Design And Implementation Of Vehicle Path Planning System Based On Multi Agent System

With the continuous improvement of car ownership in recent years,traffic congestion has become an urgent problem to be solved.Climate problems such as air pollution and heat island phenomenon have become more and more prominent.In order to alleviate the pressure caused by these problems,it is necessary to make more scientific and reasonable planning for the driving path of vehicles.Vehicle routing problem has always been one of the main research contents of motion planning since it was put forward.It has been widely used in road construction,travel planning,smart city construction and other fields,involving technology including data mining,statistical analysis and so on.Around this problem,this thesis mine and reconstruct the road network information through the historical trajectory data,use the improved ant colony algorithm based on the negative feedback mechanism to plan the vehicle path,and finally rely on the multi-agent system to build an efficient and accurate vehicle path planning system.The main contents of this thesis are as follows:1.Research the congestion area detection algorithm based on time decay model.Due to the different congestion conditions of different road sections,the road network information should be improved before the path planning.In this thesis,the trajectory data is used to mine the road section area,and the hot spot area mining algorithm based on the time decay model is used to re label the road network information.The congestion situation of different road sections is obtained,which solves the problem that the traditional clustering algorithm is difficult to find the high-density area under a specific time window.The experimental results show that the congestion detection algorithm based on time decay model can mine the congestion area more accurately and completely than the traditional clustering algorithm combined with time,improve the operation efficiency of the path planning system and calculate a more reasonable planning route.2.Research the improved ant colony optimization algorithm based on negative feedback pheromone.In view of the congestion of road network environment,this paper defines the optimal path measurement method which integrates path length and path congestion.By improving the heuristic information and pheromone update in ant colony algorithm,an improved ant colony algorithm NAA based on the feedback pheromone is proposed.The algorithm can dynamically adjust the pheromone volatilization speed of each region in the ant colony algorithm by judging the congestion of each road section,improve the diversity of solutions by using the negative feedback mechanism,and solve the local optimal solution problem which is easy to appear in the ant colony algorithm.The comparison on the open data set shows that the algorithm can reach the optimal solution on the path planning problem.3.Design and implement vehicle path planning system based on multi-agent.According to the practical application requirements of vehicle path planning,a multi-agent-based vehicle path planning model is built.At the same time,a complete requirement analysis,system design and implementation are carried out.It is developed in Java language and implemented on the open-source multi-agent development environment JADE platform.Through the function test and performance test of the system,the accuracy and efficiency of the multi-agent-based vehicle path planning system designed in this paper are verified.The research results will help to improve the travel efficiency of vehicles and help the construction of intelligent transportation and smart city.