Wi-Fi RTT/Encoder/INS-Based Robot Indoor Localization Using Smartphones
Baoding Zhou, Zhiqian Wu, Zhipeng Chen, Xu Liu, Qingquan Li
Abstract
With the rapid development of robotics, miniaturized and inexpensive robots have gradually entered the public's field of vision. Smartphone-based robots have the potential to provide high-precision indoor localization, and they are easy to promote and apply. In this paper, we propose a tight integrated positioning system based on Wi-Fi round trip time (RTT), encoders, and the inertial measurement unit (IMU) for robot indoor localization using smartphones. In our approach, we first design an error-state Kalman Filter (ESKF) to fuse the inertial information from the IMU with the measurements from the encoders to suppress the errors accumulated by the inertial navigation system (INS). Second, the Wi-Fi RTT-based localization method is implemented through an adaptive extended Kalman filter (AEKF) to fuse the ranging information. Finally, to overcome the shortcomings of the long-time drift of the INS and the instability of the Wi-Fi RTT-based localization system, we implement a tight integrated positioning system, which obtains the optimal estimation of the INS positioning error by using the Wi-Fi RTT ranging values as filtering constraints and smooths the INS positioning error through the Rauch-Tung-Striebel (RTS) algorithm. Experimental results show that compared with the Wi-Fi RTT-based method under line of sight condition (LOS) and non-line of sight condition (NLOS), the mean positioning error of the proposed method is improved by 54.62% and 58.38%, respectively, while compared with the INS/Encoder method, those is improved by 57.48% and 33.04%, respectively.