Multi-objective optimization of vehicle routing problem using evolutionary algorithm with memory

Abstract

The idea of a new evolutionary algorithm with a memory aspect included is proposed to find a multi-objective optimized solution for the vehicle routing problem with time windows. This algorithm uses a population of agents that individually search for optimal solutions. The agent memory incorporates the process of learning from the experience of each individual agent as well as from the experience of the whole population. This algorithm uses crossover operation to define each agent's evolution. In this paper, we choose the Best Cost Route Crossover (BCRC) operator as a base. This operator is well-suited for VPRTW problems; however, it does not treat both parents symmetrically, which is not natural for general evolutionary processes. A part of the paper is devoted to finding an extension of the BCRC operator in order to improve the inheritance of chromosomes from both parents. Thus, the proposed evolutionary algorithm is implemented with the use of two crossover operators: BCRC and its extendedmodified version. We analyze the results obtained from both versions applied to Solomon's and Gehring & Homberger's instances. We conclude that the proposed method with the modified version of the BCRC operator gives statistically better results than those obtained using the original BCRC. It seems that an evolutionary algorithm with the memory and modification of the Best Cost Route Crossover Operator leads to very promising results when compared to other solutions presented in the literature.