Enhancing INS/UWB Integrated Position Estimation Using Federated EFIR Filtering

Abstract

In order to obtain accurate position information, a federated extended finite impulse response (EFIR) filter which includes sub-filters and main filter is proposed for inertial navigation system (INS)/ultrawideband (UWB)-integrated human positioning in this paper. Here, the EFIR filter is employed as the sub-filter, which is able to fusion the UWB- and INS-measured distances between the reference nodes and the target human. The main filter is used to provide the optimal INS position error estimation by fusing the outputs of the sub-filters, which is able to correct the INS position. The test results show that the performance of the proposed method is better than the traditional federated extended Kalman filter in position accuracy.

Keywords

• Indoor human positioning
• inertial navigation system (INS)
• ultra wide band (UWB)
• tightly-coupled model
• federated EFIR filtering