Fusing Ultra-wideband Range Measurements with IMU for Mobile Robot Localization

Abstract

Due to the widespread use of robotics in recent years, accurate localization and tracking have become active research topic. As a low-power wireless communication and sensing technology, Ultra-wideband (UWB) has relatively accurate positioning and sensing capabilities, and has broad application prospects for precise positioning and other fields. But due to the complex environment and obstacles, the non-line-of-sight(NLOS) error generated by it will be severe. It seriously affects the position estimation of the system, resulting in low positioning accuracy and poor robustness. Improving the accuracy and robustness of the UWB positioning technology in a complex environment, a method based on the fusion of UWB and IMU data, which effectively combines global positioning and local positioning, positioning, using LSTM neural network algorithm processes the IMU data, and The EKF algorithm merge the IMU and UWB. Compared with the traditional UWB positioning method, this method can effectively suppress Control the influence of NLOS interference in positioning estimation and improve the accuracy and robustness of the positioning system.