Research on multi-sensor fusion localization algorithm based on EKF
Aiming at the problem that the accuracy of outdoor mobile robot positioning system depends on sensor fusion and has accumulated errors,a multi-sensor fusion method based on extended Kalman filter(EKF)was proposed for outdoor mobile robot positioning.The extended Kalman filter fusion localization of outdoor mobile robot was carried out through Real-Time Kinematic(RTK),odometer information and IMU information.Experiments were carried out in a real outdoor environment.The experimental results showed that the algorithm could eliminate the cumulative error,and the robot's positioning accuracy was significantly improved.The dynamic positioning accuracy could reach within 2.5 cm,compared with odometer IMU fusion localization,the error was reduced by about 92.4%;compared to the traditional RTK algorithm,the positioning accuracy was improved by 55.4%.Multiple experiments show that the algorithm has good robustness.
RTK positioningextended Kalman filtersensor fusionoutdoor mobile robot