基于扩展信息滤波的多机器人SLAM研究  

Multi-robot SLAM Based on Extended Information Filtering

作　　者：徐斌斌[1] 石守东[1] 王小波 

机构地区：[1]宁波大学信息科学与工程学院,浙江宁波315211 [2]宁波市光启通信工程有限公司,浙江宁波315030

出　　处：《宁波大学学报（理工版）》2011年第1期38-41,共4页Journal of Ningbo University：Natural Science and Engineering Edition

基　　金：宁波大学大学生科技创新科研项目(2010SRT)

摘　　要：运用信息滤波算法实现多机器人SLAM,根据对机器人运动结构和刚体约束建立多机器人运动模型,用里程计和激光测距传感器建立多机器人观测模型,并运用SLAM方法建立全局地图,不断更新该地图的同时完成校正多机器人位姿.根据理论推导,利用信息滤波算法解决多机器人SLAM问题,仿真实验也结果表明多机器人定位精度良好.Robot Simultaneous Localization and Mapping （SLAM） are two key issues in robot autonomous navigation. In recent years, with the development of robotics, the investigation of Multi-robot Simultaneous Localization and Mapping becomes the central areas of SLAM. To achieve Multi-robot SLAM, the Extended Information Filtering algorithm is applied to multi-robot system. Multi-robot motion models are established according to the robot motion structure and rigid body constraints, and observation models are constructed with the odometers and laser ranging sensors. Based on the results of theoretical analysis, the global map is utilized by SLAM to correct multi-robot pose in period of updating map. The Extended Information Filtering algorithm is used to solve the Multi-robot Simultaneous Localization and Mapping problems in the process. The simulation results are suggestive of the sufficiently good robotic positioning and performance robustness in the SLAM.

关 键 词：同时定位和地图创建 系统模型 多机器人扩展信息滤波 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]