基于激光测距仪全局匹配扫描的SLAM算法研究

针对传统的SLAM(Simultaneous Localization and Mapping)算法构建地图时容易受环境因素和外界条件的的影响，在非线性系统状态下误差修正能力不足，且当机器人位姿都处于未知状态时，移动机器人位姿获取不精确，地图构建SLAM技术特征量的获取比较繁琐、不准确等问题。以电力巡检机器人为平台，研究了基于全局匹配的扫描算法,摒弃传统的栅格地图模型的插值方法，采用双线性滤波的插值方法，保证子栅格单元的精确性，估算栅格占用函数的概率和导数。最后采用此算法解决了SLAM地图构建的问题，并分别在室内室外环境进行实验。实验结果表明：基于激光测距仪的全局匹配扫描的SALM算法，在室内室外两种不同环境下，不受复杂背景的影响，准确地进行机器人位姿定位，以及环境地图的构建

The SLAM Research of Global Matching Scanning Based on Laser Range Finders

Targeting to tackle those problems like tedious characteristic quantity, complex calculation, etc., during environment map and robot pose are unkown state, mobile robot Simultaneous Localization and Mapping (SLAM) technology. Inspection robot for power as a platform, based on matching scanning algorithm are studied, and use this algorithm solve the problems of SLAM. Finally, the experiment is conducted in the indoor and outdoor environment. The experimental results show that the SLAM of global matching scanning based on laser range finders is not affected by the complex background, accurate positioning of the robot position and the construction of the environment map.