An evolutionary approach for the optimization of production-distribution network design

Abstract. In this Thesis an evolutionary technique for finding (near) optimal solutions to the two-stage fixed charge transportation problem (finding minimum cost transportation configurations when considering per unit transportation cost, fixed charges associated to routes, limited capacity of production plants and unlimited capacity of distribution centers) is proposed. Basically, the Hybrid Adaptive Evolutionary Algorithm with three different domain specific genetic operators (one crossover: network; two mutations: distribution and production) is applied. Here a candidate solution is encoded using two matrices for each stage of the network. The crossover operator exchanges the transportation plan of the second stage between two networks. The distribution mutation operator closes a randomly selected distribution center, so products, that were distributed to customers by such center, return to their plants where those came from. The mutation operator changes the distribution plan in the first stage of the network from a randomly selected production plant. After applying an operator, a balance method is used. Finally, the fitness function is the sum of transportation costs, including the unit transportation costs and the fixed cost incurred when using a route. Computational experiments carried on twenty instances of the problem that are available in the literature, show that our approach is able to find equal or better solutions compared to those reported in the literature.

• Autores: Puerta Jaramillo, David Leonardo
• Fecha de publicación: 2016
• Institución: Universidad Nacional de Colombia
• Repositorio: Universidad Nacional de Colombia
• Idioma: spa
• OAI Identifier: oai:repositorio.unal.edu.co:unal/59123
• Acceso en línea: https://repositorio.unal.edu.co/handle/unal/59123; http://bdigital.unal.edu.co/56372/
• Palabra clave: 0 Generalidades / Computer science, information and general works; 38 Comercio, comunicaciones, transporte / Commerce, communications and transportation; 51 Matemáticas / Mathematics; 6 Tecnología (ciencias aplicadas) / Technology; 65 Gerencia y servicios auxiliares / Management and public relations; Transportation problem; Fixed charge; Evolutionary algorithm; Supply chain; Optimization; Algoritmo Evolutivo; Cadenas de suministro; Problema de Transporte; Cargo fijo; Optimización
• Rights: openAccess
• License: Atribución-NoComercial 4.0 Internacional