A Three-Stage Approach for Split Delivery Vehicle Routing Problem Solving

This paper describes a mathematical model of the split delivery vehicle routing problem. The objective is to minimize travel distance while using the lowest number of vehicles. A clustering first and routing later strategy is used to realize this objective. A three-stage approach is proposed: first, the maximum minimum distance method is employed to cluster the customer points into the lowest number groups; second, load demands in all groups are adjusted to form all possible routes according to the maximum vehicle load capacity by adopting the “push-out” and “pull-in” tactics; and third, possible routes are optimized to minimize the total travel distance. Case studies are introduced to verify the feasibility and effectiveness of this proposed algorithm. The results show that the performance of this three-stage proposed algorithm is better than the tabu search algorithm, the k-means clustering method, the BCO method and the sweep clustering algorithm in both the optimized travel distance and the less time consumption.