Solving the vehicle routing problem with stochastic demands using spiral optimization

This paper presents a research work that studied a Vehicle Routing Problem with Stochastic Demands (VRPSD), in which the customer demand is the unique stochastic variable. Moreover, this variable follows a discrete distribution, and its value is only known when the vehicle arrives to the customer location.To solve this problem, we implemented the metaheuristic called Spiral Optimization, with an apriority approach and the preventing restocking strategy by only one vehicle.In order to improve that methodology, the Nearest Neighbor heuristic methodology was used, and later the Mutation, an evolutionary operator to widen the search points’ exploration zone.Besides, the mutation and exchange 2-Opt (a local search heuristic) were applied for enhancing algorithm search strategies of diversification and intensification, respectively.On the other hand, it was carried out a design of experiments 23, in order to determine the effect of each input parameter on the objective function. The eight different instances used for this DOE, were designed and developed by Galván et al. [1].The final solutions obtained were compared with the ones obtained using the hybrid algorithm EPSO for proving the efficacy and efficiency of the developed method. The comparison showed that the proposed method, obtains better solutions in all instances and improvements of up to 5,71 %.

• Fecha de publicación: 2016
• Institución: Universidad Pedagógica y Tecnológica de Colombia
• Repositorio: RiUPTC: Repositorio Institucional UPTC
• Idioma: spa
• OAI Identifier: oai:repositorio.uptc.edu.co:001/14139
• Acceso en línea: https://revistas.uptc.edu.co/index.php/ingenieria/article/view/4626; https://repositorio.uptc.edu.co/handle/001/14139
• Palabra clave: metaheuristics; spiral optimization; stochastic demands; vehicle routing; demandas estocásticas; metaheurísticas; optimización por espiral; ruteo de vehículos
• License: http://purl.org/coar/access_right/c_abf357