针对RoboCup标准组比赛平台仿人机器人NAO定位的特殊问题,在研究通用Monte Carlo定位算法基础上,构建NAO的运动模型和感知模型。通过增加一组随机粒子,改进通用Monte Carlo定位算法,增强MCL算法对于定位失败及仿人机器人NAO被"绑架"问题的适应性。最后通过仿人机器人NAO的静态定位、动态定位以及被"绑架"后的定位实验,验证了算法的有效性和鲁棒性。
Aiming at the special problems in localisation of the humanoid robot NAO of RoboCup standard group competition platform, and based on studying the general Monte Carlo localisation algorithm, we construct the NAO motion model and the perceptual model. By increasing a group of stochastic particles, we improve the general Monte Carlo localisation algorithm, and enhance the adaptability of MCL algorithm for the localisation failure and the robot-kidnapped problem of robot NAO. At last, through the static and dynamic localisations of humanoid robot NAO as well as the localisation experiment after to be "kidnapped", we verify the effectiveness and robustness of the algorithm.