基于改进遗传算法的移动机器人路径规划

Path planning is the key technology to realize autonomous navigation of mobile robots. For the problem that the path length is not the shortest and the path is not coherent in the two plan cycles with conventional path planning method, a new method for inter-frame correlation smooth path planning based on improved genetic algorithm is proposed. Firstly, the candidate paths were generated by combining random and directional search methods. Then, the insertion operator and deletion operator were added to conventional genetic operators, and the path coherence of two plan cycles was considered in the fitness function to calculate the fitness value of each candidate path. Finally, the path with the highest fitness value was output as the current optimal path. Simulation results show that the proposed method is correct and feasible. Experimental results show that, compared with A^* algorithm and conventional genetic algorithm, the path length of mobile robot is reduced by 3.05% and 1.85%, the variation of maximum yaw angle is reduced by 38.02% and 32.43%, and the sum of absolute value of turning angle is reduced by 23.97% and 19.94% respectively during the movement of mobile robot. It shows that the resulting path of this method is more optimal, which observably improves the moving efficiency and stationarity of the mobile robot.