基于小生境伪并行遗传算法的自主机器人路径优化

Niche Pseudo-parallel Genetic Algorithms for Path Optimization of Autonomous Robot

提出了一种新的遗传算法——小生境伪并行遗传算法(NPPGA ) ,用于对自主机器人运动路径进行优化。这种遗传算法是对小生境遗传算法和并行遗传算法的一个有效改进,该方法兼顾了对局部最优解和全局最优解的搜索,维持了群体的多样性,避免了早期收敛现象的发生;同时也增强了自然群体进化的并行性,加快了搜索进程。实验结果表明,该方法是有效可行的,通过合理选取选择、交叉、变异及小生境算子,能得到机器人路径规划最优解。

A new method named Niche Pseudo-Parallel Genetic Algorithms is presented for path optimization of autonomous robot.The NPPGA is an effective betterment to maintain the community diversity for the sake of avoiding premature convergence and strengthen the evolutionary parallelism of a colony to accelerate the search process combined niche genetic algorithms with pseudo-parallel genetic algorithms.Experimental results show that NPPGA is feasible and effective.It can obtain the optimal path by choosing selection,crossover,mutation and niche operators logically.