A Navigation and Positioning Method for Indoor Mobile Robots

Abstract

For the precise location information in mobile robot indoor scenarios, we design a method based on graph optimization to realize the combination of UWB/IMU/ODOM navigation and positioning system. first of all, using EKF to IMU/ODOM data fusion to get a current estimates of the robot pose, and then use the values of UWB data validity check, build residual, figure optimization calculation, get the optimized trajectory. Experimental results show that the average positioning accuracy of UWB/IMU fusion based on EKF is about 0.18m and the maximum positioning error is less than 0.5m, while the average positioning accuracy of UWB/IMU fusion based on graph optimization is less than 0.07m under the line-of-sight condition. Under the condition of non-line-of-sight or data loss, the UWB/IMU fusion localization method based on graph optimization can keep the robot trajectory tracking for quite a long time until the signal is recovered.