欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
针对传统蒙特卡罗定位（MCL）算法在结构化相似环境中容易出现定位失败的问题,提出一种基于多假设粒子群优化的改进蒙特卡罗定位方法（MPSO-MCL）。以激光传感器的观测信息作为适应度函数,对MCL算法的采样粒子进行多假设粒子群优化更新,使得采样粒子向当前群体中多个最优粒子方向移动,从而使得粒子迅速收敛到后验概率密度分布取值较大的区域,实现了移动机器人高效精确自主定位。实验结果表明,MPSO-MCL算法克服了相似环境中定位的粒子匮乏问题,并且提高了定位的精确度。
英文摘要：
According to the failure of the conventional Monte Carlo localization （MCL） algorithm in structural similar environment, an algorithm called MPSO-MCL is presented which combines MCL with multi-hypotheses particle swarm optimization （MPSO）. It takes the perceptual information of laser range finder as the fitness function and carries out a multi-objects heuristic searching step on the samples generated in MCL, which flies the samples towards the regions where the value of the desired posterior density function is large. Simulation experiments demonstrate that MPSO-MCL can overcome the impoverishment of particle filter in similar environment and improve the localization precise.
同期刊论文项目
同项目期刊论文
位置：