基于扩展卡尔曼滤波的传感器融合SLAM研究  

Research on sensor fusion SLAM based on extended Kalman filter

作　　者：杜俊峰 郁汉琪[1] 刘义亭 张昊[1] Du Junfeng;Yu Hanqi;Liu Yiting;Zhang Hao(Nanjing Institute of Technology,Nanjing 211167,China)

机构地区：[1]南京工程学院,江苏南京211167

出　　处：《无线互联科技》2023年第21期133-137,共5页Wireless Internet Technology

基　　金：2021年度江苏省重点研发计划(产业前瞻与关键核心技术),项目编号:BE2021016-5。

摘　　要：同时定位与地图构建(Simultaneous Localization And Mapping,SLAM)技术是当今自主导航机器人的关键部分,移动机器人的里程计误差会直接降低SLAM效果,造成机器人定位丢失和地图不匹配问题。文章基于里程计和惯性测量单元(Inertial Measurement Unit,IMU)传感器融合,提出同时使用线性变换函数和扩展卡尔曼滤波(Extended Kalman Filter,EKF)减小里程计误差的方法。实验结果表明,里程计误差最大降低了91.72%,机器人SLAM建图稳定无漂移,能有效提高SLAM的准确性和鲁棒性。Simultaneous Localization And Mapping(SLAM)technology is a key part of today’s autonomous navigation robots.The odometer error of mobile robots will directly reduce the SLAM effect,resulting in robot positioning loss and map mismatch.Based on the fusion of odometer and Inertial Measurement Unit(IMU)sensors,this paper proposes a method that uses linear transform function and Extended Kalman Filter(EKF)to simultaneously reduce odometer errors.The experimental results show that the maximum error of the odometer is reduced by 91.72%,and the robot SLAM is stable without drift,which can effectively improve the accuracy and robustness of SLAM.

关 键 词：SLAM 传感器融合 线性变换 扩展卡尔曼滤波 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]