Research On Mobile Robot Positioning System Based On Multi-sensor Fusion

In recent years,with the rapid development of artificial intelligence,intelligent mobile robots have been widely used in various industries,especially in traditional manufacturing,logistics and services.Localization is one of the core technologies for mobile robot to achieve autonomy.The accuracy and reliability of localization determine the work of mobile robot.Therefore,aiming at the problems existing in the positioning of mobile robot with single sensor,a mobile robot positioning system based on multi-sensor fusion is proposed to improve the reliability and accuracy of mobile robot in the working process.UWB positioning technology has high positioning accuracy,but when the UWB signal is blocked,the positioning accuracy is greatly affected.The strapdown inertial navigation system based on IMU does not need to rely on external information in normal operation and is not easy to be affected and disturbed by the environment.Although the positioning accuracy is high in the short term,there is error accumulation and the positioning effect is not good in the long term.In this paper,a positioning method combining UWB and IMU is proposed to provide high-precision and reliable position information for mobile robot.The main research contents of this paper are as follows:(1)According to the principle of strapdown inertial navigation,the position and attitude solution in inertial coordinate system is realized.According to the error model of IMU,IMU is calibrated to obtain the Allan variance diagram of accelerometer and gyroscope.By analyzing the Allan variance diagram,the values of Gaussian white noise and bias random walk are obtained.Through the comparative analysis of path and trajectory,it is verified that IMU positioning error will accumulate with time.(2)Several positioning principles and methods of UWB are studied.Through comparison,tw-tof positioning algorithm is selected,and the causes of positioning error caused by this positioning method are analyzed;This paper introduces the establishment and configuration of UWB positioning system base station,carries out the static positioning experiment of UWB positioning system under line of sight and non line of sight,and designs the data processing algorithm using MATLAB.Through the filtered experimental results,it is concluded that the positioning error of UWB under line of sight is increased from the initial10 cm to 2cm,the positioning error under non line of sight is increased from 35 cm to 8cm,and the positioning accuracy is significantly improved.(3)The principle of multi-sensor data fusion is introduced,and the localization method0 f IMU assisted UWB fusion is proposed.The extended Kalman filter fusion algorithm is designed using MATLAB.Through the comparative analysis of experimental motion trajectory,the feasibility of the fusion algorithm is verified and the positioning accuracy is improved.(4)The experimental platform is built,the motion trajectory is drawn through the experimental data,and the trajectory map similar to the grid is drawn.Quantitative analysis of its trajectory shows that the positioning error of UWB / IMU combined positioning system is12.4cm and that of UWB alone is 21.3cm.