Research On Path Planning Of Mobile Robot Based On Multi-factor Ant Colony Algorithm

With the rapid development of artificial intelligence,Intemet+ and 5G communication,mobile robots are increasingly used in all walks of life.These mobile robots work in complex and changing spaces,and require advanced path planning algorithms to improve intelligence and security.Sexuality and efficiency.The path planning algorithm not only requires the calculation results to be safe and reliable,but also meets the needs of actual use.It also requires superior algorithm performance and strong anti-interference ability.Existing path planning algorithms tend to pursue only a single improvement,while ignoring the overall performance of the algorithm.Therefore,how to find a comprehensive optimal path and how to effectively avoid obstacles is a hot and difficult problem of robot technology.Firstly,the grid is used to model the map,the path planning problem is described as the optimization problem in mathematical language,and the principle of the classical ant colony algorithm is introduced.The existing improvement mechanism is summarized and added to the classical ant colony algorithm.in.Secondly,the storage mechanism of the grid method map model is improved,so that the algorithm can read the map information more quickly,and propose an improved map turning mechanism to make the robot turn safe and reliable.Thirdly,a multi-factor ant colony algorithm is proposed to improve the heuristic function in the state transition probability formula,and to add heuristic information such as distance,turn,and height difference,and guide the ants to select the next transfer grid in various aspects;and improve the ant colony pheromone.The release mechanism correlates the concentration of pheromone release with the comprehensive index of the path,thereby bringing the ant colony closer to the integrated optimal path.Finally,a dynamic collision avoidance strategy based on rolling window is proposed.The local collision prediction of rolling window is analyzed.The local collision avoidance response strategy of rolling window is designed.The solution to the problem of "sub-object unreachable" is proposed.The adaptive step size algorithm and multi-factor ant colony local path planning algorithm form a multi-factor ant colony dynamic path planning algorithm based on rolling window.The matlab simulation software is used to simulate the static path planning of multi-factor ant colony algorithm and the dynamic path planning of multi-factor ant colony algorithm in rolling window.The experiments are carried out from contrast experiments,specific map experiments and algorithm reliability experiments.Verification,simulation results show the feasibility and superiority of the article algorithm.