论文部分内容阅读
基于广泛使用的概率栅格地图,提出了一种移动机器人可定位性估计方法.通过对定位Fisher信息矩阵进行栅格离散化,提出了静态可定位性矩阵,该矩阵适用于已知地图条件下的离线估计.在此基础上,针对在线估计中环境存在的非预期动态变化问题,采用局部感知的未知障碍物影响因子来修正静态可定位性矩阵,进而得到动态可定位性矩阵,该矩阵定量描述了机器人可定位性能力及其方向性.各种典型环境下的机器人实验结果表明了所提方法的有效性.
Based on the widely used probability grid map, a method to estimate the locality of mobile robots is proposed. By locating the Fisher information matrix for grid discretization, a static locatable matrix is proposed, which is suitable for the known map conditions Based on this, aiming at the unpredictable dynamic change of the environment in online estimation, the locally perceived unknown obstacle influence factor is used to correct the static locatability matrix, then the dynamic locatability matrix is obtained, which is quantitative The robot locatable ability and its directivity are described.The experimental results of robots in various typical environments show the effectiveness of the proposed method.