论文部分内容阅读
提出一种基于二维激光雷达的移动机器人实时位姿估计算法。该算法首先将环境数据聚类为不同的障碍物,然后利用障碍物的特征匹配来关联相邻2帧的障碍物,最后通过优化位置序列间的方向角实现机器人的位姿估计。与ICP(迭代最近点)算法对比验证表明:该算法与ICP具有相同数量级的位姿估计精度,且明显提高了计算效率。
A real-time pose estimation algorithm for mobile robot based on two-dimensional lidar is proposed. Firstly, the environment data is clustered into different obstacles. Then, the obstacles of two adjacent obstacles are correlated by the feature matching of the obstacle. Finally, the position and orientation of the robot are estimated by optimizing the orientation angle between the sequences. Compared with the ICP (iterative nearest neighbor) algorithm, the verification shows that this algorithm has the same order of accuracy of attitude estimation as ICP, and significantly improves computational efficiency.