| Vehicle path planning is one of the key elements in the intelligent transportation system. It can alleviate the congestion, reduce the energy consumption, and decrease the air pollution from vehicles and save the travel time of people. It makes the maximum utilization of resources and time. With the development of embedded technology and wireless sensor technology in the intelligent transportation network, the vehicle can not only fetch the map information, historical information, but also get the real-time traffic information. From a theoretical perspective, vehicle path planning problem in the transportation system has been put forward as early as 50 years ago. So its theoretical study has been very mature. However, few studies have pointed out the exact improvements of the vehicle path planning by using the real-time information under the real traffic network till now.In this paper, we focus on the processing of traffic information, the implementation and performance analysis of vehicle path planning algorithm. Through our co-operation with the Shanghai Dazhong Taxi Company, we collect lots of GPS data from more than 4,000 taxis. And with the steps of map-matching algorithm, routing algorithm, the real-time speed and the historical speed, we build up the Shanghai urban vehicle network (SUVnet). Then, we find the experience driving paths (EDP) and carry out the shortest path algorithm (SPA), the shortest time algorithm (STA), historical based algorithm (HBA), adaptive real-time algorithm (ARA) and the optimal algorithm (OPT). As real-time based algorithm has been proved to be NP hard, we propose an adaptive algorithm with the real-time information updating dynamically. Finally, we compared the performance of path planning algorithms at different dimensions. Through the experiment, we find the real-time information plays a key role on vehicle path planning. It has been shown that using real-time traffic information provided by vehicular sensor networks could improve the quality of vehicle path planning by 33%. Time-delay associated with the real-time information also has an important impact on the quality of paths. In addition, we find that the paths selected by taxi drivers are not as good as expected. We also find that GPS navigation system is helpful to the drivers and it can reduce the traveling time. And the same time, history information is important to the vehicle path planning, especially when the traffic is at non-peak time. Finally, we discuss some important guidelines of vehicle routing algorithm design for an intelligent transportation system. |
