Application of an Improved and Unscented Kalman Filtering Algorithm in Target Tracking

Abstract

The paper introduces and adopts an improved and unscented kalman filtering algorithm to track moving targets. For issues like large calculation amount, unfavorable real-time performance and non-local effects of samples of this algorithm, scale factors are adaptive selected to simplex sampling with minimum skewness. According to simulation results, on one hand, introduction of the algorithm can reduce calculation amount and increase arithmetic speed. On the other hand, it can decrease errors of non-local effects and high order, and enhance the accuracy of target tracking.