The solving strategy for the real-world vehicle routing problem

The vehicle routing problem is an important problem in logistics. And it has taken as the NP hard problem. Most of the researchers want to get the best resources by many kinds of ways. But the real-world problem is much more difficult than the VRP. In order to make the model and get the solving strategy for the real-world VRP, the paper analysis the customers´ demands and divides the problem into three-stage. For the different stage, the paper puts the different strategy and approach, including the data mining for dynamic demands, two-stage approach based on GA and combinational optimize and real-time control for the stochastic demands based on GIS/GPS and FCFS. Then the human-computer interaction mode is given to make the sub-parts together.