基于非线性多传感器复合定位融合算法应用研究  

Research on Application of Nonlinear Multi-sensor Compound Positioning Fusion Algorithm

作　　者：李航 黄波 廖映华 LI Hang;HUANG Bo;LIAO Yinghua(School of Mechanical Engineering,Sichuan University of Science and Engineering,Yibin Sichuan 644000,China)

机构地区：[1]四川轻化工大学机械工程学院,四川宜宾644000

出　　处：《机床与液压》2023年第5期41-47,共7页Machine Tool & Hydraulics

基　　金：四川省科技厅重点研发项目(2021YFG0056);过程装备与控制工程四川省高校重点实验室项目(GK202203)。

摘　　要：针对移动机器人单传感器数据短时丢失、定位精度低、传感器频率异步等问题,采用激光雷达、IMU、轮式里程计获取定位信息,提出基于扩展卡尔曼滤波和互补融合的组合数据融合方法。先通过S-G滤波算法对初始定位数据进行预处理,利用扩展卡尔曼滤波融合算法实现IMU和轮式里程计传感器的定位数据融合,得到融合数据1;再利用互补融合算法将融合数据1和激光雷达进行融合得到融合定位数据2。其中融合数据1对激光雷达进行实时补正,解决频率异步的位移偏差,从而明显提高定位精度。最后采用Gazebo仿真平台,搭建移动机器人模型以及设置传感器的基本参数,验证算法的有效性和稳定性。实验结果表明:数据融合算法提高了非线性传感器的定位精度和稳定性,并且平均定位误差在8 cm内。Aiming at the problems of single sensor data loss,low positioning accuracy and asynchronous sensor frequency of mobile robot,lidar,IMU and wheeled odometer were used to obtain positioning information,and a combined data fusion method based on extended Kalman filter and complementary fusion was proposed.The initial positioning data were preprocessed by S-G filtering algorithm,and the extended Kalman filter fusion algorithm was used to fuse the positioning data of IMU and wheel odometer sensor to obtain the fused data 1.Then,the fused data 1 and lidar data were fused by complementary fusion algorithm to obtain the fused positioning data 2.The fused data 1 was used to correct the lidar in real time,solving the displacement deviation caused by asynchronous frequency,so as to significantly improve the positioning accuracy.Finally,the Gazebo simulation platform was used to build the mobile robot model and set the basic parameters of the sensor to verify the effectiveness and stability of the algorithm.The experimental results show that the data fusion algorithm improves the positioning accuracy and stability of nonlinear sensors,and the average positioning error is within 8 cm.

关 键 词：多传感器 数据融合 扩展卡尔曼滤波 互补融合 定位精度 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]