Urban Hierarchical And Partitioned Path Planning Based On Traffic Guidance Sub Region

With the acceleration of urbanization process, car ownership of residents has increased significantly and transportation problem has gradually become an important issue which has a great impact on the economic development and residents’livelihood. As the core of traffic guidance system, path planning technology has become an important means to solve the city transportation problem. Taking the stratification characteristics of urban road network as the breakthrough point, this paper has proposed a method of dividing the road network into several traffic guidance sub regions. The research contents of path planning based on traffic guidance sub regions in this paper are summarized as follows:In chapter II, a general method for sub-region division of road network and a hierarchical and partitioned path planning algorithm are proposed. By analyzing the properties of different traffic guidance sub regions and the influence factors of division, this chapter has proposed the models of key indicators like the road congestion a and the influencer τ of dynamic road grade classification, and has designed the steps for the division of static or dynamic traffic guidance sub regions; in addition, the hierarchical and partitioned road network storage method is studied in this chapter by using adjacency matrix and gathered of node information together to store the hierarchical partitioned road network under the Arc-Node model, which lays a foundation for the implementation of path planning algorithm.In chapter III, basing on the hierarchical and partitioned road network generated by the diversion of traffic guidance sub regions, an improved hierarchical and partitioning algorithm of path planning has been proposed. By analyzing the principles of general path planning algorithms and the shortcomings of other hierarchical algorithms, we have designed an IHP algorithm by combining the Dijkstra algorithm and bidirectional search phase strategies to reduce the complexity in searching switch points, introduced a method named "virtual layer" to prevent detour in road selection by using some other hierarchical and partitioned path planning algorithms. Test data based on actual road network shows that this proposed algorithm has great practical value and better performance than other algorithms.Focusing on the dynamic characteristics of road network, chapter IV has proposed a dynamic hierarchical and partitioned path planning algorithm by using DTW model to store dynamic road weight and dynamically improving the Dijkstra algorithm. Under the principle of FIFO whose applicability has been analyzed, the DTW model has been proposed and the path planning algorithm has been dynamically improved. Combining with characteristics of the hierarchical partitioned road network, a DIHP algorithm which is based on the initial planning of historical data and the rolling planning of real-time data has been designed. Finally, with a simplified example, the rationality of the algorithm and the ability to respond to an unexpected congestion have been verified.