Intelligent logistics is a modern comprehensive logistics system supported by information tech‐nology. It realizes system perception, comprehensive analysis, timely processing and self‐adjustment in transportation, warehousing, packaging, handling, circulation processing, distri‐bution, information service and so on. In the modern logistics industry, the logistics distribution carries two main features: the various and numerous products to be transported, and the rapidly increasingly transport demand. For a distribution system, the total cost relies heavily on the path length of the vehicle. In light of these, vehicle routing must be optimized to enhance the transport efficiency and lower the distribution cost. This contributes to the focus on the vehicle routing problem (VRP) in the intelligent logistics industry and the academia. Since 1959, the VRP has always been a hot topic in operations research (Dantzig and Ram‐ster)[1]. Over the years, computer simulation, logistics planning and many other strategies have been introduced to the VRP research, yielding fruitful results. Recent years has seen the rise of the multi‐compartment VRP (MCVRP) with capacity constraints. This problem mainly considers the multiple compartments of the vehicle, and ensures the separation between different products. Both the VRP and the MCVRP are NP‐hard problems, for the traveling salesman problem (TSP), a special case of the VRP, has been proved as NP‐hard (Garey, 1979)[2]. It is difficult to solve a large‐scale NP‐hard problem with traditional mathematical optimization algorithm.