Robot Dynamic Path Planning And Its Simulation Application In Maritime Traffic

With the rapid development of artificial intelligence,robot technologies have become more and more matured.Among them,robot path planning and multi-robot formation research are two hotspots which stand for a potential trend of current robot technologies.At present,robot navigation and collaboration can be seen in many fields such as military rescue,marine survey,family service and many other high-tech fields.However,many related works are still focused on the global path planning problem in which the environmental information is totally known and there is no dynamic obstacle environment.As a result,the robot path planning problem in the dynamic environment is still a difficult problem to be solved.In this thesis,considering the classical ant colony algorithm which is applied for robot path planning in dynamic environment has the problems such as slow convergence speed,large path integral turning angle and low adaptability to environmental changes,an adaptive dynamic path planning algorithm based on ant colony-clustering algorithm is proposed firstly.Moreover,the multi-robot formation and its application in sea transportation of the proposed algorithm are discussed.The main works of this thesis are been listed as follows.1.An new self-adaptive path planning algorithm based on ant colony-clustering algorithm in unknown and dynamic environment is proposed.In this proposed algorithm,the optimal searching radius are decided accurately by introducing the clustering algorithm which can judge the environmental complexity automatically so as to fully utilize the limited computing capacity of robot and improve the convergence speed.By identifying the diagonal obstacles and generating virtual obstacles,the pair angle obstacles are avoided to cross.Finally,the obtained dynamic planning path is smoothed by the smoothing mechanism in order to reduce the path length and reducing the cumulative turning angle effectively.The simulation results show that the proposed algorithm can automatically select the appropriate search radius according to the complexity of the obstacles,complete the dynamic path planning,and show good environmental adaptability and better comprehensive path optimization performance.2.The path planning problem of multi-agent triangle formation in dynamic environment is studied.Based on the research results of single agent path planning in dynamic environment,the path planning algorithm for multi-agent formation in dynamic environment is designed.Aiming at the selection problem of local search radius in the third chapter,a fuzzy mechanism which can perform a fuzzy weight assignment is introduced to bridge between the-clustering method and the duty cycle method,and the more reasonable local search radius is obtained.Then,the multi-robot coordination strategy is designed based on the leading-follower method.Finally,the simulation experiment shows that the effectiveness and the robust performance of the multi-robot formation is achieved.3.Self-adaptive path planning algorithm design for sea transportation.Aiming at the path planning problem of unmanned surface craft in the dynamic environment,an ant colony-clustering-based autonomous path planning method based on electronic chart rasterization is proposed.The grid map of the marine environment is established based on S-57 electronic chart.The proposed algorithm in chart 3 should be improved according characters of the unmanned surface craft such that it is suitable for dynamic change of the marine environment.Compared with the traditional global path planning method,the proposed algorithm can plan suitable path successfully in complicate and dynamic sea environment.What is more,it also improves the dynamic path planning capacity and keeps the unmanned surface craft in safe state.