Fuzzy Variational Calculus in Linearly Correlated Space: Part I

Gastão S. F. Frederico, Estevão Esmi and Laécio C. Barros · 2025

Read Fuzzy Variational Calculus in Linearly Correlated Space: Part I by Gastão S. F. Frederico, Estevão Esmi and Laécio C. Barros (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-07612.txt.

Introduction

Gastão Frederico, Estevão Esmi, and Laécio Barros wrote a 2025 math paper opening a series on fuzzy variational calculus in a special number space. Their space holds sums of a plain real number and a scaled fuzzy number, a setup readable as a sure value plus noise. Unlike most fuzzy settings, this space carries a full Banach space structure, making analysis tools available.

Variational calculus, the mathematics of optimizing curves and functions, needs many foundations laid first. Part one sets a total order on the space, so best and worst values make sense for fuzzy functions. It then gives fuzzy versions of the first and second optimality conditions, stated through derivatives. A generalized du Bois-Reymond lemma, a workhorse of the classical theory, closes the installment.

Catalog record

CREATORGastão S. F. Frederico, Estevão Esmi and Laécio C. Barros
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE57 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-07612

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Fuzzy Variational Calculus in Linearly Correlated Space: Part I Gastão S. F. Frederico, Estevão Esmi and Laécio C. Barros Abstract Abstract This article is the first part of series of articles that aim to present the foundations for fuzzy variational calculus for functions taking values in the space of linearly correlated fuzzy numbers \mathbb{R}_{\mathcal{F}(A)} . Recall that the space \mathbb{R}_{\mathcal{F}(A)} is composed by all sums of a real number r and a fuzzy number qA , where A is a given asymmetric fuzzy number. Two advantages of this space are that it can be equipped with a Banach space structure and, similar to additive models in Statistics, its elements can be interpreted as the sum of a deterministic expected/predictable value with an uncertain/noise component. The foundation of variational calculus theory requires the definition and establishement of many concepts and results. This aritcle presents a total order relation on \mathbb{R}_{\mathcal{F}(A)} for wihch the notions of local minimal and maximal of a \mathbb{R}_{\mathcal{F}(A)} -valued function f can be derived. We present a fuzzy version of the first and second optimality conditions in terms of derivatives of f . Finally, we present a generalized fuzzy version of du Bois–Reymond lemma which is essential in variational calculus theory. keyword: Fuzzy Variational Calculus; Linearly correlated fuzzy numbers; Order relation; Optimality conditions; Lagrange lemma; du Bois-Reymond lemma. 1 Introduction Nowadays, there are several approach about calculus for fuzzy number-valued functions and all them extends the classical calculus for real functions. As a consequence, this can also be seen in fuzzy approach to calculus of variations by considering fuzzy functions and their derivatives into the variational integrals to be extremized. This may occur naturally in many problems of physics and mechanics. Recently, Farhadinia [ 17 ] studied necessary optimality conditions for fuzzy variational problems using the fuzzy differentiability concept due to Buckley and Feuring [ 5 ] . In [ 16 ] , Fard and Zadeh use the concept of \alpha –differentiability to obtain an extended fuzzy Euler-Lagrange condition. Subsequently, Fard et al. present the fuzzy Euler–Lagrange conditions for fuzzy constrained and unconstrained variational problems under the generalized Hukuhara ( gH ) differentiability [ 15 ] . Base on the notion of gH -differentiability, Zhang et al. study Euler-Lagrange equations for fuzzy fractional variational problems under gH-Atangana-Baleanu Differentiability [ 36 ] . However, their approach does not bring light in the concept of extremals (i.e., the concepts of minimal and maximal), since there is not a consensual (total) order relation on fuzzy numbers. In the literature, there are many proposals of order relation on fuzzy numbers, for instance, see [ 18 , 26 , 32 , 37 , 35 ] and the references therein. Despite all these order relations, one of them has been usual in optimization problems that is defined in terms of endpoints of the \alpha -levels [ 20 , 27 , 28 ] . In [ 33 ] , authors investigate optimizations problems in terms of other order relation that is provided in [ 18 ] . Recently, Antczak presents optimality conditions for fuzzy optimization problems under granular differentiable fuzzy objective functions [ 1 ] . Our study do not deal with the general class of fuzzy number ( \mathbb{R}_{\mathcal{F}} ) nor does it consider the fuzzy calculus from generalized Hukuhara derivative. Instead, here we use the subclass \mathbb{R}_{\mathcal{F}(A)} of \mathbb{R}_{\mathcal{F}} which is a subset composed by linearly correlated fuzzy numbers or linearly interactive fuzzy numbers (in fact, the precise term it should be linearly interactive fuzzy variables, the interested reader can see [ 9 , 8 ] . Here, we present a different and new total order relation on \mathbb{R}_{\mathcal{F}(A)} for which we state necessary and sufficient optimality conditions. In optimization problems, the objective function often involves squaring operations. Similarly, this can also be the case when we are dealing with fuzzy optimization problems. To this propose, it is reasonable to expected that the square of a fuzzy number is greater or equal than zero. However, in general, such a property does not hold true if we define the square of a fuzzy number in terms of the usual multiplication operation. Moreover, \mathbb{R}_{\mathcal{F}(A)} is not closed under the usual multiplication operation. In contrast, Longo et al. defined the \Psi -cross product on \mathbb{R}_{\mathcal{F}(A)} for which \mathbb{R}_{\mathcal{F}(A)} is closed under it [ 24 ] . In this work, we define a total order on \mathbb{R}_{\mathcal{F}(A)} such that the \Psi -cross product of any element of \mathbb{R}_{\mathcal{F}(A)} by itself is greater or equal zero. These results are important to establish the fuzzy version of the Langrange du Bois-Reymond lemmas that are two fundamental results that have major implications in several areas of mathematics including variational calculus. To the best of our knowledge, this is the first time that these lemmas are established in the fuzzy context. The paper is organized as follows. Section 2 presents some basic notions of fuzzy set theory and properties of the subclass of fuzzy number \mathbb{R}_{\mathcal{F}(A)} . A brief recall of the calculus theory for \mathbb{R}_{\mathcal{F}(A)} -valued functions is provided in this section. In Section 3, we present an order relation on \mathbb{R}_{\mathcal{F}(A)} and states some of its properties with respect to the \Psi -cross product. Sufficient and necessary optimality conditions regarding the proposed order relation is provided in Section 4. In Section 5, we present the main results of this first article which includes a fuzzy version of the well-known Lagrange and du Bois-Reymond lemmas that will serve as a basis for establishing a theory of fuzzy variational calculus in subsequent articles of this series of articles. Finally, we conclude this paper present some final comments and concluding remarks in the last section. 2 Mathematical Background on Fuzzy Sets and Calculus on \mathbb{R}_{\mathcal{F}(A)} A fuzzy (sub)set A of the universe X\neq\emptyset is described by its membership function \mu_{{}_{A}}:X\longrightarrow[0,1] , where \mu_{A}(x) means the degree to which x belongs to A . For notational convenience, we write A(x) instead of \mu_{A}(x) . The \alpha -levels of the fuzzy subset A are classical subsets defined as follows: [A]_{\alpha}=\{x\in X:A(x)\geq\alpha\} for 0<\alpha\leq 1 and, when X is a topological space, [A]_{0}=\overline{\{x\in X:A(x)>0\}} , where \overline{Y} stands for the closure of Y\subseteq X . Here, we focus on a particular classes of fuzzy sets of \mathbb{R} called fuzzy numbers, which will be denoted by \mathbb{R}_{\mathcal{F}} . Definition 1 (Fuzzy number) . [ 2 , 12 ] The fuzzy subset A of \mathbb{R} is a fuzzy number if i. all their \alpha -levels are closed and nonempty intervals of \mathbb{R} , and ii. the support of A , supp(A)=\{x\in\mathbb{R}:A(x)>0\} , is bounded. For every \alpha\in[0,1] , we denote the \alpha -level of A\in\mathbb{R}_{\mathcal{F}} by [A]_{\alpha}:=[\underline{a}_{\alpha},\overline{a}_{\alpha}] . Note that every real number r can be regarded as a fuzzy number such that its \alpha -level is given by [r,r] for all \alpha\in[0,1]. Therefore, the set of real numbers can be embedded in \mathbb{R}_{\mathcal{F}} and we simply denote it by the symbol \mathbb{R}\subset\mathbb{R}_{\mathcal{F}} . For every C,D\in\mathbb{R}_{\mathcal{F}} and \lambda\in\mathbb{R} , the usual sum and product of C and D and the product of \lambda and C are defined level-wise respecatively by: [C+D]_{\alpha}=[\underline{c}_{\alpha}+\underline{d}_{\alpha},\overline{c}_{\alpha}+\overline{d}_{\alpha}], (1) [C\cdot D]_{\alpha}=[\min P_{\alpha},\max P_{\alpha}], (2) and [\lambda C]_{\alpha}=\left\{\begin{array}[]{cl}\left[\lambda\underline{c}_{\alpha},\lambda\overline{c}_{\alpha}\right]&,\mbox{ if }\lambda\geq 0,\\ \left[\lambda\overline{c}_{\alpha},\lambda\underline{c}_{\alpha}\right]&,\mbox{ if

Read the complete text