中文摘要：

扫描匹配算法被广泛应用于基于视觉、声纳、激光等传感器数据的特征匹配中，其中迭代最近点扫描匹配算法（ICP）是最常见的扫描匹配算法，但该算法存在匹配误差较大、对角度误差修正较差等缺点；针对基于ICP的激光传感器数据配准中存在的问题，提出了一种遗传迭代最近点扫描匹配算法（GICP）；通过遗传算法搜索当前扫描数据和参考扫描数据的最优匹配，修正初始里程计读数的误差以及机器人的位姿；实验结果表明，提出的算法能够有效地解决扫描匹配算法中任意的配准问题，提高了机器人的定位精度。

英文摘要：

Scan matching algorithms are widely used in solving sensor based feature matching problem. Iterative Closest Point （ICP） is one of the most popular scan matching algorithms. However, such algorithm fails to match accurately, especially when the orientation between the reference scan and the current scan is large. In this paper, Genetic Iterative Closest Point （GICP） is proposed to solve this problem, which integrates Genetic Algorithm into ICP. GICP can correct the odometry error and further the estimation of robot pose by finding the best scan match. It is implemented by replacing the least squares method in the ICP with genetic evolution process. Experimental results demonstrate that the new algorithm can cope with various scan matching situations and produce accurate localization results.