基于遗传进化算法的机器人全局路径规划  

The Path Planning for Robots Based on the Genetic Algorithm

作　　者：张东亮[1] 孙秀云[1] 田新诚[1] 张乃标[2] 

机构地区：[1]山东大学,山东济南250061 [2]枣庄学院,山东枣庄277160

出　　处：《机械与电子》2005年第4期49-51,共3页Machinery & Electronics

摘　　要：研究了机器人在已知环境下用遗传进化算法进行最优路径搜索的方法。在此方法中,把二维编码简化为一维编码,提高了搜索的速度;把障碍物作为排斥子,提出了简单、确定的适应度函数;最后提出了对路径的转折节点处进行平滑处理的方法。仿真实验证明了该方法能够得到机器人行进的最优化路径。A new method based on the genetic algorithm is proposed for searching the best path for robots when the positions of obstacles are determined.For this method,the onedimensional codes replace the traditional twodimensional codes,resulting in a significant increase of searching speed. A simple,definite fitness function is designed by considering the obstacles as repellents.Replacing the broken lines around the corners with appropriate arcshaped lines can smooth the path containing sharp corners.Finally,the simulation results confirm that this method can generate the optimum operating path for robots.

关 键 词：全局路径规划 遗传进化算法 路径平滑 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]