In this paper, we present a solution for the Open Vehicle Routing Problem with Capacity Restriction and Balance between Routes (OVRPCRBR) using bio-inspired algorithms. The flexibility, ease of use and efficiency make these algorithms a recurring option to solve problems. For OVRPCRBR, an objective function was designed with the flexibility of adaptation to various algorithms or hybridizations of them. For our experiments, we implemented the next algorithms: (a) Genetic Algorithm (GA), (b) Multi-Objective Genetic Algorithm (MOGA), (c) Particle Swarm Optimization algorithm (PSO), (d) Multi-Objective variant of Particle Swarm Optimization (MOPSO) and (e) Ant Colony Optimization (ACO) algorithm. Results shown that all these algorithms performed optimal outcomes. The ACO algorithm presents a good speed of convergence and acceptable solutions to make it a very suitable algorithm for this type of problems. Finally, a co-hybridization between ACO and MOGA algorithms was implemented, which achieved the best results for the problem in hand.

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

A Hybrid ACO-MOGA Algorithm and Bio-inspired Algorithms for Optimization of Vehicle Routing Problems

  • Eliandis Matos,
  • Fernando Gaxiola,
  • Alain Manzo-Martinez,
  • Luis González-Gurrola,
  • Graciela Ramírez-Alonso

摘要

In this paper, we present a solution for the Open Vehicle Routing Problem with Capacity Restriction and Balance between Routes (OVRPCRBR) using bio-inspired algorithms. The flexibility, ease of use and efficiency make these algorithms a recurring option to solve problems. For OVRPCRBR, an objective function was designed with the flexibility of adaptation to various algorithms or hybridizations of them. For our experiments, we implemented the next algorithms: (a) Genetic Algorithm (GA), (b) Multi-Objective Genetic Algorithm (MOGA), (c) Particle Swarm Optimization algorithm (PSO), (d) Multi-Objective variant of Particle Swarm Optimization (MOPSO) and (e) Ant Colony Optimization (ACO) algorithm. Results shown that all these algorithms performed optimal outcomes. The ACO algorithm presents a good speed of convergence and acceptable solutions to make it a very suitable algorithm for this type of problems. Finally, a co-hybridization between ACO and MOGA algorithms was implemented, which achieved the best results for the problem in hand.