Research On Path Planning Algorithm Based On Multiple Moving Points

With the rapid development of science and technology,various mobile robots are used in people’s life,and due to the limitation of single robot’s constraint ability,multi-robot collaborative task implementation becomes the current research hotspot when facing the task conditions of complex environment.The paper investigates the path planning algorithm for multiple mobile robots by considering various robots as mobile points.Based on the development trend of path planning and the current status of research on task goal matching of multiple moving points,the thesis starts from the technical analysis of path planning of single objective and further studies the task goal matching algorithm of multiple moving points,thus realizing the algorithm research of path planning problem of multiple moving points.The main research contents of this thesis are as follows.(1)The single-objective path planning problem is to find a safe path for the moving point to move from the starting position to the target position without collision,and the path length is as optimal as possible.The particle swarm algorithm has a simple structure and is easy to combine with other algorithms,but it also has its own problems of converging too fast and falling into local optimal solutions easily with premature maturity.Based on this problem,the thesis proposes a hybrid particle swarm algorithm that adds a second global optimal solution by adding a genetic operator,and improves the efficiency of finding the optimal solution by increasing the diversity of solutions.By building a map model and conducting simulation experiments for comparison,the experiments show that the improved hybrid algorithm can effectively improve the path planning results.(2)Task planning for multiple targets is essentially an allocation of resources.The path task planning of multiple moving points needs to take into account the condition constraints of the moving points themselves and the demand constraints of the objectives.Genetic algorithms are mostly used for multi-objective solution problems.The basic genetic algorithm may converge too slowly at the beginning due to the low fitness of the randomly generated population and the problem of slow merit seeking.The paper proposes an elite operator that increases the individuals with higher fitness values in the population and reduces the individuals with lower fitness values by intervening in the selection operation of the algorithm,which increases the merit seeking speed of the algorithm.The VRP problem with multiple moving points is used as a simulation experiment,and the results effectively improve the probability of the algorithm reaching the theoretical optimal solution.Finally,the thesis combines the above research on path planning for single target and task planning for multiple targets,and designs and implements a multi-moving point path planning optimization algorithm,whose core idea is to find the distance between all moving points and task target points with a hybrid particle swarm algorithm,substitute into an improved genetic algorithm,and assign task targets,effectively realizing path planning for multi-moving point task target assignment.