Collaborative multi-depot pickup and delivery vehicle routing problem with split loads and time windows

Abstract

Optimization of collaborative multi-depot pickup and delivery logistics networks (CMDPDLN) with split loads and time windows involves a customer demand splitting strategy and a multi-depot pickup and delivery vehicle routing problem under time window constraints. In the collaborative network, the customer demand splitting scheme based on customer clustering aims to achieve the balance of demands’ spatial distribution and improve the efficiency of logistics transportation. The multi-depot pickup and delivery vehicle routing problem focuses on establishing a collaborative network optimization model to coordinate the pickup and delivery services among multiple depots and determine the optimal routes with reduced operating cost through logistics resource sharing. A 3D customer clustering algorithm with split load strategies is developed to reassign each customer to its favorable service provider considering multiple customer service characteristics. A hybrid genetic algorithm with tabu search is designed to optimize the pickup and delivery routes and maximize the logistics resource utilization. A realistic logistics network in Chongqing, China is used to test the performance of the proposed solution methods for the CMDPDLN optimization. Computational results show the effectiveness of customer clustering and demand splitting in simplifying and improving the large-scale collaborative network, and the adaptability of the hybrid algorithm in finding the minimal-cost vehicle routes. Therefore, the collaboration and demand split strategy adopted in network optimization can provide a reference for logistics operational management and facilitate sustainable pickup and delivery networks.