摘要
为避免救援机器人路径寻优过程陷入不安全和局部的最优路径,提升救援机器人的救援能力,采用改进的Sine混沌映射初始化种群,设计带有惩罚权重的适应度安全评价函数T_(e),通过自适应选择、非线性交叉与线性变异策略改进遗传算法,实现救援路径的优化。结果表明:在两种栅格地图下,改进的算法与传统遗传算法相比,路径分别减少了5.24%和2.16%,收敛速度分别提高了30.00%和27.08%;与蚁群算法相比,收敛速度分别提高了46.15%和37.50%,搜索时间分别减少了88.50%和91.67%,说明了改进算法在复杂地图中搜索效率的有效性。
This paper seeks to improve the robots’rescue ability by avoiding the unsafe path and local optimization in selecting the optimal path.The study is focused on the effort to use the improved Sine chaotic map to initialize the population,to design the fitness safety evaluation function T_(e) with penalty weight,and to realize the optimization of rescue path by improving genetic algorithm with different strategies of adaptive selection operation,nonlinear crossover operation and linear mutation operation.The results show that based on the two grid maps,compared with the traditional genetic algorithm,the improved algorithm reduces the paths by 5.24%and by 2.16%,the convergence rates respectively increase by 30.00%and 27.08%;compared with the ant colony algorithm,the convergence rate increased by 46.15%and 37.50%,the search time decreases by 88.50%and 91.67%respectively,indicating the availability of the improved algorithm in the search efficiency of complex maps.
作者
赵杰
王馨阳
王贺
Zhao Jie;Wang Xinyang;Wang He(School of Electrical & Control Engineering, Heilongjiang University of Science & Technology, Harbin 150022, China)
出处
《黑龙江科技大学学报》
2022年第3期393-400,共8页
Journal of Heilongjiang University of Science And Technology
基金
黑龙江省省属高校基本科研业务费项目(2020-KYYWF-0694)。