The Research & Implementation Of Traffic Path Planning Method Based On Description Of Feature And Generalized Topology On National Road

With the growing of National Economy and the improving of the people's living, regional intercourse has become more frequent, and driving also has become an important part in daily life. Accompanying with providing more conveniences and choices, the improvement of traffic condition causes a lot of problems. Therefore, how to select a satisfied route, which costs the least time for drivers, has become a newly urgent topic to be solved.The fundamental goal of this paper is to find the optimal driving routes for robotics and motor vehicles in a broad area, which extends the driving environment from a simple small area to extremely heterogeneous one such as the whole province, nation, continent or even inter-continent. The experimental system bases on the national road model which is the description of the relationship between feature and generalized topology, together with the urban road network modeling and traffic path-planning, forms an intelligent transport system. Subjects uses A* heuristic search method and taut rubber band strategy to enhance the path planning efficiently in the wide-area.The main theoretic framework of this dissertation is the Knowledge Library Theory, Heuristic Search and Machine Learning, within the field of Artificial Intelligence. The development platform is Borland C++ Builder. With the assistance of taut rubber band strategy, location coding, and machine learning results, after the user provides start and end, the system gives an optimal route which meets the evaluation benchmarks (such as time, cost, safety, etc.). The fundamental goal is to plan out the route for a given task according to the shortest mileage within the scope of Liaoning province, and then apply this technology to other provinces. The entire system and every function module design and implementation details are introduced in this paper, the results of the system operation is also given. Finally, the paper points out the remaining disadvantages, and puts forward a future plan in accordance with the discipline trend.