Hybrid multiobjective evolutionary algorithm with fast sampling strategy-based global search and route sequence difference-based local search for VRPTW

Abstract

The vehicle routing problem with time windows (VRPTW) is an important and widely studied combinatorial optimization problems. This paper aims at VRPTW with the objectives of reducing the number of vehicles and minimizing the time-wasting during the delivery process caused by early arrival. To solve this NP-hard problem, a hybrid multiobjective evolutionary algorithm with fast sampling strategy-based global search and route sequence difference-based local search (HMOEA-GL) is proposed. Firstly, fast sampling strategy-based global search (FSS-GS) of HMOEA-GL extensively explores the entire solution space to quickly guide the search direction towards the center and edge areas of Pareto frontier. Secondly, route sequence difference-based local search (RSD-LS) is executed on the individuals with poor performance in the population obtained by FSS-GS to enhance the search ability of HMOEA-GL. In addition, the suitable coding method and proper genetic operators are designed, especially, a simple insertion search is used to reduce the number of vehicles in VRPTW. Comparing with NSGA-II, SPEA2, and MOEA/D, experimental results on 12 Solomon benchmark test problems indicate that the proposed HMOEA-GL is effective, and more excellent in convergence, while maintaining a satisfying distribution performance. HMOEA-GL could be an effective intelligent algorithm for expert and intelligent decision support system to help logistics companies to make decisions.