理性遗传算法及其在多机器人运动协调中的应用

面对传统遗传算法在解决一些复杂问题时所存在的收敛慢或早熟等困难 ,基于仿人理性决策原则 ,提出一种具有更丰富进化含义的进化算法——理性遗传算法 .其通过遗传信息的反馈或理性规则的建立来指导遗传操作的进行 ,从而将种群内部知识与经验的继承和学习更有效地结合在遗传算法之中 .相对于传统遗传算法 ,较好地解决了多机器人确知环境下协调运动规划问题 .理论分析和仿真实验结果都是令人鼓舞的 .

When conventional genetic algorithm(GA) is used to cope with some complex problems, slow convergence or prematurity often occurs. A novel evolutionary algorithm, based on the rational decision  making of human, the rational genetic algorithm (RGA) is proposed to solve these problems. The key point of RGA is to use the genetic information feedback and set up rational rules to guide the evolution of genetic individuals. The proposed RGA effectively incorporates inheriting and learning behaviors of knowledge and experiences of species into GA.The problem of multi  robot motion cooperation under known circumstance can be solved better by RGA than conventional GA. Theoretical analysis and simulation results show the validity of RGA.

国家“8 6 3”计划智能机器人主题项目 ( 86 3-512 - 982 0 - 0 6 );;中国科学院沈阳自动化研究所机器人学开放实验室基金 ( RL 199910 )资助