摘要
为解决机器人路径规划问题,在极坐标系下利用遗传算法,依据多属性决策理论提出了新的综合适应度函数。采用基于该适应度函数的遗传算法可首次规划出满足路径、时间和耗能3个约束属性的最优路径。同时引入理想适应度函数,并基于引入的理想适应度函数,提出一种新的变异算子,该变异算子可保证个体变异的方向性,对优异的父代个体有较小变异、劣质个体有较大变异。仿真结果验证了算法的可行性和有效性。
We use genetic algorithm to plan robot＇s path in the polar coordinate. A new compositive fitness function based on multiple attribute decision theory is proposed. Using genetic algorithm with this fitness function can plan the optimal path. The path satisfies the constrained attributes of path, time, and energy dissipation optimal. Based on a new introduced perfect fitness function, a new mutation operator is pro- posed. The advantage of the mutation operator is that it can guarantee the outstanding individual with smaller variable rate, inferior individual with larger variable rate. The simulation results demonstrate the effectiveness and real-time of the proposed algorithm
出处
《吉林大学学报（信息科学版）》
CAS
2012年第3期228-233,共6页
Journal of Jilin University（Information Science Edition）
基金
黑龙江省教育厅科学技术研究基金资助项目(12511002)
关键词
机器人
遗传算法
综合适应度函数
决策理论
多属性
变异算子
robot
genetic algorithm
compositive fitness function
decision theory
multiple attribute
mutation operator
