基于激光EKF-SLAM的车厢内定位方法  

Car Positioning Based on Lidar of Extended Kalman Filter(EKF) Algorithm and Simultaneous Localization and Mapping(SLAM)

作　　者：蔡杨华 谭金 王冠 

机构地区：[1]广东电网有限责任公司电力科学研究院广东电科院能源技术有限责任公司

出　　处：《工业控制计算机》2019年第12期12-14,17,共4页Industrial Control Computer

基　　金：省级计量检定中心无人化、智能化技术研究及应用-课题1：计量检定中心库前区无人化运行关键技术研究(GDKJXM20172799)

摘　　要：针对传统物流行业中工作强度大、效率低的弊端,使用自动移动机器人按对车厢内的货物进行转移,需要对车厢构图,并对移动机器人进行车厢内精确定位。使用扩展卡尔曼滤波(EKF)同时定位与构图(SLAM),借助使用车轮驱动编码器信号,建立状态模型,获取移动机器人的里程计信息,使用激光雷达获取环境点云信息,建立观测模型;通过提取直线特征,利用状态模型预测位姿完成观测模型同地图中直线特征相互匹配,进而通过卡尔曼滤波算法完成对实际位姿的估计过程。所提出定位方法可以实现在室内、车厢等有限环境中定位,平均定位精度达2cm,姿态角度误差小于1°。In view of the drawbacks in the traditional logistics industry,the use of automated mobile robots to transfer goods in the car requires the composition of the car and precise positioning of the mobile robot.In this paper,the simultaneous positioning and composition using extended Kalman filter is used to establish the state model by using the encoder placed at the wheel,obtain the odometer information of the mobile robot,and use the laser radar to obtain the surrounding environment point cloud information.The observation model is established.By extracting the linear features,the straight line matching process in the observation model is completed by using the predicted pose in the state model,and the estimation process of the actual pose is completed by using the Kalman filter algorithm(EKF).Using this method,the average positioning accuracy can be up to 5cm in a limited environment such as indoors and cars,and the attitude error is less than 1°,which effectively solves the positioning problem in the cabin.

关 键 词：激光雷达 扩展卡尔曼滤波 定位与构图 直线特征 

分 类 号：TP2[自动化与计算机技术—检测技术与自动化装置]