An improved LNS algorithm for real-time vehicle routing problem with time windows

This paper studies the dynamic vehicle routing problem with hard time windows (DVRPTW). The main study course of this problem was briefly reviewed. The solving strategy and algorithm of the problem are put forward. First of all, DVRPTW problem is decomposed into a series of static VRPTW. When and how to decompose the DVRP is the issue, that must be addressed. An event-trigger mechanism has been proposed and used to decompose the DVRPTW into a series of system delay-snapshots. The trigger event to be adopted is a new request arrival during the stable operation. And each snapshot is regarded as a static VRPTW. Whether each static VRPTW can quickly and efficiently be solved within a given time or a shorter time, i.e. the solving time is another issue for the DVRPTW. In the solving process, how to merge the latest requirement to the current solution is the third issue that must be solved. An improved large neighborhood search (LNS) algorithm is proposed to solve the static problem. Utilizing the remove-reinsert process of the LNS, the latest request nodes are regarded as a part of the removed nodes; these nodes can be inserted into the original or current solution in good time in the reinsertion process; meanwhile, its computing speed is high and effective for it does not need to resolve primal problem each time. Computational results of a large number of test problems, which cited from Solomon's static benchmarks and Lacker's dynamic data set, show that our method is superior to other methods in most instances.