摘要
为了提高车间内机器人导航规划的实时性、减少导航路径长度,提出了混沌狼群围捕算法。建立了工作环境的矢量模型与路径的适应度函数;在狼群围捕算法基础上,使用Tent混沌映射改进了狼群初始化方法,使狼群初始分布更加均匀,有利于算法初期对整个工作区域的遍历式搜索;鉴于levy飞行长时间短距离来回搜索与偶尔长距离搜索相互穿插的特点,将其应用于改进围捕步长,有利于算法长期的细致搜索并保持跳出局部极值的能力;借鉴遗传思想改进狼群进化方法,将算法向收敛的方向进行引导;基于以上改进,提出了混沌狼群围捕算法。在车间环境下进行仿真验证,与传统狼群围捕算法相比,混沌狼群围捕算法的导航规划用时减少了37.5%,最优导航路径长度减少了16.7%。
To improve real-time of workshop used robot navigation planning and lessen navigation path length,navigation planning method by chaotic wolf pack besieging algorithm is put forward. Vector model of working environment and fitness function of the path are built. On the basis of wolf pack besieging algorithm,Tent chaotic mapping is used to initialize wolf pack location,which makes wolf pack distribute evenly,and it benefits to ergodic search of the whole working area.Considering short distance scanning in longtime and long-distance scanning occasionally interspersed with each other for levy flight,the method is applied to improve besieging step,which benefits to detailed search in longtime and overstep the local extremum occasionally. Wolf pack evolution method is improved by heredity ideology,which can guide the algorithm upward to convergence. Based on these three improved aspects above,chaotic wolf pack besieging algorithm is proposed. Simulation is executed in workshop environment,and the result shows that compared with basic algorithm,time-consuming of navigation planning by chaotic wolf pack besieging algorithm decrease by 37.5%,and length of optimal path decrease by 16.7%.
作者
周璟
ZHOU Jing(Department of Electronic Information,Wuxi Vocational Institute of Arts&Technology,Jiangsu Yixing214200,China)
出处
《机械设计与制造》
北大核心
2020年第1期251-255,共5页
Machinery Design & Manufacture
基金
江苏省大学生创新创业训练计划项目(201713749011Y)