Path Planning of Mobile Robot Based on Genetic-Ant Colony Algorithm
Chenfei Hou,Chenxu Li,Jing Yang,Jianhui Zhu,Wenxuan Wang
DOI: https://doi.org/10.1088/1742-6596/2216/1/012011
2022-01-01
Journal of Physics Conference Series
Abstract:Mobile robots are widely used in daily life, and ant colony algorithm is often adopted for its path planning. However, the classic ant colony algorithm has poor search ability and search speed due to the monotony of individuals within the population. In order to solve the problems of local optimal solution, slow convergence speed and low search efficiency of classic ant colony algorithm, a genetic-ant colony algorithm is proposed in this paper. The algorithm parameters are optimized by simulation experiment to determine the best parameter combination. The method is to give a certain range of random pheromone values to the primary ant individuals to enrich the population diversity. In each cycle, m ants are sent for foraging. After each foraging cycle, aiming at the growth of individual pheromone value, the fitness function of genetic algorithm is established to evaluate individual fitness value, and the parental individuals carry out genetic operations such as selection, crossover and variation to produce offspring individuals with the same number and higher pheromone level, the offspring individuals carry out the next cycle of foraging and keep cycling. The population diversity is maintained at a high level, while the increase in the level of individual pheromone speeds up the convergence and operation of the algorithm. The simulation results show that the running time of the improved algorithm is significantly shorter than that of the traditional ant colony algorithm in the 40 * 40 grid map, and the running time of the algorithm is reduced by 21.34%; in the complex grid environment, the frequency of obtaining the optimal path of the algorithm is significantly higher than that of ordinary ant colonies, and the frequency is increased; and the number of iterations when the optimal path is reached for the first time is significantly reduced,–which has verified the effectiveness, stability and superiority of genetic-ant colony algorithm.