RLC_ACS: An Improved Ant Colony Algorithm for VRPSDP

This paper studies the reverse logistics vehicle routing problem with simultaneous distribution of the goods and collection of the ones as same as the initial state by a homogeneous fleet of vehicles with capacities constraint and maximum distance constraint under a single depot. It presents a mixed integer programming model. For the complex feature of the fluctuating vehicle load, this paper uses an ant colony system (ACS) approach combining with the pheromone updating strategy of ASRank and MMAS ant algorithm. Also a heuristic factor based on residual loading capacity is designed to improve the vehicle loading ability. Additionally, the paper proposes a candidate list based on saving-ant, and uses a local search with sweeping in the process of tour improvement to accelerate the searching. By making comparisons with different algorithms of other researches, the experimental study indicates that RLCACS could obtain the satisfactory solution in the acceptable time.