A Novel Maneuvering Target Tracking Algorithm Using Polynomial Filter

Abstract

In order to improve the performance of maneuvering target tracking in complex observation scenarios, a composite expanding memory polynomial (EMP)/fading memory polynomial (FMP) filtering algorithm based on the polynomial filter is proposed. According to the analysis of the advantages and disadvantages of four forms of polynomial filters, the composite EMP/FMP filter is constructed by combining the self-initialization of EMP filter with the indefinite tracking of FMP filter. Additionally, there is no transient phenomenon when switching two filters. Then, the filter which applies the EMP/FMP filter as the pre-filter, and the unscented Kalman filter (UKF) as the main filter is simulated. The experimental results indicate that the EMP/FMP pre-filter enhances the quality of measurements, and effectively improves the target tracking performance of UKF for maneuvering target.