Research On Path Planning Of Mobile Robot Based On Improved Particle Swarm Optimization Algorithm

Path planning of mobile robot is an important technique in robotics. To some extent, it marks the level of intelligence of robots. Path planning of mobile robot requires the robot moves from the starting position to target position at minimum cost(for example, the shortest path, the least time, the lowest energy consumption, etc) safely, in this process, the collision between robot and obstacles or other robots is avoided. Particle Swarm Optimization(PSO) is a new intelligent optimization algorithm. Since PSO algorithm has a simple concept and it is easy to be implemented. Many researchers have been trying to use PSO algorithm for solving path planning of mobile robots. But PSO algorithm is easy to converge to local optimal solutions, when PSO algorithm is been used to solve the problem of mobile robot path planning, in many cases, the optimized solution is not the global optimal path, but one of the local optimal paths.In this paper, a new PSO algorithm named Particle Swarm Optimization based on Jumping Mechanism and Pulling Operation(JMPOPSO) has been proposed. In this algorithm, two improvements have been made compared to Basic PSO(BPSO) algorithm. First, the diversity of population at the later part of the algorithm decreases drastically, BPSO algorithm is easy to converge to local optimal solutions. A jumping mechanism has been introduced to JMPOPSO algorithm to enhance the global searching ability of the algorithm. Second, a pulling operation has been introduced to accelerate the convergence of BPSO algorithm. The simulation results show that the JMPOPSO algorithm get a better global optimization performance and a faster convergence speed than the BPSO algorithm.In this paper, it is difficult to get the global optimal path using BPSO algorithm. The JMPOPSO algorithm has been applied to optimize the path planning of single mobile robot. Firstly, model of the environment of robot worked have been established. Secondly, the fitness function has been established according to the optimization objectives of the path. Finally, optimize the path using JMPOPSO algorithm based on the fitness function that has been established. Simulation results show that the JMPOPSO algorithm has better global search ability and can obtain a better path.Path planning of multi-robot system has been studied in this paper, and a new PSO algorithm named Multi-population Particle Swarm Optimization based on Coordination Mechanism(CMMPPSO) has been proposed. In this algorithm, the optimization problem of multi-robot path planning is decomposed into optimization problem of multiple single robot and each sub-population optimizes a robot. Representative individuals have been selected to exchange coordinate information for generating the elite individuals from each sub-population as the optimization path of multi-robot system. In multi-robot system, not only the problem of collisions between robots and obstacles have been taken into account, but also the collisions between robots have been considered, and a collision evaluation function has been established to evaluate the representative paths. Simulation results show that CMMPPSO algorithm can achieve the path of multi-robot.