Adaptive Genetic Algorithm Enhancements for Path Planning of Mobile Robots

Abstract

An adaptive Genetic Algorithm (GA) is proposed, which focuses on the automatic adjustments of crossover probability and mutation probability with the changeable environmental parameters. The improved algorithm can overcome some disadvantages of traditional GA, such as, early falling into local optimum, lower convergence speed and large calculation etc. In sequence, the complementary characteristic between crossover probability and mutation probability is obtained through carrying out the numerical simulation. The results demonstrate that, compared with the traditional GA, the adaptive one leads to better performance in path curves and fitness, when 30 generations operations is implemented. This solution mentioned above, is proved to a better choice for practical application in path planning for mobile robots.