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

net/publication/374229814

Relative Fuzzy Linear Programming

Article · September 2023

DOI: 10.14445/22315373/IJMTT-V69I9P503

CITATIONS READ
0 1

2 authors, including:

Owin Olowu
University of Benin
25 PUBLICATIONS 12 CITATIONS

SEE PROFILE

All content following this page was uploaded by Owin Olowu on 27 September 2023.

The user has requested enhancement of the downloaded file.

International Journal of Mathematics Trends and Technology Volume 69 Issue 9, 13-22, September 2023
ISSN: 2231-5373/ https://doi.org/10.14445/22315373/IJMTT-V69I9P503 © 2023 Seventh Sense Research Group®

Original Article

𝕋-Relative Fuzzy Linear Programming

Kelly Osawaru1, Owin Olowu2
1,2
Department of Mathematics, University of Benin, Nigeria.
2Corresponding Author : oghenewaire.olowu@uniben.edu

Received: 13 July 2023 Revised: 19 August 2023 Accepted: 04 September 2023 Published: 27 September 2023

Abstract - We introduce and develop in this article linear programming in the context of the 𝕋-Relative fuzzy sets that were
introduced by Osawaru, Olaleru and Olaoluwa [3]. Here, the objective function and (or) its set of constraints expressed as a
function of a parameter (or variable), say time, is considered. By relaxing the pretenses of optimization using a subjective
gradation relative to the parameter, we model and obtain optimal solutions by employing the tools of the relative fuzzy
membership functions. Thus, fuzzy optimal values obtained are expressed relative to the parameter of control defining the
dynamics of the fuzziness. The results of this study generalize the results obtained for fuzzy linear programming in literature.

Keywords - 𝕋-Relative fuzzy sets, 𝕋-Relative fuzzy mapping, 𝕋-Relative fuzzy linear programming, Fuzzy sets.

1. Introduction
Complexities due to uncertainty in the form of ambiguity, uncertainty related to errors, vagueness, sparsity of data,
subjectivity of expert judgements, and chance or incomplete knowledge characterize most real-world problems. Thus, fuzzy
methods are best suited to define, express and solve such problems. So, formulating linear programming problems using fuzzy
membership maps as well as employing fuzzy concept tools to obtain optimal solutions, has yielded better and richer results.
This approach provides better results as the manager’s knowledge is considered a fuzzy one.

Fuzzy set was introduced by L. A. Zadeh to represent these complexities; several authors have extended and applied it in
different fields. Our interest in this study is the Fuzzy Linear Programming (FLP). The concept of FLP as a traditional Linear
Programming (LP) in a fuzzy environment was introduced in 1978 by Zimmermann [28] following “Decision Making in Fuzzy
Environment” and “On Fuzzy Mathematical Programming” proposed by Bellman R. E., Zadeh L. A. [1] and Tanaka H. et al.
respectively. Gasimov and Yenilmez [24] studied mathematical programming with a single objective expressed by fuzzy
parameters. Since the introduction of the concept, FLP has been constructed in a number of directions with several positive
applications, including the proposed parametric programming approach for solving FLP. The authors of [29] formulated an FLP
problem for a fuzzy system endowed with a fuzzy Boolean coverage, and recently, the authors of [31] simplified with certain
substitutions on nonlinear terms the formulated FLP of [29] to a 0 − 1 integer program. The model was used to determine
optimality conditions for a set covering problem in a fuzzy environment.

Recently, [3] introduced a more profound generalization of the fuzzy set. It is called 𝕋-Relative fuzzy set (TRFS). The
authors gave its characterization and showed that they extended the concept of the fuzzy set by taking into account the variability
of membership grades of elements of the set. The applications of the TRFS sets were also buttressed by [3]. [23] defined Relative
fuzzy maps in the Heilpern. [2] sense and studied fixed point results for Relative Fuzzy Maps. This further enriched the fuzzy
map definition with the capacity to represent and express dynamic fuzzy maps in this sense. Thus, many fuzzy maps results were
deductions from the relative fuzzy maps results.

This article focused on the introduction of 𝕋-Relative Fuzzy Linear Programming (RFLP) as a Fuzzy Linear Programming
(FLP) whose objective function and (or) constraints are defined as 𝕋-Relative fuzzy membership functions. It represents the
fuzzy version of an LP problem whose constraints and or objective functions are expressed 𝕋-Relative to a variable. As a
generalization of the fuzzy point, the concept of a 𝕋-Relative fuzzy point and relative fixed fuzzy point was defined by [23],
which is suitable for an RFLP and is useful in the sequel. Their results also provide sufficient theoretical background for this
study.

This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

 Kelly Osawaru & Owin Olowu / IJMTT, 69(9), 13-22, 2023

2. Preliminaries
The following definitions and results from [3,23] will suffice in this article.
Definition 2.1. A 𝕋-Relative Fuzzy Set (TRFS) 𝑅 of 𝑋 𝕋 is a set with the membership growth function 𝜇𝑅 : 𝑋 × 𝕋 → [0,1]
such that
0 if 𝑥 ∉ 𝑅
𝜇𝑅 (𝑥, 𝜎(𝑡)) = {
𝑘 ∈ (0,1] if 𝑥 ∈ 𝑅 and 𝑡 ∈ 𝕋

∀𝑥 ∈ 𝑋 and 𝑡 ∈ 𝕋, where 𝕋 is any time scale, 𝑅 ⊂ 𝑋 and 𝜎: 𝕋 → 𝕋 a forward difference operator.

𝑛
Definition 2.2. Let {𝑅𝑖 }𝑛𝑖=1 be TRFSs of 𝑋, (𝑋, any nonempty set) with membership functions {𝜇𝑅𝑖 } respectively. Then
𝑖=1
(i) the union of {𝑅𝑖 }𝑛𝑖=1 over 𝕋 (respectively for any 𝑡 ∈ 𝕋 ) is the TRFS membership function defined as

𝜇(∪𝑛𝑖=1𝑅𝑖) (𝑥, 𝑡)𝕋 = max {𝜇𝑅𝑖 (𝑥, 𝑡)}

𝑥∈𝑋,𝑡∈𝕋
(respectively 𝜇(∪𝑛𝑖=1𝑅𝑖) (𝑥, 𝑡)𝑡 = max {𝜇𝑅𝑖 (𝑥, 𝑡)} for any 𝑡 ∈ 𝕋 )
𝑥∈𝑋

(ii) the intersection of {𝑅𝑖 }𝑛𝑖=1 over 𝕋 (respectively for any 𝑡 ∈ 𝕋 ) is the TRFS with 𝕋- membership function defined as

𝜇(∩𝑛𝑖=1𝑅𝑖) (𝑥, 𝑡)𝕋 = min {𝜇𝑅𝑖 (𝑥, 𝑡)}

𝑥∈𝑋,𝑡∈𝕋
(respectively 𝜇(∩𝑛𝑖=1𝑅𝑖) (𝑥, 𝑡)𝑡 =min {𝜇𝑅𝑖 (𝑥, 𝑡)} for any 𝑡 ∈ 𝕋 )
𝑥∈𝑋

(iii) 𝑅𝑖 ⊂ 𝑅𝑗 over 𝕋 (respectively for any 𝑡 ∈ 𝕋 ) if

𝜇𝑅𝑖 (𝑥, 𝑡)𝕋 ≤ 𝜇𝑅𝑗 (𝑥, 𝑡)𝕋 ∀𝑥 ∈ 𝑈 and ∀𝑡 ∈ 𝕋

(respectively 𝜇𝑅𝑖 (𝑥, 𝑡)𝑡 ≤ 𝜇𝑅𝑗 (𝑥, 𝑡)𝑡 ∀𝑥 ∈ 𝑈 and for any 𝑡 ∈ 𝕋 ) for any 𝑖, 𝑗 with 𝑖 ≠ 𝑗.

Definition 2.3. For a metric space (𝑋, 𝑑), from [23] the level sets of a TRFS with 𝛼 ∈ (0,1] is defined by

𝑅(𝛼)𝕋 = {𝑥 ∈ 𝑋: 𝜇𝑅 (𝑥, 𝑡) ≥ 𝛼, for all 𝑡 ∈ 𝕋}

(respectively 𝑅(𝛼)𝑡 = {𝑥 ∈ 𝑋: 𝜇𝑅 (𝑥, 𝑡) ≥ 𝛼, for any 𝑡 ∈ 𝕋})

𝜙 − 𝕋-Relative fuzzy point for all members of 𝕋 denoted 𝑥𝜙𝕋 is called a 𝜙 − 𝕋-Relative fuzzy fixed point of 𝑇𝕋 if 𝑥𝜙𝕋 ⊂ 𝑇𝕋 𝑥.
That is 𝑥 ∈ 𝑇𝕋 𝑥(𝜙) for all 𝑡 ∈ 𝕋, where 𝜙 is a function on 𝑋 × 𝕋 into 0,1].

Definition 2.4. For a metric space (𝑋, 𝑑), let 𝜋(𝑋)𝕋 be a family of approximate quantities and 𝛼 ∈ (0,1]. Then the 𝕋-Relative
fuzzy map 𝑇𝕋 : 𝑋 → 𝜋(𝑋)𝕋 for all members of 𝕋 is said to be a 𝛼 − 𝕋-Relative fuzzy contraction map if

𝐷𝛼 (𝑇𝕋 𝑥, 𝑇𝕋 𝑦)𝕋 ≤ 𝑎𝑑(𝑥, 𝑦)

for all 𝑥, 𝑦 ∈ 𝑋, and 𝑡 ∈ 𝕋.

Definition 2.5. For a metric space (𝑋, 𝑑), let 𝜋(𝑋)𝕋 be a family of approximate quantities and 𝛼 ∈ (0,1]. Then for 𝑇𝕋 : 𝑋 →
𝜋(𝑋)𝕋 for all members of 𝕋,

(i) 𝐹𝑇 (𝛼) = {𝑥 ∗ ∈ 𝑋: 𝑥 ∗ ∈ 𝑇𝕋 𝑥 ∗ (𝛼)} is the set of 𝛼 − 𝕋-Relative fuzzy fixed points of 𝑇𝕋 for all members of 𝕋.
(ii) for any 𝑥0 ∈ 𝑋, the sequence {𝑥𝑛 }∞ 𝑛=0 defined by 𝑥𝑛+1 ∈ 𝑇𝕋 𝑥𝑛 (𝛼) where 𝑛 ≥ 0, is called the 𝛼 − 𝕋-Relative fuzzy Picard
iterative scheme iff for each 𝑥 ∈ 𝑋 and any 𝑦 ∈ 𝑇𝕋 𝑥 the sequence {𝑥𝑛 }∞ 𝑛=0 converges to a fixed point of 𝑇𝕋 with 𝑥0 = 𝑥 and
𝑥1 = 𝑦.

Theorem 2.1. The 𝜙 - 𝕋-Relative fuzzy contraction function, where 𝑋 is a complete metric space. Then
(i) there exists 𝑥 ∗ ∈ 𝑋 (called the 𝜙 − 𝕋-Relative fuzzy fixed point) i.e., 𝑥 ∗ ∈ 𝑇𝕋 𝑥 ∗ (𝜙).

14
 Kelly Osawaru & Owin Olowu / IJMTT, 69(9), 13-22, 2023

(ii) 𝑥 ∗ 𝑖𝑠 a unique if 𝑑(𝑥 ∗ , 𝑦 ∗ ) ≤ 𝑝𝜙 (𝑇𝕋 𝑥 ∗ , 𝑇𝕋 𝑦 ∗ )𝕋 for any 𝑦 ∗ ∈ 𝐹𝑇 (𝜙).

(ii) the scheme, 𝑥𝑛+1 ∈ 𝑇𝕋 𝑥𝑛 (𝜙) converges to 𝑥𝜙∗ 𝕋 of 𝑇𝕋 strongly.

3. Main Result
In optimization theory, problems are modelled in the form of optimizing objective(s) at the instance of some constraints.
Obtaining a feasible solution for the modeled problem is usually challenging; hence, the task of a manager making a decision is
usually complex, especially in multi-criteria problems. In operations research, linear programming techniques are robust tools
used to solve such problems. In real-life problems, it is usually difficult to determine precise targets for set goals. Fuzzy
membership functions are introduced to deal with such difficult situation. The goal is to relax the pretenses of optimization by
means of a subjective gradation, which can be modelled into fuzzy membership functions. 𝜇𝑖 . So, that if 𝐹 =∩ 𝜇𝑖 the objective
will be to search for 𝑥 such that max𝐹 = 𝐹(𝑥). If max𝐹 = 1, then there exists 𝑥 such that 𝐹(𝑥) = 1, but if max𝐹 = 𝛼 ∈ (0,1),
the solution of the multiobjective optimization is a fuzzy point. 𝑥𝛼 and 𝐹(𝑥) = 𝛼. In a more general sense than the one given by
Heilpern [2], a mapping 𝐹: 𝑋 → 𝐼 𝑋 is a fuzzy function over 𝑋 and (𝐹(𝑥))(𝑥) is the fixed degree of 𝑥 for 𝐹. Now if an objective
function and or its set of constraints is expressed as a function of a parameter, say time, then relaxing the pretenses of optimization
by means of a subjective gradation, which can be modelled into fuzzy membership functions is not possible. The way out would
be to resolve it using TRS tools. The stages are as summarized below.

4. 𝕋-Relative Fuzzy Optimization

Here we propose in the sense of [28] and [30] 𝕋-Relative Fuzzy Linear Programming (RFLP) problem.

(i) The set of objective functions that maximizes (or minimizes) a problem may change with respect to some factors. Thus, such
problems are presented as optimizing some goal function 𝕋-Relative to the factors (for example time) given certain constraints.
Such set of objective functions can be modeled into 𝕋-Relative fuzzy membership function 𝜇𝕋 . It is achieved by relaxing the
pretenses of optimization by means of a subjective gradation with respect to a time scale value.

(ii)Examine the following standard form of the symmetric time scale dependent LP problem:
max 𝑧 = 𝑡𝑐𝑥
s.t. 𝑡𝐴𝑥 ≲ b
𝑥 ≥0

where, 𝐴 ∈ ℝ𝑚×𝑛 , b ∈ ℝ𝑚 , 𝑐 𝑇 ∈ ℝ𝑛 , 𝑥 ∈ ℝ𝑛 , 𝑡 ∈ 𝕋 and ≲ and ≳ are the fuzzy versions of ≤ and ≥ respectively.

To represent the 𝕋-Relative fuzzy goal, we impose that the objective function 𝑐𝑡𝑥 with respect to 𝑡 be essentially greater
than or equal to an aspiration level 𝑏0 , chosen by the manager. The following problems will be adequate:

find 𝑥
s.t.tcx ≳ 𝑏0
𝑡𝐴𝑥 ≲b
𝑥 ≥ 0.

(iii) For treating fuzzy inequalities, we propose a 𝕋-Relative linear membership function as follows:

1 if 𝑡𝑐𝑥 > 𝑏0
𝑏0 − 𝑡𝑐𝑥
𝜇𝐴𝑜 (𝑥, 𝑡) = 1 − if 𝑏0 − 𝑝0 ≤ 𝑡𝑐𝑥 ≤ 𝑏0
𝑝0
{ 0 if 𝑡𝑐𝑥 < 𝑏0 − 𝑝0
1 if (𝑡𝐴𝑥𝑗 ) < 𝑏𝑗
𝑗 (𝑡𝐴𝑥𝑗 ) − 𝑏𝑗
𝜇𝐴 (𝑥, 𝑡) = 1 − if 𝑏𝑗 (𝑡𝐴𝑥𝑗 ) < 𝑏𝑗 + 𝑝𝑗 𝑗 = 1,2, … , 𝑚
𝑝𝑗
{ 0 if (𝑡𝐴𝑥𝑗 ) > 𝑏𝑗 − 𝑝𝑗

15
 Kelly Osawaru & Owin Olowu / IJMTT, 69(9), 13-22, 2023

where for 𝑗 = 1, … , 𝑚(𝑡𝐴𝑥𝑗 ) is the 𝑗𝑡ℎ row of 𝐴𝑡𝑥, 𝑏𝑗 is the 𝑗𝑡ℎ element of b and for 𝑗 = 0,1, … , 𝑚, 𝑝𝑗 is a subjectively chosen
constant by the manager expressing the limit of the admissible violation of the 𝑗𝑡ℎ inequality.

(iv) Suppose {𝜇𝕋 }𝑚

𝑗 is the 𝑚 set of 𝕋-Relative fuzzy membership functions maximizing the set of objective functions 𝕋-Relative
to the 𝕋. Using the “min” operator of Bellman and Zadeh (1970) together with the above 𝕋-Relative linear membership functions
then

(a) the optimal region (solution space) for any fixed 𝑡 ∈ 𝕋 is given by the 𝕋-Relative fuzzy membership function

𝜇𝐷 (𝑥, 𝑡) = min {𝜇 𝑗 (𝑥, 𝑡)} or 𝐹𝑡 = min{{𝜇𝑡 }𝑚

𝑗 }.
𝑗=1,…,𝑚

If max𝐹𝑡 = 1, then there exists 𝑥, the optimal solution of the multi objective optimization with respect to 𝑡 is a fuzzy point 𝑥1𝑡
such that 𝐹(𝑥, 𝑡)𝑡 = 1 for a fixed 𝑡 ∈ 𝕋. More generally if max𝐹𝑡 = 𝛼𝑡 , 𝛼𝑡 ∈ (0,1] then the optimal solution of the multi objective
optimization with respect to 𝑡 is a fuzzy point 𝑥𝛼𝑡 such that 𝐹𝑡 (𝑥, 𝑡) = 𝛼𝑡 i.e. there exists 𝑥 such that 𝐹(𝑥, 𝑡)𝑡 = 𝛼 for a fixed
𝑡 ∈ 𝕋.

(b) the optimal region (solution space) for all 𝑡 ∈ 𝕋 is given by the 𝕋-Relative fuzzy membership function
𝜇𝐷 (𝑥, 𝑡) = min { min {𝜇 𝑗 (𝑥, 𝑡)}} or 𝐹𝕋 = min {min{{𝜇𝑡 }𝑚 𝑚
𝑗 }} = min{{𝜇𝕋 }𝑗 }.
𝑡∈𝕋 𝑗=1,…,𝑚 𝑡∈𝕋

If max𝐹𝕋 = 1, then there exists 𝑥 the optimal solution of the multi objective optimization with respect to 𝕋 is a fuzzy point 𝑥1𝕋
such that 𝐹(𝑥, 𝑡)𝕋 = 1 for all 𝑡 ∈ 𝕋. More generally, if max𝐹𝕋 = 𝛼𝕋 , 𝛼𝕋 ∈ (0,1] then the optimal solution of the multi-objective
optimization with respect to 𝕋 is a fuzzy point 𝑥𝛼𝕋 such that 𝐹𝕋 (𝑥, 𝑡) = 𝛼𝕋 i.e there exists 𝑥 such that 𝐹(𝑥, 𝑡)𝕋 = 𝛼 for all 𝑡 ∈ 𝕋.

(v) Applying Theorem 2.1 of [26], we can determine the existence of an 𝛼 relative fixed point. Thus, we have to determine if

(a) 𝐷𝛼 ({𝑥 ∈ 𝑋: 𝜇𝐷 (𝑥, 𝑡) = 𝛼, for any 𝑡 ∈ 𝕋}, {𝑦 ∈ 𝑋: 𝜇𝐷 (𝑦, 𝑡) = 𝛼, for any 𝑡 ∈ 𝕋})𝑡 ≤ 𝑎𝑑(𝑥, 𝑦), for all 𝑥, 𝑦 ∈ 𝑋

(b) 𝐷𝛼 ({𝑥 ∈ 𝑋: 𝜇𝐷 (𝑥, 𝑡) = 𝛼, for all 𝑡 ∈ 𝕋}, {𝑦 ∈ 𝑋: 𝜇𝐷 (𝑦, 𝑡) = 𝛼, for all 𝑡 ∈ 𝕋})𝕋 ≤ 𝑎𝑑(𝑥, 𝑦), for all 𝑥, 𝑦 ∈ 𝑋

(vi) The 𝛼 relative fixed point if it exists coincide with the 𝛼-optimal solution of the multi-objective optimization problem. So

(a) If max𝐹𝑡 = 1, then there exists 𝑥, the optimal solution of the multi objective optimization with respect to 𝑡 is a fuzzy point
𝑥1𝑡 such that 𝐹(𝑥, 𝑡)𝑡 = 1 for a fixed 𝑡 ∈ 𝕋. More generally if max𝐹𝑡 = 𝛼𝑡 , 𝛼𝑡 ∈ (0,1] then the optimal solution of the multi
objective optimization with respect to 𝑡 is a fuzzy point 𝑥𝛼𝑡 such that 𝐹𝑡 (𝑥, 𝑡) = 𝛼𝑡 i.e. there exists 𝑥 such that 𝐹(𝑥, 𝑡)𝑡 = 𝛼 for
a fixed 𝑡 ∈ 𝕋.

(b) If max𝐹𝕋 = 1, then there exists 𝑥 the optimal solution of the multi objective optimization with respect to 𝕋 is a fuzzy point
𝑥1𝕋 such that 𝐹(𝑥, 𝑡)𝕋 = 1 for all 𝑡 ∈ 𝕋. More generally, if max𝐹𝕋 = 𝛼𝕋 , 𝛼𝕋 ∈ (0,1] then the optimal solution of the multi
objective optimization with respect to 𝕋 is a fuzzy point 𝑥𝛼𝕋 such that 𝐹𝕋 (𝑥, 𝑡) = 𝛼𝕋 i.e. there exists 𝑥 such that 𝐹(𝑥, 𝑡)𝕋 = 𝛼
for all 𝑡 ∈ 𝕋. (vii) Alternatively, the RFLP problem can be transformed into an equivalent LP problem by introducing an auxiliary
variable 𝜆 as follows:
max 𝜆
𝑗
s.t. 𝜆 ≤ 𝑡𝜇𝐴 (𝑥, 𝑡)
𝑥 ≥ 0, 𝜆 ∈ [0,1]
which gives
max 𝜆
s.t. 𝑐𝑡𝑥 ≥ 𝑏0 − (1 − 𝜆)𝑝0
( Atx )𝑗 ≥ 𝑏0 + (1 − 𝜆)𝑝𝑗
𝑥 ≥ 0, 𝜆 ∈ [0,1]

Thus, we have the equivalent LP problem for each 𝑡 ∈ 𝕋. So the optimal solution for 3.1(𝑖𝑣) is also optimal for 3.1(𝑣)

16
 Kelly Osawaru & Owin Olowu / IJMTT, 69(9), 13-22, 2023

Example 4.5.2. Suppose a firm’s problem is to design an attractive dividend structure depending on the volume of shares units
held by its shareholders limited by the government regulation of what a modest dividend should be. If the interval of dividend
𝑥 and share units 𝑢 are 𝑋 = [0,5.8] and 𝕌 = [0.1,2.0] respectively. The fuzzy set of the objective function attractive dividend
𝕋-Relative to units of shares could, for instance, be defined by:

1 𝑥 ≥ 5, 𝑢 ≥ 1.5
0.9 𝑥 ≥ 5,0.1 ≤ 𝑢 < 1.5
𝑢 + 0.9
𝜇𝑋𝑜 (𝑥, 𝑢) = − [ (29𝑥 3 − 366𝑥 2 − 877𝑥 + 540)] 0.5 < 𝑥 < 5,0.1 ≤ 𝑢 ≤ 2
2100
0.05 𝑥 ≤ 0.5,0.2 < 𝑢 ≤ 2
{ 0 𝑥 ≤ 0.5, 𝑢 ≤ 0.2 }

The fuzzy set (constraint) “modest dividend” could be represented by

1 𝑥≥5
1
𝜇𝑋𝑐 (𝑥) = { [−29𝑥 3 − 243𝑥 2 + 16𝑥 + 2388] 0.5 < 𝑥 < 4.9
2100
0 𝑥 ≤ 0.5

The 𝕋-Relative fuzzy set “decision” at any 𝑢 is then characterized by its membership function 𝜇𝑋𝐷 (𝑥, 𝑢) = min{𝜇𝑋𝑜 (𝑥, 𝑢), 𝜇𝑋𝑐 (𝑥)}
𝑢+0.9
i.e. the fixed fuzzy point is 𝑥𝛼𝑢 such that max𝜇𝑋𝐷 (𝑥, 𝑢) = 𝛼 for a fixed 𝑢 For 𝑢 = 1.5, 𝜇𝑋𝐷 (𝑥, 𝑢) = min {− [ (29𝑥 3 −
2100
366𝑥 2 − 877𝑥 + 540)] , 0.05, 𝜇𝑋𝑐 (𝑥)} and the 𝕋-Relative fuzzy set “decision” for all 𝑢 is then characterized by its membership
function
𝜇𝑋𝐷 (𝑥) = min {min{𝜇𝑋𝑜 (𝑥, 𝑢), 𝜇𝑋𝑐 (𝑥)}}
𝑢∈𝕌

for all 𝑢 ∈ 𝕌 i.e. the fixed fuzzy point is 𝑥𝛼𝕌 such that max𝜇𝑋𝐷 (𝑥) = 𝛼 for all 𝑢 ∈ 𝕌

Fig. 1 The 𝕋-Relative fuzzy set “decision”

The graph in the figure below shows the optimal region. Thus max𝜇𝑋𝐷 (1.7) ≈ 0.7. The next example considers the
application of the fuzzy cover problem in [29] in the context of a relative fuzzy set and a fuzzy Boolean coverage setting. A
generalized mathematical programming problem is formulated and used to obtain relative fuzzy optimality conditions. The results
are compared with results for the fuzzy set covering problem and (or) Boolean coverage.

17
 Kelly Osawaru & Owin Olowu / IJMTT, 69(9), 13-22, 2023

Example 4.5.3. Suppose there are five goals 𝐼 = {1,2,3,4,5} to be attained (reached) by levels (degrees) of the combination
of four means 𝐽 = {1,2,3,4} in three seasons 𝐾 = {1,2,3}. Let 𝜒𝑗 (𝑖, 𝑘) ∈ [0,1], 𝑖 ∈ 𝐼, 𝑗 ∈ 𝐽, 𝑘 ∈ 𝐾, with 𝜒𝑗 : 𝐼 × 𝐾 → [0,1] such
that

𝑝 ∈ (0,1] if mean 𝑗 is used to achieve Goal 𝑖 in Season 𝑘

𝜒𝑗 (𝑖, 𝑘) = {
0 if mean 𝑗 is not used to achieve Goal 𝑖

be the measure of application of mean 𝑗 to achieve goal 𝑖 in Season 𝑘. Thus, the grade of attained goal 𝑖 by using mean 𝑗 = 𝑗 ′
is more than the grade of attained goal 𝑖 at any season 𝑘 if 𝜒𝑗 ′ (𝑖, 𝑘) ≥ 𝜒𝑗 ′′ (𝑖, 𝑘) and the grade of attained goal 𝑖 is full at any
season 𝑘 if 𝜒𝑗 ′′′ (𝑖, 𝑘) = 1.

Let 𝑐𝑗 be the cost of using mean 𝑗 and 𝑥𝑗 , 𝑗 = 1,2,3,4, the fuzzy (partial usage) means usable defined as follows;

𝑝 ∈ (0,1] if mean 𝑗 is used

𝑥𝑗 = {
0 if mean 𝑗 is not used

Thus, the usage of mean 𝑗 = 𝑗 ∗ is more than the usage of mean 𝑗 ∗∗ if 𝑥𝑗 ∗ ≥ 𝑥𝑗 ∗∗ and any mean 𝑗 ∗∗ is fully used if 𝑥𝑗 ∗∗∗ = 1.

The values for 𝜒𝑗 (𝑖, 𝑘), 𝑐𝑗 for the seasons are summarized in the Table below:

Table 1. The Array of 𝝌𝒋 (𝒊, 𝒌)

Goals (𝑰) 1 2 3 4 5
Means (𝐽) Seasons (𝑲)
1 0.4 0.1 0.5 0.7 0.8
1(𝑐1 = 4) 2 0.2 0.2 0.4 0.8 0.7
3 0.1 0.3 0.3 0.9 0.6
1 0.1 0.3 0.8 0.2 0.6
2(𝑐2 = 2) 2 0.05 0.4 0.7 0.3 0.4
3 0.05 0.5 0.6 0.4 0.4
1 0.3 0.7 0.2 0.9 0.4
3(𝑐3 = 5) 2 0.2 0.8 0.1 0.9 0.3
3 0.1 0.9 0.05 1.0 0.02
1 0.5 0.9 0.4 0.1 0.2
4(𝑐4 = 2) 2 0.4 0.9 0.3 0.2 0.1
3 0.3 1.0 0.2 0.3 0.05

Table 1. The Array of 𝜒𝑗 (𝑖, 𝑘)

We need to find
(i) a system of fuzzy means which ensures that the goal 𝑖 is reached at each 𝑘 season with 𝛼 - level and the minimum total cost
of mean usage.

(ii) a system of fuzzy means which ensures that the goal 𝑖 is reached for all 𝑘 season with 𝛼 - level and the minimum total cost
of mean usage for all 𝑘 seasons.

5. Remark 4.5.4.
(i) We refer to 𝜒𝑗 (𝑖, 𝑘) defined above as the Relative fuzzy Boolean coverage.

(ii) 𝑥𝑗 defined above is called the fuzzy system [29].

(iii) If for each 𝑖 and 𝑗, 𝜒𝑗 (𝑖, 𝑘) is constant for all 𝑘 ∈ 𝐾 then the problem becomes a fuzzy Boolean coverage problem. Thus,
we have the following relative fuzzy sets,

18
 Kelly Osawaru & Owin Olowu / IJMTT, 69(9), 13-22, 2023

𝑃1 = {(1,1,0.4), (1,2,0.2), (1,3,0.1), (1,1,0.1), (1,2,0.2), (1,3,0.3), (1,1,0.5), (1,2,0.4), (1,3,0.3),

(1,1,0.7), (1,2,0.8), (1,3,0.9), (1,1,0.8), (1,2,0.7), (1,3,0.6)}
𝑃2 = {(1,1,0.1), (1,2,0.05), (1,3,0.05), (1,1,0.3), (1,2,0.4), (1,3,0.5), (1,1,0.8), (1,2,0.7), (1,3,0.6),
(1,1,0.2), (1,2,0.3), (1,3,0.4), (1,1,0.6), (1,2,0.4), (1,3,0.4)}
𝑃3 = {(1,1,0.3), (1,2,0.2), (1,3,0.1), (1,1,0.7), (1,2,0.8), (1,3,0.9), (1,1,0.2), (1,2,0.1), (1,3,0.05),
(1,1,0.9), (1,2,0.9), (1,3,1.0), (1,1,0.4), (1,2,0.3), (1,3,0.02)}
𝑃4 = {(1,1,0.5), (1,2,0.4), (1,3,0.3), (1,1,0.9), (1,2,0.9), (1,3,1.0), (1,1,0.4), (1,2,0.3), (1,3,0.2),
(1,1,0.1), (1,2,0.2), (1,3,0.3), (1,1,0.2), (1,2,0.1), (1,3,0.05)}

Then the RFLP problem can be expressed as a FLP problem for each 𝑘 as follows:
𝑛

Min ∑ 𝑐𝑗 𝑥𝑗
𝑗=1
𝑛

s.t [1 − ∏ (1 − 𝜒𝑗 (𝑖, 𝑘)𝑥𝑗 )] ≥ 𝛼, 𝑖 = 1.2 … , 𝑚

𝑗=𝑗
𝑥𝑗 ∈ [0,1], 𝑗 = 1,2, … , 𝑛, 𝛼 ∈ (0,1]

So, the problem in example above can be formulated as follows

𝑛

Min ∑ 𝑐𝑗 𝑥𝑗 = 4𝑥1 + 3𝑥2 + 5𝑥3 + 2𝑥4

𝑗=1
s.t 1 − [(1 − 0.4𝑥1 )(1 − 0.1𝑥2 )(1 − 0.3𝑥3 )(1 − 0.5𝑥4 )] ≥ 𝛼,
1 − [(1 − 0.1𝑥1 )(1 − 0.3𝑥2 )(1 − 0.7𝑥3 )(1 − 0.9𝑥4 )] ≥ 𝛼,
1 − [(1 − 0.5𝑥1 )(1 − 0.8𝑥2 )(1 − 0.2𝑥3 )(1 − 0.4𝑥4 )] ≥ 𝛼,
1 − [(1 − 0.7𝑥1 )(1 − 0.2𝑥2 )(1 − 0.9𝑥3 )(1 − 0.1𝑥4 )] ≥ 𝛼,
1 − [(1 − 0.8𝑥1 )(1 − 0.6𝑥2 )(1 − 0.4𝑥3 )(1 − 0.2𝑥4 )] ≥ 𝛼,
𝑥𝑗 ∈ [0,1], 𝑗 = 1,2, … , 𝑛, 𝑘 = 1.
𝑛

Min ∑ 𝑐𝑗 𝑥𝑗 = 4𝑥1 + 3𝑥2 + 5𝑥3 + 2𝑥4

𝑗=1
s.t 1 − [(1 − 0.2𝑥1 )(1 − 0.05𝑥2 )(1 − 0.2𝑥3 )(1 − 0.4𝑥4 )] ≥ 𝛼,
1 − [(1 − 0.2𝑥1 )(1 − 0.4𝑥2 )(1 − 0.8𝑥3 )(1 − 0.9𝑥4 )] ≥ 𝛼,
1 − [(1 − 0.4𝑥1 )(1 − 0.7𝑥2 )(1 − 0.1𝑥3 )(1 − 0.3𝑥4 )] ≥ 𝛼,
1 − [(1 − 0.8𝑥1 )(1 − 0.3𝑥2 )(1 − 0.9𝑥3 )(1 − 0.2𝑥4 )] ≥ 𝛼,
1 − [(1 − 0.7𝑥1 )(1 − 0.4𝑥2 )(1 − 0.3𝑥3 )(1 − 0.1𝑥4 )] ≥ 𝛼,
𝑥𝑗 ∈ [0,1], 𝑗 = 1,2, … , 𝑛, 𝑘 = 2.
𝑛

Min ∑ 𝑐𝑗 𝑥𝑗 = 4𝑥1 + 3𝑥2 + 5𝑥3 + 2𝑥4

𝑗=1
s.t 1 − [(1 − 0.1𝑥1 )(1 − 0.05𝑥2 )(1 − 0.1𝑥3 )(1 − 0.3𝑥4 )] ≥ 𝛼,
1 − [(1 − 0.3𝑥1 )(1 − 0.5𝑥2 )(1 − 0.9𝑥3 )(1 − 1.0𝑥4 )] ≥ 𝛼,
1 − [(1 − 0.3𝑥1 )(1 − 0.6𝑥2 )(1 − 0.05𝑥3 )(1 − 0.2𝑥4 )] ≥ 𝛼,
1 − [(1 − 0.9𝑥1 )(1 − 0.4𝑥2 )(1 − 1.0𝑥3 )(1 − 0.3𝑥4 )] ≥ 𝛼,
1 − [(1 − 0.6𝑥1 )(1 − 0.4𝑥2 )(1 − 0.02𝑥3 )(1 − 0.05𝑥4 )] ≥ 𝛼,
𝑥𝑗 ∈ [0,1], 𝑗 = 1,2, … , 𝑛, 𝑘 = 3.

19
 Kelly Osawaru & Owin Olowu / IJMTT, 69(9), 13-22, 2023

At an 𝛼 = 0.5-level degree, the following optimal solutions are obtained using LINGO 18.0. See Appendix 1-3 for the
output values in the Table below. They were obtained from the LINGO 18.0 program written to execute this problem.
Seasons 𝒙∗𝟏 𝒙∗𝟐 𝒙∗𝟑 𝒙∗𝟒 Min 𝒁 Remark
𝒌𝟏 1 0 0 1 6 Crisp
0.7 0 0 0.5 3.8 Fuzzy
𝒌𝟐 1 0 0 1 6 Crisp
0.7 0 0 0.5 3.8 Fuzzy
𝒌𝟑 1 0 1 1 11 Crisp
0.7 0 0.2 0.5 4.8 Fuzzy

From the Table above, a system of fuzzy means which ensures that the goal 𝑖 is reached at each 𝑘 season with 0.5 – level,
the minimum total cost of mean usage smaller with the FLP than that obtained with the traditional LP. This shows the advantage
of the FLP over LP. Also, a system of fuzzy means which ensures that the goal 𝑖 is reached for all 𝑘 season with 0.5 – level, the
minimum total cost of mean usage for all 𝑘 seasons is as summarized in the Table below;

𝒙∗𝟏 𝒙∗𝟐 𝒙∗𝟑 𝒙∗𝟒 Min 𝒁 Remark

Seasons 𝑘1 , 𝑘2 , 𝑘3 (1) - 𝑘3 (1) 𝑘1 , 𝑘2 𝑘3 (1) 11 Crisp
𝑘1 , 𝑘2 , 𝑘3 (0.7) - 𝑘3 (0.2) 𝑘1 , 𝑘2 𝑘3 (0.5) 4.8 Fuzzy

From the Table above, 4.8 is the minimum total cost of mean usage for all 𝑘 seasons. This shows that a general minimal cost
can be ascertained. This cannot be obtained with the LP and FLP problem formulation and solution approach. Thus, the RFLP is
best suited for problems exhibiting membership value changes.

6. Conclusion
RFLP, a generalization of FLP is studied through the approaches of [28, 30], where the objective and (or) constraints are
subjectively graded, incorporating manager’s input and [29, 31] which represent the objective and (or) constraints crisply but
with a Relative fuzzy Boolean coverage and fuzzy systems. The problem formulation and solution for problems characterized by
dynamic membership values are achievable and the formulations and results shows the strength of the RFLP over FLP and LP
in literature. One can also explore other approaches e.g. Tanaka’s approach [25] where the variables are fuzzy numbers. Our
future work is the application of RFLP to a generalized target coverage problem formulation.

Appendix
Appendix 1

20
 Kelly Osawaru & Owin Olowu / IJMTT, 69(9), 13-22, 2023

Appendix 2

Appendix 3

References
[1] R.E. Bellman, and L.A. Zadeh, “Decision-Making in a Fuzzy Environment,” Management Sciences, vol. 17, no. 4, 1970. [CrossRef]
[Google Scholar] [Publisher Link]
[2] Stanisław Heilpern, “Fuzzy Mappings and Fixed Point Theorem,” Journal of Mathematical Analysis and Applications, vol. 83, no. 2, pp.
566-569, 1981. [CrossRef] [Google Scholar] [Publisher Link]
[3] K.E. Osawaru, J.O. Olaleru, and Hallowed Olaoluwa, “Relative Fuzzy Sets,” Journal of Fuzzy Set Valued Analysis, vol. 2, pp. 86-110,
2018. [Publisher Link]
[4] Shizuo Kakutani, “A Generalization of Brouwer’s Fixed Point Theorem,” Duke Mathematical Journal, vol. 8, no. 3, pp. 457-459, 1941.
[CrossRef] [Google Scholar] [Publisher Link]
[5] Samuel Eilenberg, and Drane Montgomery, “Fixed Point Theorems for Multivalued Transformations,” American Journal of Mathematics,
vol. 68, no. 2, pp. 214-222, 1946. [CrossRef] [Google Scholar] [Publisher Link]
[6] Wayman L. Strother, “On an Open Question Concerning Fixed Points,” Proceedings of the American Mathematical Society, vol. 4, no. 6,
pp. 988-993, 1953. [CrossRef] [Google Scholar] [Publisher Link]

21
 Kelly Osawaru & Owin Olowu / IJMTT, 69(9), 13-22, 2023

[7] Robert L. Plunkett, “A Fixed Point Theorem for Continuous Multivalued Transformations,” Proceedings of the American Mathematical
Society, vol. 7, no. 1, pp. 160-163, 1956. [Google Scholar] [Publisher Link]
[8] L.E. JR. Ward, “Characterization of the Fixed Point Property for a Class of Set-Valued Mappings,” Fundamenta Mathematicae, vol. 50,
pp. 159-164, 1961. [CrossRef] [Google Scholar] [Publisher Link]
[9] J.R. Sam Bernard Nadler, “Multivalued Contraction Mappings,” Pacific Journal of Mathematics, vol. 30, no. 2, pp. 475-488, 1969.
[CrossRef] [Google Scholar] [Publisher Link]
[10] J.R. Sam Bernard Nadler, Some Results on Multivalued Contraction Mappings, Fleischman, W.M. (eds) Set-Valued Mappings, Selections
and Topological Properties of 2x, Lecture Notes in Mathematics, Springer, Berlin, Heidelberg, vol. 171, pp. 1-2, 2006. [Google Scholar]
[Publisher Link]
[11] R.K. Bose, and D. Sahani, “Fuzzy Mappings and Fixed Point Theorems,” Fuzzy Sets and Systems, vol. 21, no. 1, pp. 53-58, 1987.
[CrossRef] [Google Scholar] [Publisher Link]
[12] Singh, Shyam Lal, and Talwar, Rekha, “Fixed Points of Fuzzy Mappings,” Soochow Journal of Mathematics, vol. 19, no. 1, pp. 95-102,
1993. [Google Scholar] [Publisher Link]
[13] Tanmoy Som, and R.N. Mukherjee, “Some Fixed Point Theorems for Fuzzy Mappings,” Fuzzy Sets and Systems, vol. 33, no. 2, pp. 213-
219, 1989. [CrossRef] [Google Scholar] [Publisher Link]
[14] Pu Pao-Ming, and Liu Ying-Ming, “Fuzzy Topology. I. Neighborhood Structure of a Fuzzy Point and Moore-Smith Convergence,”
Journal of Mathematical Analysis and Applications, vol. 76, no. 2, pp. 571-599, 1980. [CrossRef] [Google Scholar] [Publisher Link]
[15] Byung Soo Lee, and Sung Jin Cho, “A Fixed Point Theorem for Contractive-Type Fuzzy Mappings,” Fuzzy Sets and Systems, vol. 61, no.
3, pp. 309-312, 1994. [CrossRef] [Google Scholar] [Publisher Link]
[16] Jong Yeoul Park, and Jae Ug Jeong, “Fixed Point Theorems for Fuzzy Mappings,” Fuzzy Sets and Systems, vol. 87, no. 1, pp.111-116,
1997. [CrossRef] [Google Scholar] [Publisher Link]
[17] L.A. Zadeh, “Fuzzy Set,” Information and Control, vol. 8, no. 3, pp. 338-353, 1965. [CrossRef] [Google Scholar] [Publisher Link]
[18] Ming-Po Chen, and Mau-Hsiang Shih, “Fixed Point Theorems for Point-to-Point and Point-to-Set Maps,” Journal of Mathematical
Analysis and Applications, vol. 71, no. 2, pp. 515-524, 1979. [CrossRef] [Google Scholar] [Publisher Link]
[19] R.K. Bose, and R.N. Mukherjee, “Common fixed points of some set-valued mappings”, Tamkang Journal of Mathematics, vol. 8, no. 2,
pp. 245-249, 1977.
[20] Ismat Beg, and Akbar Azam, “Fixed Points of Asymptotically Regular Multivalued Mappings,” Journal of the Australian Mathematical
Society, vol. 53, no. 3, pp. 313-326, 1992. [CrossRef] [Google Scholar] [Publisher Link]
[21] Vicente D. Estruch, and Anna Vidal, “A Note on Fixed Fuzzy Points for Fuzzy Mappings,” In: Rendiconti dell’Istituto di Matematica
dell’Università di Trieste, An International Journal of Mathematics, vol. 32, no. 2, pp. 39-45, 2001. [CrossRef] [Google Scholar]
[Publisher Link]
[22] Stefan Banach, “On Operations in Abstract Sets and their Application to Integral Equations,” Fundamenta Mathematicae, vol. 3, no. 1,
pp. 133-181, 1992. [Google Scholar] [Publisher Link]
[23] K.E. Osawaru, J.O. Olaleru, and H. Akewe, “T-Relative Fuzzy Maps and Some Fixed Point Results,” Journal of Mathematics Statistics,
vol. 13, no. 1, pp. 61-88, 2020. [Google Scholar] [Publisher Link]
[24] Rafail R. Gasimov, and Kürşat Yenilmez, “Solving Fuzzy Linear Programming with Linear Membership Functions,” Turkish Journal of
Mathematics, vol. 26, no. 4, pp. 375-396, 2002. [Google Scholar] [Publisher Link]
[25] Hideo Tanaka, Tetsuji Okuda, and Kiyoji Asai, “On Fuzzy-Mathematical Programming,” Journal of Cybernetics, vol. 3, no. 4, pp. 37-46,
1974. [CrossRef] [Publisher Link]
[26] Hans-Jürgen Zimmermann, “Fuzzy Programming and Linear Programming with Several Objective Functions,” Fuzzy Sets and Systems,
vol. 1, no. 1, pp. 45-55, 1978. [CrossRef] [Google Scholar] [Publisher Link]
[27] Hans-Jürgen Zimmermann, Fuzzy Set Theory-and its Applications, 4th ed., New York: Springer Science Business Media, 2002. [CrossRef]
[Google Scholar] [Publisher Link]
[28] Hans-Jürgen Zimmermann, “Fuzzy Programming and Linear Programming with Several Objective Functions,” Fuzzy Sets and Systems,
vol. 1, no. 1, pp. 45-55, 1978. [CrossRef] [Google Scholar] [Publisher Link]
[29] K. Zimmermann, “Fuzzy Set Covering Problem,” International Journal of General Systems, vol. 20, no. 1, pp. 127-131, 1991. [CrossRef]
[Google Scholar] [Publisher Link]
[30] Brigitte Werners, “An Interactive Fuzzy Programming System,” Fuzzy Sets and Systems, vol. 23, no. 1, pp. 131-147, 1987. [CrossRef]
[Google Scholar] [Publisher Link]
[31] M.J. Hwang, C.I. Chiang, and Y.H. Liu, “Solving a Fuzzy Set-Covering Problem,” Mathematical and Computer Modelling, vol. 40, no.
7-8, pp. 861-865, 2004. [CrossRef] [Google Scholar] [Publisher Link]

22

View publication stats