A memetic algorithm based on two_Arch2 for multi-depot heterogeneous-vehicle capacitated arc routing problem

Abstract

With the rapid growth in the number of motor vehicles, traffic pollution has become an increasingly serious problem, due to high carbon emission and low load utilization rate. It is reasonable to plan vehicle driving routes on urban roads to deliver services, (i.e., to tackle the capacitated arc routing problem (CARP)). Previous studies on CARP have typically considered at most two objectives simultaneously; however, two objectives are not sufficient to solve the pollution problem. Consequently, a many-objective optimization model of multi-depot heterogeneous-vehicle CARP is constructed in this study that considers four objective functions: total cost, makespan, carbon emission and load utilization rate. Furthermore, based on the ideas of two-archive and information entropy, we propose a memetic algorithm based on Two_Arch2 (MATA) to tackle the constructed model. The experimental results show that MATA effectively optimizes the many-objective model and solves the problem.