Study of virtual balise information fusion method based on attenuation factor

To address the problem that the Kalman filter was prone to filter divergence when using the virtual balise (VB) information fusion, an information fusion method based on the improved Sage-Husa adaptive filtering algorithm was proposed: firstly, the adaptive filtering was used to dynamically adjust the noise statistical characteristics parameters to suppress filter divergence, secondly, an attenuation factor was introduced into the prediction error variance matrix to reduce the influence of stale data and thus improve the filtering accuracy, and finally, simulation experiment was conducted to compare the proposed algorithm with the Kalman filter and Sage-Husa adaptive filter in terms of position and velocity error of the VB. The simulation outcome reveals that the algorithm has an obvious advantage in the positioning error of the VB with better stability in the same time.