 − τ
 , 
 =1 (25)
 
 
 
=

,

    !Θ      !Θ   
 n Λ−1 sëσ( ) ,β σ( ) n Λ−1 süσ( ) ,γσ( ) 
 Λ τ ∏  , Λ τ ∏  
τ τ
 =1  =1

where the permutation of (1, ..., n) are (σ (1), ..., σ (n), and defined as =
e
σ ( −1) ≥ =σ ( ) ∀  = 2, ..., n.
e

Definition 18. Let =

e = h(s ä , α) , (së , β) , (sü , γ)i (  = 1, ..., n) be the set of 2-tuple picture fuzzy linguistic
numbers. Then, the 2-tuple picture fuzzy linguistic hybrid average (2TPFLHA) operator is a function of
Ωn → Ω, such that
∗
!
n
2TPFLH AΘ (=
e 1 , ..., = Θ =
e n) L
= e
σ( )
 =1
   ∗ ∗
 !Θ   
n Λ−1 s äσ( ) ,ασ( )
 Λ  τ 1 − ∏ 1− τ
 , 
  =1 



 (26)
=  ,
 
∗
 Θ  
Θ
    
∗
 ∗ ∗
Λ −1 s
 !
 n ë σ( )
, β σ (  ) n Λ−1 süσ( ) ,γσ( )
 
 Λ τ ∏    , Λ τ ∏
   
 
τ τ
 =1  =1

n ∗
where Θ = (Θ1 , ..., Θn ) T denote the associated weights, such that Θ  > 0, ∑ Θ  = 1; = e
σ ( ) is the
 =1
∗ ∗
!
th
j biggest element of the 2-tuple picture fuzzy linguistic arguments = = = nω  =  ,  = 1, ..., n ;
σ( ) σ( )
e e e

n
e  , with ω  > 0, ∑ ω  = 1;
(ω = ω1 , ..., ωn ) is the weight vector of 2-tuple picture fuzzy linguistic arguments =
 =1
and n is the balancing coefficient.

5. Geometric Aggregation Operators on 2-Tuple Picture Fuzzy Linguistic Numbers

In this section, we used the information of 2-tuple picture fuzzy linguistic numbers, developed
some geometric aggregation operators, and discussed their properties.
Mathematics 2019, 7, 1000 11 of 19

Definition 19. Let = e = h(s ä , α) , (së , β) , (sü , γ)i (  = 1, ..., n) be the set of 2-tuple picture fuzzy linguistic
numbers. Then, the function of Ωn → Ω are the 2-tuple picture fuzzy linguistic weighted geometric (2TPFLWG)
operator, defined as
n  Θ
e  ,
O
2TPFLWGΘ (= e 1 , ..., =
e n) = = (27)
 =1

n
where Θ = (Θ1 , ..., Θn ) T is the weighting vector of =
e  , such that Θ  > 0, ∑ Θ  = 1.
 =1

Theorem 2. The aggregated value by using 2TPFLWG operator is also a 2-tuple picture fuzzy linguistic
numbers, such that
n  Θ 
2TPFLWGΘ (=
e 1 , ..., = =
e n) N
= e
 =1
 Θ  ! Θ  !! 
n n
 
Λ−1 (s ä  ,α  ) Λ−1 (së  ,β  )
 Λ τ ∏ τ ,Λ τ 1− ∏ 1− τ , 

  =1  =1 
 (28)
=

,

Θ  !!
n

Λ−1 (sü  ,γ  )
 
Λ 1− ∏ 1−
 
τ τ
 =1

n
where Θ = (Θ1 , ..., Θn ) T is the weights of =
e  , such that Θ  > 0, ∑ Θ  = 1.
 =1

Proof. We used the mathematical induction principle to prove Equation (28).

(1). When n = 2, we get
   Θ1 !  Θ1 !! 
Λ−1 (s ä1 ,α1 ) Λ−1 (së1 ,β 1 )
 Λ τ τ ,Λ τ 1− 1− τ , 
  Θ1 



=
e1 =  .
  Θ1 !! 
Λ−1 (sü1 ,γ1 )
 
Λ 1− 1−
 
τ τ

   Θ2 !  Θ2 !! 
Λ−1 (s ä1 ,α2 ) Λ−1 (së2 ,β 2 )
 Λ τ τ ,Λ τ 1− 1− τ , 
  Θ2 



=
e2 =  .
  Θ2 !! 
Λ−1 (sü2 ,γ2 )
 
,Λ 1− 1−
 
τ τ

Then,
  Θ1   Θ2
2TPFLWGΘ (=
e 1, =
e 2) = =
e1 ⊗ =
e2

  Θ 1  Λ −1  Θ2 ! 
Λ−1 (s ä1 ,γ1 ) (sä2 ,γ2 )

 Λ τ τ τ , 
 
 
 
   Θ1  Θ2 !! 
Λ−1 (së1 ,α1 ) Λ−1 (së2 ,α2 )
 
=  Λ τ 1− 1− 1− , .
 
 τ τ 
 
 
   Θ1  Θ2 !! 
Λ−1 (sü1 ,α1 ) Λ−1 (sü2 ,α2 )
 
Λ τ 1− 1− 1−
 
τ τ
Mathematics 2019, 7, 1000 12 of 19

(2). Assume that Equation (28), true for n = κ; that is,

κ 
M Θ 
2TPFLWGΘ (=
e 1 , ..., =
eκ) = =
e
 =1
  Θ  !  Θ  !! 
κ Λ−1 (s ä  ,α  ) κ Λ−1 (së  ,β  )
 Λ τ∏ τ ,Λ τ 1− ∏ 1− τ , 
  =1  =1 
 
=  
  Θ  !! 
Λ−1 (sü  ,γ  )
 κ 
Λ 1− ∏ 1−
 
τ τ
 =1

and we prove Equation (28) for n = κ + 1; then,

  Θ1   Θκ  Θ
2TPFLWGΘ (=
e 1 , ..., =
e κ +1 ) = =
e1 ⊗ ... ⊗ =
eκ ⊗ =e κ +1 κ +1

  Θ  !  Θ  !! 
κ Λ−1 (s ä  ,α  ) κ Λ−1 (së  ,β  )
 Λ τ∏ τ ,Λ τ 1− ∏ 1− τ , 
  =1  =1 
 
 
  Θ  !! 
Λ−1 (sü  ,γ  )
 κ 
Λ 1− ∏ 1−
 
τ τ
 =1

=     ! Θ κ +1      ! Θ κ +1   
Λ−1 s äκ +1 ,ακ +1 Λ−1 sëκ +1 ,β κ +1
 Λ τ  , Λ  τ 1 − 1−  , 
 τ τ 
 
 
⊕ 
 ! Θ κ +1  
   

 
 Λ−1 süκ +1 ,γκ +1 
 Λ  τ 1 − 1− τ
 

 Θ  ! Θ  !! 
κ +1

κ +1

Λ−1 (s ä  ,α  ) Λ−1 (së  ,β  )
 Λ τ ∏ τ ,Λ τ 1− ∏ 1− τ , 
  =1  =1 
 
=  
 Θ  !! 
κ +1

Λ−1 (sü  ,γ  )
 
Λ 1− ∏ 1−
 
τ τ
 =1

which shows that the aggregated value is also a 2TPFLN. Therefore, Equation (28) holds for all n.

The 2TPFLWG operator have the following properties are satisfied.

Property 4 (Idempotency). If =
e = =
e for all , then

2TPFLWGΘ (=
e 1 , ..., =
e n) = =
e. (29)

Property 5 (Boundedness). Let = e + = max=

e  (  = 1, ..., n) be a set of 2TPFLNs, and = e − = min=
e , = e  , then
 

e − ≤ 2TPFLWGΘ (=
= e 1 , ..., = e +.
e n) ≤ = (30)
Mathematics 2019, 7, 1000 13 of 19

Property 6 (Monotonicity). Let =

e  (  = 1, ..., n) and = e / (  = 1, ..., n) be a collection of 2TPFLNs, if =
e ≤
e / , ∀ , then
=
2TPFLWGΘ (= e 1 , ..., =
e n ) ≤ 2TPFLWGΘ (= e / , ..., =
e / ). (31)
1 n

Definition 20. Let =

e = h(s ä , α) , (së , β) , (sü , γ)i (  = 1, ..., n) be the set of 2-tuple picture fuzzy linguistic
numbers. Then, the function of Ωn → Ω are 2-tuple picture fuzzy linguistic ordered weighted geometric
n
(2TPFLOWG) operator, defined as Θ = (Θ1 , ..., Θn ) T , which are the associated weights, and Θ  > 0, ∑ Θ  = 1.
 =1
Then,
n  Θ 
2TPFLOWGΘ (=
e 1 , ..., = =
e n) = N
σ( )
e
 =1
    !Θ       !Θ   
n Λ−1 s ä ,ασ( ) n Λ−1 së ,β σ( )
 Λ τ ∏ σ( )  , Λ  τ 1 − ∏ 1− σ( )  , 
τ τ
 =1  =1
 



 (32)
=
 ,
 !Θ  
  

Λ−1 sü
 
 n ,γσ( ) 
 Λ  τ 1 − ∏ 1− σ( )
τ
 
 =1

The permutation of (1, ..., n) are (σ (1), ..., σ(n), and =

e
σ ( −1) ≥ =σ ( ) ∀  = 2, ..., n.
e

Definition 21. Let = e = h(s ä , α) , (së , β) , (sü , γ)i (  = 1, ..., n) be the set of 2-tuple picture fuzzy linguistic
numbers. Then, the function of Ωn → Ω are 2-tuple picture fuzzy linguistic hybrid geometric (2TPFLHG)
operator, defined as
!Θ 
n ∗
2TPFLHGΘ (=
e 1 , ..., = =
e n) = N
σ( )
e
 =1
   
∗
 Θ   
 !Θ    
∗ ∗ ∗
n Λ−1 s ä ,α σ (  ) n Λ−1 s ë , βσ( )
 Λ τ σ( )
∏
σ( ) ,Λ
 τ 1 − ∏ 1 −

 , 
   
 τ τ

  =1  =1 
 (33)
= ,
 
 !Θ  
   ∗

∗
Λ−1 s ü
 
 n ,γ σ (  ) 
 Λ  τ 1 − ∏ 1− σ( )
τ
 
 =1

n ∗
where Θ = (Θ1 , ..., Θn ) T are the associated weights, Θ  > 0, ∑ Θ  = 1, and = e th
σ ( ) is the  biggest element of
 =1
∗ ∗
!
the 2-tuple picture fuzzy linguistic arguments. = e =e = nω  =
e  ,  = 1, ..., n , where (ω = ω1 , ..., ωn )
σ( ) σ( )

n
e  , with ω  > 0, ∑ ω  = 1, where n
is the weighting vector of 2-tuple picture fuzzy linguistic arguments =
 =1
denotes the balancing coefficient.

6. An Approach for MADM with 2-Tuple Picture Fuzzy Linguistic Information

In this section, based on proposed two operators (2TPFLWA or 2TPFLWG operators), we proposed
an approach for MADM problem, with the information of 2-tuple picture fuzzy linguistic information.
Suppose that the discrete set of alternatives is Q = (Q1 , ..., Qm ), and the attributes set is N = (N1 , ..., Nn ),
n
where Θ = (Θ1 , ..., Θn ) T is the weights of the attribute N  and Θ  ∈ [0, 1], ∑ Θ  = 1. Assume that
 =1


Z = erı m×n = h(s ä , α) , (së , β) , (sü , γ)im×n is the 2-tuple picture fuzzy linguistic decision matrix,
Mathematics 2019, 7, 1000 14 of 19

where e rı , takes the form of the 2-tuple picture fuzzy linguistic numbers, and (s ä , α) , (së , β) , (sü , γ)
denote the positive, neutral, and negative grades, respectively, such that the alternative Qı satisfies the
attribute N  given by the decision-makers. Where (s ä , α) , (së , β) , (sü , γ) ∈ Ś, αı , β ı , γı ≤ [−0.5, 0.5),
ı = 1, ..., m;  = 1, ..., n. Now, we used the 2-tuple picture fuzzy linguistic information and apply the
2TPFLWA or 2TPFLWG operator for the MADM problem.
Step 1: To find the total preference values = e ı (ı = 1, ..., m) of the alternative Q  , we used the given
information of the matrix Z, the 2TPFLWA operator (Equation (21)), and the 2TPFLWG operators
(Equation (28)).
n  
2TPFLWAΘ (=
e 1 , ..., = Θ =
e n) L
= e
 =1
 Θ  !! 
n

Λ−1 (s ä  ,α  )
Λ τ 1− ∏ 1− τ ,
(34)
 
  =1 
 
=

,

Θ  ! Θ  !
n n
  −1
Λ−1 (së  ,β  ) Λ (sü  ,γ  )
 
Λ τ ∏ ,Λ τ ∏
 
τ τ
 =1  =1

Or
n  Θ 
2TPFLWGΘ (=
e 1 , ..., = =
e n) N
= e
 =1
 Θ  ! Θ  !! 
n n
 
Λ−1 (s ä  ,α  ) Λ−1 (së  ,β  )
 Λ τ ∏ τ ,Λ τ 1− ∏ 1− τ , 

  =1  =1 
 (35)
=

,

Θ  !!
n

Λ−1 (sü  ,γ  )
 
Λ 1− ∏ 1−
 
τ τ
 =1

Step 2: Find the scores Sc(=

e ı )(ı = 1, ..., m) values, using Equation (18) of the total 2-tuple picture
fuzzy linguistic numbers =e ı.
Step 3: According to the score value Sc(= e ı )(ı = 1, ..., m), give ranking to the alternatives Qı and
select the best one.

6.1. Practical Example

To demonstrate the application of the developed approach, we present a numerical application
about the robot selection by manufacturing unit. A robot can execute instruction on its own, either as a
machine or as a virtual intelligent agent. In practice, robots are usually electromechanical machines that
are controlled by computers, as well as electronic programming. An industrial robot is an automatically
controlled, reprogrammed, multifunctional works and can be programmed on three or more axes.
It can be installed permanently or mobile for use in industrial application. It is well known that
many industrial robots are present in almost all production or manufacturing industries to improve
quality and productivity. Depending on the work performed, choosing the right robot for a critical
task has become a challenging task for automated production cells. Various criteria and alternatives
are considered to be responsible for the operation of a particular robot. Therefore, to facilitate this
evaluation and selection process a strong multicriteria decision support model is indeed required. To
address this issue, we present a work that explores the concept of 2TPFLSs, integrated with averaging
and geometric operators to help such a decision-making problems.
The given model has been applied for the selection of an industrial robot carried out by
three production unit of a famous manufacturing industry in Pakistan. After initial selection,
three alternatives robots {Q1 , Q2 , Q3 } were selected for further evaluation. The given set of criteria
{N1 , N2 , N3 , N4 } has been considered, where N1 represent the speediness (measure in m/s, which shows
the quickness of the response during the transportation); N2 shows payload capacity (operation
limitation or ranged of the robot’s payload capacity, measured in kg); N3 represent the programming
Mathematics 2019, 7, 1000 15 of 19

flexibility (readability, coordination, and degree of common robotic integration with other robotic
system); N4 shows the human–machine interface (capability of easy human–robot interaction). By the
existing experience and knowledge, the manufacturing unit uses the criteria weighting vector
Θ = (0.27, 0.24, 0.23, 0.26) T . The evaluation matrix is shown in Table 1. In the following, to choose
the best robot for the manufacturing unit, we used the proposed approach. The final ranking of robot,
see Tables 2–4.

Table 1. The 2-tuple picture fuzzy linguistic decision matrix.

N1 N2 N3 N4
Q1 h(s6 , 0), (s1 , 0)(s1 , 0i h(s4 , 0), (s2 , 0)(s1 , 0i h(s4 , 0), (s2 , 0)(s1 , 0i h(s4 , 0), (s2 , 0)(s3 , 0i
Q2 h(s4 , 0), (s2 , 0)(s2 , 0i h(s5 , 0), (s1 , 0)(s2 , 0i h(s3 , 0), (s1 , 0)(s2 , 0i h(s1 , 0), (s3 , 0)(s4 , 0i
Q3 h(s5 , 0), (s1 , 0)(s2 , 0i h(s4 , 0), (s2 , 0)(s1 , 0i h(s1 , 0), (s2 , 0)(s3 , 0i h(s2 , 0), (s2 , 0)(s2 , 0i

To choose the most desirable alternative, we used the following steps.

Table 2. The aggregated value of the alternatives by using the 2-tuple picture fuzzy linguistic weighted
average (2TPFLWA).

2TPFLWA 2TPFLWG
Q1 h(s4 , 0.45), (s2 , −0.35), (s1 , 0.33)i h(s4 , 0.44), (s2 , −0.25), (s2 , −0.42)i
Q2 h(s3 , 0.42), (s2 , −0.41), (s2 , −0.36)i h(s3 , −0.28), (s2 , −0.16), (s3 , −0.41)i
Q3 h(s4 , −0.28), (s2 , −0.39), (s2 , −0.18)i h(s3 , −0.46), (s2 , −0.19), (s2 , 0.03)i

Table 3. Alternatives score values.

2TPFLWA 2TPFLWG
Q1 (s3 , 0.48) (s3 , 0.38)
Q2 (s3 , 0.06) (s2 , 0.42)
Q3 (s3 , 0.09) (s3 , −0.44)

Table 4. Ranking of the alternatives.

Operators Ranking
2TPFLWA Q1 > Q3 > Q2
2TPFLWG Q1 > Q3 > Q2

6.2. A Comparative Analysis with Linguistic Picture Fuzzy Sets

The notion of linguistic picture fuzzy set (LPFS) was developed by Zeng et al. [53]. Zeng et al. [53]
introduced the extended version of the linguistic picture fuzzy TOPSIS method and solved the
problem of enterprise resource planning systems. We solved our created problem, using the concept of
Zeng et al. [53].
First, we convert the data of created problem to linguistic picture fuzzy numbers, which are
shown in Table 5.

Table 5. The linguistic picture fuzzy decision matrix.

N1 N2 N3 N4
Q1 h(s6 , s1 , s1 )i h(s4 , s2 , s1 )i h(s4 , s2 , s1 )i h(s4 , s2 , s3 )i
Q2 h(s4 , s2 , s2 )i h(s5 , s1 , s2 )i h(s3 , s1 , s2 )i h(s1 , s3 , s4 )i
Q3 h(s5 , s1 , s2 )i h(s4 , s2 , s1 )i h(s1 , s2 , s3 )i h(s2 , s2 , s2 )i

Then, we apply all of the steps of the approach of Zeng et al. [53], which, using the weights of the
attributes, are (0.27, 0.24, 0.23, 0.26)T . We obtain the following ranking.
Mathematics 2019, 7, 1000 16 of 19

6.3. Comparative Discussion

To illustrate the effectiveness of the developed algorithm using 2TPFLSs, we give a numerical
example and analyze the selection of the best alternative using the developed approach and 2TPFL
information. In Table 4, we derived the ranking of the alternatives by utilizing the developed
method. From Table 6, we observed that the ranking orders of the alternatives using the developed
method are totally matched with the result derived from the Zeng et al. [53] method. Therefore,
the proposed approach is also validated. Using the proposed method, the best alternative is Q1 ,
and using the Zeng et al. [53] method, the best alternative is aslo Q1 , both are same best alternative
Q1 (see Table 6). Thus, the proposed technique is rather prominent, because it can effectively avoid
any loss of information which formerly occur during the linguistic information processing. Both
numerical and linguistic information are taken into consideration using 2-tuple picture fuzzy linguistic
information, which makes the developed techniques more prominent, flexible, and realistic.

Table 6. Ordering of the alternatives.

Approaches Ranking
LPF TOPSIS Method [53] Q1 > Q3 > Q2
2TPFLWA operator Q1 > Q3 > Q2
2TPFLWG operator Q1 > Q3 > Q2

7. Conclusions
In this article, we study a MADM problem with the 2-tuple picture fuzzy linguistic environment.
First, we introduced some 2-tuple picture fuzzy linguistic aggregation operators: 2-tuple picture
fuzzy linguistic weighted average (2TPFLWA), 2-tuple picture fuzzy linguistic weighted geometric
(2TPFLWG), 2-tuple picture fuzzy linguistic ordered weighted average (2TPFLOWA), 2-tuple picture
fuzzy linguistic ordered weighted geometric (2TPFLOWG), 2-tuple picture fuzzy linguistic hybrid
average (2TPFLHA), and 2tuple picture fuzzy linguistic hybrid geometric (2TPFLHG) operators. We
study some basic properties of the defined operators. Then, we use the developed operators and
write an algorithm for the solution of MADM problem. The practical application of robot selection
by manufacturing unit is given to prove the importance of the proposed method and to establish its
practicability and effectiveness. Finally, we compare our proposed method with the Zeng et al. [53]
method, to show that our proposed method is more validate, practical, and effective than the other
existing methods. In our future work, we will apply the application of 2TPFLNs in many other
researches [22,23,27,30,32].

Author Contributions: Project administration, A.A.K.; Conceptualization, Software, Writing—original draft,

M.Q.; Writing—review and editing, M.Q. and M.B.; Supervision, S.A.; Funding acquisition, J.L.
Funding: This paper is supported by “Research on the Effectiveness Evaluation, Risk Control and System
Construction of the Agricultural Credit Guarantee Policy”, the National Natural Science Foundation of China
(NSFC), Jan 2019-Dec 2022, No. 71873100. Sponsor and Host: Jianchao Luo. This research work is supported
by “Research on the Effect Evaluation, Operational Pattern, Supporting Policy of the Contracted Management of
Farmland Mortgage Finance”, the National Natural Science Foundation of China (NSFC), Jan 2016. Dec 2019, No.
71573210. Sponsor and Host: Jianchao Luo.
Acknowledgments: The authors would like to thank the Editor in Chief, the Associate Editor, and the anonymous
referees for their detailed and valuable comments, which helped improve this manuscript.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Zadeh, L.A. Fuzzy sets. Inf. Control 1965, 8, 338–353. [CrossRef]
2. Bellman, R.E.; Zadeh, L.A. Decision-making in a fuzzy environment. Manag. Sci. 1970, 17, B-141. [CrossRef]
3. Buckley, J.J. Fuzzy decision-making with data: Applications to statistics. Fuzzy Sets Syst. 1985, 16, 139–147.
Mathematics 2019, 7, 1000 17 of 19

[CrossRef]
4. Atnassov, K. Intuitionistic Fuzzy Sets: Theory and Applications. Physica-Verlag: Heidelberg, Germany, 1999.
5. Garg, H. Generalized intuitionistic fuzzy interactive geometric interaction operators using Einstein t-norm
and t-conorm and their application to decision-making. Comput. Ind. Eng. 2016, 101, 53–69. [CrossRef]
6. Garg, H. Novel intuitionistic fuzzy decision-making method based on an improved operation laws and its
application. Eng. Appl. Artif. Intell. 2017, 60, 164–174. [CrossRef]
7. Kaur, G.; Garg, H. Cubic intuitionistic fuzzy aggregation operators. Int. J. Uncertain. Quantif. 2018, 8, 405–427.
[CrossRef]
8. Kaur, G.; Garg, H. Generalized cubic intuitionistic fuzzy aggregation operators using t-norm operations and
their applications to group decision-making process. Arab. J. Sci. Eng. 2018, 1–20. [CrossRef]
9. Garg, H.; Arora, R. Novel scaled prioritized intuitionistic fuzzy soft interaction averaging aggregation
operators and their application to multi criteria decision-making. Eng. Appl. Artif. Intell. 2018, 71C, 100–112.
[CrossRef]
10. Garg, H.; Rani, D. Some generalized complex intuitionistic fuzzy aggregation operators and their application
to multicriteria decision-making process. Arab. J. Sci. Eng. 2018, 1–20. [CrossRef]
11. Shen, K.W.; Wang, J.Q. Z-VIKOR method based on a new weighted comprehensive distance measure of
z-number and its application. IEEE Trans. Fuzzy Syst. 2018. TFUZZ.2018.2816581. [CrossRef]
12. Wang, X.K.; Peng, H.G.; Wang, J.Q. Hesitant linguistic intuitionistic fuzzy sets and their application in
multicriteria decision-making problems. Int. J. Uncertain. Quantif. 2018, 8, 321–341. [CrossRef]
13. Hoskova-Mayerova, S.; Maturo, A. Decision-making process using hyperstructures and fuzzy structures in
social sciences. In Soft Computing Applications for Group Decision-making and Consensus Modeling; Springer:
Cham, Switzerland, 2018; pp. 103–111.
14. Collan, M.; Kacprzyk, J. (Eds.) Soft Computing Applications for Group Decision-making and Consensus Modeling;
Springer International Publishing: Cham, Switzerland, 2018
15. Cuong, B.C. Picture fuzzy sets-first results. In Part 1 “Seminar” Neuro-Fuzzy Systems with Applications; Institute
of Mathematics: Hanoi, Vietnam, 2013
16. Cuong, B.C. Picture fuzzy sets-first results. In Part 2 “Seminar” Neuro-Fuzzy Systems with Applications; Institute
of Mathematics: Hanoi, Vietnam, 2013
17. Cuong, B.C. Picture fuzzy sets. J. Comput. Sci. Cybern. 2014, 30, 409.
18. Cuong, B.C.; Van Hai, P. Some fuzzy logic operators for picture fuzzy sets. In Proceedings of the 2015 Seventh
International Conference on Knowledge and Systems Engineering (KSE), Ho Chi Minh City, Vietnam,
8–10 October 2015; pp. 132–137.
19. Cuong, B.C.; Kreinovitch, V.; Ngan, R.T. A classification of representable t-norm operators for picture fuzzy
sets. In Proceedings of the 2016 Eighth International Conference on Knowledge and Systems Engineering
(KSE), Hanoi, Vietnam, 6–8 October 2016; pp. 19–24.
20. Phong, P.H.; Hieu, D.T.; Ngan, R.T.; Them, P.T. Some compositions of picture fuzzy relations. In Proceedings
of the 7th National Conference on Fundamental and Applied Information Technology Research (FAIR’7),
Thai Nguyen, Vietnam, 19–20 June 2014; pp. 19–20.
21. Wei, G.; Alsaadi, F.E.; Hayat, T.; Alsaedi, A. Projection models for multiple attribute decision-making with
picture fuzzy information. Int. J. Mach. Learn. Cybern. 2018, 9, 713–719. [CrossRef]
22. Wei, G.; Gao, H. The generalized Dice similarity measures for picture fuzzy sets and their applications.
Informatica 2018, 29, 107–124. [CrossRef]
23. Wei, G. Some similarity measures for picture fuzzy sets and their applications. Iran. J. Fuzzy Syst. 2018,
15, 77–89.
24. Singh, P. Correlation coefficients for picture fuzzy sets. J. Intell. Fuzzy Syst. 2015, 28, 591–604.
25. Son, L.H. DPFCM: A novel distributed picture fuzzy clustering method on picture fuzzy sets. Expert Syst.
Appl. 2015, 42, 51–66.. [CrossRef]
26. Thong, P.H. A new approach to multi-variable fuzzy forecasting using picture fuzzy clustering and picture
fuzzy rule interpolation method. In Knowledge and Systems Engineering; Springer: Cham, Switzerland, 2015;
pp. 679–690.
27. Son, L.H. Generalized picture distance measure and applications to picture fuzzy clustering. Appl. Soft
Comput. 2016, 46, 284–295.
Mathematics 2019, 7, 1000 18 of 19

[CrossRef]
28. Son, L.H. Measuring analogousness in picture fuzzy sets: From picture distance measures to picture
association measures. Fuzzy Optim. Decis. Mak. 2017, 16, 359–378. [CrossRef]
29. Van Viet, P.; Van Hai, P. Picture inference system: A new fuzzy inference system on picture fuzzy set.
Appl. Intell. 2017, 46, 652–669.
30. Thong, P.H. Picture fuzzy clustering for complex data. Eng. Appl. Artif. Intell. 2016, 56, 121–130. [CrossRef]
31. Thong, P.H. A novel automatic picture fuzzy clustering method based on particle swarm optimization and
picture composite cardinality. Knowl.-Based Syst. 2016, 109, 48–60. [CrossRef]
32. Wei, G. Picture fuzzy cross-entropy for multiple attribute decision-making problems. J. Bus. Econom. Manag.
2016, 17, 491–502. [CrossRef]
33. Yang, Y.; Liang, C.; Ji, S.; Liu, T. Adjustable soft discernibility matrix based on picture fuzzy soft sets and its
applications in decision-making. J. Intell. Fuzzy Syst. 2015, 29, 1711–1722. [CrossRef]
34. Garg, H. Some picture fuzzy aggregation operators and their applications to multicriteria decision-making.
Arab. J. Sci. Eng. 2017, 42, 5275–5290. [CrossRef]
35. Peng, X.; Dai, J. Algorithm for picture fuzzy multiple attribute decision-making based on new distance
measure. Int. J. Uncertain. Quantif. 2017, 7, 177–187. [CrossRef]
36. Phuong, P.T.M.; Thong, P.H. Theoretical analysis of picture fuzzy clustering: Convergence and property.
J. Comput. Sci. Cybern. 2018, 34, 17–32. [CrossRef]
37. Thong, P.H.; Fujita, H. Interpolative picture fuzzy rules: A novel forecast method for weather nowcasting.
In Proceedings of the 2016 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE), Vancouver, BC,
Canada, 24–29 July 2016; pp. 86–93.
38. Van Viet, P.; Chau, H.T.M.; Van Hai, P. Some extensions of membership graphs for picture inference systems.
In Proceedings of the 2015 Seventh International Conference on Knowledge and Systems Engineering (KSE),
Ho Chi Minh City, Vietnam, 8–10 October 2015; pp. 192–197.
39. Ashraf, S.; Abdullah, S.; Qadir, A. Novel concept of cubic picture fuzzy sets. J. New Theory 2018, 24, 59–72.
40. Herrera, F.; Martinez, L. The 2-tuple linguistic computational model: Advantages of its linguistic description,
accuracy and consistency. Int. J. Uncertain. Fuzziness Knowl.-Based Syst. 2001, 9 (Suppl. 1), 33–48. [CrossRef]
41. Herrera, F.; Martınez, L.; Sánchez, P.J. Managing non-homogeneous information in group decision-making.
Eur. J. Oper. Res. 2005, 166, 115–132. [CrossRef]
42. Wang, W.P. Evaluating new product development performance by fuzzy linguistic computing. Expert Syst.
Appl. 2009, 36, 9759–9766. [CrossRef]
43. Wei, G.W. Extension of TOPSIS method for 2-tuple linguistic multiple attribute group decision-making with
incomplete weight information. Knowl. Inf. Syst. 2010, 25, 623–634. [CrossRef]
44. Chang, K.H.; Wen, T.C. A novel efficient approach for DFMEA combining 2-tuple and the OWA operator.
Expert Syst. Appl. 2010, 37, 2362–2370. [CrossRef]
45. Jiang, X.P.; Wei, G.W. Some Bonferroni mean operators with 2-tuple linguistic information and their
application to multiple attribute decision-making. J. Intell. Fuzzy Syst. 2014, 27, 2153–2162.
46. Li, C.C.; Dong, Y. Multi-attribute group decision making methods with proportional 2-tuple linguistic
assessments and weights. Int. J. Comput. Intell. Syst. 2014, 7, 758–770. [CrossRef]
47. Wang, J.Q.; Wang, D.D.; Zhang, H.Y.; Chen, X.H. Multi-criteria group decision-making method based on
interval 2-tuple linguistic information and Choquet integral aggregation operators. Soft Comput. 2015,
19, 389–405. [CrossRef]
48. Qin, J.; Liu, X. 2-tuple linguistic Muirhead mean operators for multiple attribute group decision-making and
its application to supplier selection. Kybernetes 2016, 45, 2–29. [CrossRef]
49. Zhang, W.; Xu, Y.; Wang, H. A consensus reaching model for 2-tuple linguistic multiple attribute group
decision-making with incomplete weight information. Int. J. Syst. Sci. 2016, 47, 389–405. [CrossRef]
50. Herrera, F.; Martínez, L. A 2-tuple fuzzy linguistic representation model for computing with words.
IEEE Trans. Fuzzy Syst. 2000, 8, 746–752.
51. Herrera, F.; Martinez, L. An approach for combining linguistic and numerical information based on the
2-tuple fuzzy linguistic representation model in decision-making. Int. J. Uncertain. Fuzziness Knowl.-Based
Syst. 2000, 8, 539–562.
Mathematics 2019, 7, 1000 19 of 19

[CrossRef]
52. Cuong, B.C.; Kreinovich, V. December. Picture Fuzzy Sets-a new concept for computational intelligence
problems. In Proceedings of the 2013 Third World Congress on Information and Communication
Technologies (WICT 2013), Hanoi, Vietnam, 15–18 December 2013; pp. 1–6.
53. Zeng, S.; Qiyas, M.; Arif, M.; Mahmood, T. Extended Version of Linguistic Picture Fuzzy TOPSIS Method and
Its Applications in Enterprise Resource Planning Systems. Math. Probl. Eng. 2019, 2019, 8594938. [CrossRef]

c 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license ([Link]