Research On Intelligent Vehicle Path Planning Based On Improved PSO And RRT

In the new era full of modern technology,mobile robots play an indispensable role in daily life and production.Wheeled mobile robots integrate multiple disciplines such as artificial intelligence,sensor technology,computers,and mechanical engineering.Not only can they accomplish intelligent handling,terrain exploration,and unmanned driving tasks,but their development is also an important manifestation of the level of industrial automation and high-tech in various countries.With the development of autonomous driving technology,autonomous obstacle avoidance path planning has also become the focus and focus of research by scholars from various countries.However,in the research of path planning algorithms proposed by many scholars,there is still much room for optimization and improvement in the face of different scenarios and objects.Therefore,this topic conducts research on intelligent obstacle avoidance algorithms for intelligent vehicle path planning to improve the performance of intelligent algorithms.This article takes ITA9710-08 pure electric unmanned intelligent vehicle as the research object,takes the campus roads of Central North University as the working scene,builds a simulation model based on MATLAB software,selects C language as the programming tool for research,and explores the effectiveness of intelligent path planning algorithms.The specific work content is as follows:(1)Aiming at the problems of premature convergence,slow convergence speed,and low search accuracy in traditional particle swarm optimization(PSO)algorithm for robot path planning,the vigilance mechanism of Sparrow Algorithm(SSA)is combined with the population of PSO algorithm to optimize the inertia weight factor and learning factor in PSO algorithm to prevent premature convergence while the algorithm converges rapidly.Finally,the new algorithm was validated through Matlab simulation experiments,and the results were analyzed.After that,the ITA9710-08 pure electric driverless intelligent vehicle was used as the research object for real vehicle verification on the campus roads of Central North University.The simulation results show that the SSA-PSO algorithm reduces the path length by 2m,3.071 m,and 10 m on different size maps compared to the PSO algorithm in solving robot path planning problems,and improves its performance by 7%,10%,and 14%,respectively.The complexity of the SSA-PSO algorithm is relatively low,and the path planning time is relatively short.The actual vehicle verification results show that the SSA-PSO algorithm has shorter response time,higher planning efficiency,and is not easy to fall into local optimization in local path planning.(2)Aiming at the limitations of traditional RRT-Connect algorithm in solving the optimization problem of intelligent vehicle path planning,a RRT-Connect algorithm based on triangular inequality is proposed to solve the optimization limitations using the principle of triangular inequality.Finally,the sensors of the ITA9710-08 pure electric driverless intelligent vehicle platform collect the road environment on the campus of Central North University,generate 8 different environmental maps,and conduct comparative analysis with RRT and RRT-Connect algorithms through simulation.The results show that the new algorithm improves the path length by an average of 20% compared to RRT algorithm,increases the path length by an average of 16% compared to RRT-Connect algorithm,increases the planning time by an average of 47% compared to RRT algorithm,and decreases the planning time by an average of 2% compared to RRT-Connect algorithm.(3)The two new algorithms proposed in this paper are verified on a real vehicle.Through the establishment of the ITA9710-08 pure electric unmanned intelligent vehicle experimental platform,the path planning experiment was conducted on the campus roads of Central North University.After collecting the internal roads of the campus through a smart car,different environmental maps are generated for experimental verification.Among them,SSA-PSO algorithm is used for local path planning,and RRT-Connect algorithm based on triangular inequality is used for global path planning.The intelligent vehicle safely arrives at its destination without collision,verifying the effectiveness of the algorithm.The results show that the SSA-PSO algorithm is more efficient for local path planning and the RRT algorithm based on triangular inequalities is more efficient for global path planning.