A Hybrid Algorithm for Vehicle Routing Problem with Time Windows

The VRPTW is a well-known and complex combinatorial problem which has considerable attention in recent years. Combinatorial optimization problems of this kind are NP-hard and are best solved to near optimum by heuristics. Here, we propose a two-stage optimization strategy for VRPTW. Firstly, for constructing a good initial solution, we use the stochastic PFIH, guaranteeing the diversity of the initial solution. And then a hybrid system based on the combination of SA and LNS is proposed to optimize the initial solution. Secondly, tune time windows for the Customers using a regression iterative strategy which is put forward to and figure out the optimum time for each vehicle departing. It can make the total waiting time zero. The test of this work is executed over the type C101 in Solomon's VRPTW instances. It is proved by the experiment that our algorithm can solve VRPTW efficiently and quickly.