Scholarlink

This work addresses the challenge of optimizing the management of water resources based on a model inspired by the transportation problem, which considers different aspects such as types of water, carriers, supply nodes, demand nodes, and an objective function to optimize the cost of water transport. Mathematical programming algorithms and Grammatical Differential Evolution (GDE) have been used to optimize this model. However, a fundamental problem arises: the model incorporates several real-world constraints that must be fully met for a solution to be feasible. When using population-based metaheuristics such as GDE, evolutionarily generated individuals that occasionally do not satisfy such constraints, causing stagnation and premature convergence towards local optimums. In this paper, we propose a novel heuristic that generates solutions that satisfy the problem constraints. Through a convergence criterion, the heuristic is activated, injecting a percentage of feasible solutions to the current population during the evolutionary process, thereby improving the convergence process. The results demonstrate that the incorporation of the heuristic in GDE provides an effective strategy that allows improving the optimization of water use while respecting real-world constraints.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Generating Feasible Solutions Through a Heuristic Approach in a Water Use Optimization Model Based on the Transportation Problem

  • 章节

  • 发表日期：

    2025年

查看全文
  • José Alejandro Cornejo-Acosta,
  • Blanca Verónica Zúñiga-Núñez,
  • Valentín Calzada-Ledesma

摘要

This work addresses the challenge of optimizing the management of water resources based on a model inspired by the transportation problem, which considers different aspects such as types of water, carriers, supply nodes, demand nodes, and an objective function to optimize the cost of water transport. Mathematical programming algorithms and Grammatical Differential Evolution (GDE) have been used to optimize this model. However, a fundamental problem arises: the model incorporates several real-world constraints that must be fully met for a solution to be feasible. When using population-based metaheuristics such as GDE, evolutionarily generated individuals that occasionally do not satisfy such constraints, causing stagnation and premature convergence towards local optimums. In this paper, we propose a novel heuristic that generates solutions that satisfy the problem constraints. Through a convergence criterion, the heuristic is activated, injecting a percentage of feasible solutions to the current population during the evolutionary process, thereby improving the convergence process. The results demonstrate that the incorporation of the heuristic in GDE provides an effective strategy that allows improving the optimization of water use while respecting real-world constraints.