An improved positioning method based on compensation and optimization of ultra-wideband ranging results

Abstract

Abstract Achieving precision in positioning under conditions of significant interference remains an unresolved challenge in research. This study introduces a low-cost ultra-wideband (UWB) distance compensation model that addresses electromagnetic wave loss in practical indoor settings. This paper employs kurtosis to detect non-line-of-sight environments, which are frequently induced by pedestrian movement. The Generalized Extreme Studentized Deviate algorithm is utilized to discern and eliminate outliers in ranging values and the Piecewise Cubic Hermite Interpolating Polynomial algorithm compensates for the eliminated data points. Finally, Kalman filtering is used to improve UWB ranging results, allowing for better error elimination and compensation. Experimental results demonstrate that our proposed algorithm has higher accuracy and the mean square error improvement ratio can reach more than 20% in dynamic positioning tests.