论文部分内容阅读
针对移动机器人定位研究中的位姿跟踪、全局定位和“绑架”三类问题,提出一种基于遗传算法的移动机器人自定位方法.设计基于位置相似度的种群适应度计算方法,利用实值编码方式实现种群的交叉、变异,有效提高算法的实时性.针对机器人定位过程中的“绑架”现象,在常规遗传算法的基础上引入种群发散算子,减小种群匮乏效应.在此基础上,利用机器人运动模型更新种群状态实现机器人的连续定位.在实际室内环境进行机器人定位实验,证实本文算法的有效性.
In order to solve the three problems of pose tracking, global positioning and “abduction ” in positioning research of mobile robots, a self-positioning method of mobile robot based on genetic algorithm is proposed. The calculation method of population fitness based on position similarity is designed, Value coding to realize the crossover and mutation of population and improve the real-time of the algorithm effectively.Aiming at the phenomenon of “kidnapping ” in the process of robot localization, the population divergence operator is introduced based on the conventional genetic algorithm to reduce the effect of population starvation. Based on this, the robotic motion model was used to update the population status to realize the robot’s continuous positioning, and robot localization experiment was conducted in the actual indoor environment, which verified the effectiveness of the proposed algorithm.