Research On A Dynamic Path Planning Method

With the growing problem brought by urban development,traffic congestion is becoming a a typical phenomenon of “big city disease”.People’s common life need the help and the convenience of the GPS,it not only contributes the planning of the routes,but also eases the congestion in the city.The road condition in urban is becoming more and more complex along with the increase of vehicles,and thus there is a need to have a navigation system with real-time dynamic characteristics.It puts forward a new request to the accuracy of the navigation system,whereas the vehicle can be given a guidance once a emergency condition appears.The existing navigation system can meet with the demands of recommending the shortest path,which covers the time-optimal and distance-optimal.But it lacks a real-time and effective path updating scheme to deal with unexpected situations.This paper proposes a dynamic path planning method based on partitioning strategy by the combination of historical and real-time data.To improve the situation of the high cost caused by frequent updating of road conditions,the paper proposes a partitioning method.Considering the different performance about vehicles in the weekdays and weekends,the paper uses polynomial regression method associated with the linear model to improve the accuracy.The proposed dynamic path update algorithm can achieve congestion avoidance and complete the following route recommendation process according the real-time traffic information.The paper uses the map of New York City and the local traffic data as the experimental data of the system.Finally,the feasibility and reliability of the algorithm are verified by experiments.