Evaluation of the applicability of state estimation filters for indoor robot positioning

Abstract

Indoor robot positioning plays an important role in ensuring autonomous navigation in smart environments where Global Positioning System (GPS) signals are unavailable. Ultra-Wideband (UWB) technology offers high accuracy and multipath interference resistance, providing a suitable foundation for indoor positioning systems. Therefore, this study focuses on evaluating the performance of three state estimation algorithms---the Extended Kalman Filter (EKF), the Unscented Kalman Filter (UKF), and the Particle Filter (PF)---in handling nonlinear models of positioning systems based on UWB.The results show that the EKF has advantages in processing speed and low computational cost, making it suitable for real-time systems, though it remains sensitive to noise and highly nonlinear models. The UKF demonstrates a balance between estimation accuracy and computational efficiency due to its ability to provide reliable estimates under nonlinear conditions without increasing computational cost. The PF offers high stability and the ability to handle non-Gaussian noise; however, it requires significant computational resources, especially when the number of particles is large.