基于指数加权的改进衰减记忆自适应滤波算法

Adaptive Fading Memory Kalman Filtering Algorithm Based on Exponential Weighting

体系对抗条件下预警机作用日益突出,但是我方离其距离远,量测数据不准确,对其跟踪困难,针对这个问题提出了一种基于指数加权的改进衰减记忆自适应滤波算法。标准卡尔曼滤波由于模型不准确会造成滤波发散,传统衰减记忆滤波在解决此问题的同时由于其衰减因子为常值会造成目标跟踪精度不足,为此设计了一个指数型的衰减因子对传统算法进行改进,该衰减因子能够自适应计算,并将改进后的方法应用于交互式多模型算法当中。仿真结果表明,该算法可以对预警机目标进行有效跟踪,且与标准卡尔曼滤波和常值衰减因子滤波对比发现,该算法估计误差明显减小,跟踪精度显著提高。

The threat of airborne warning and control system（AWACS）is increasing under systems counterwork condition,and when the target is far away from our devices, that will cause the measurement data inaccuracy and the tracking difficult. To solve this problem, an improved fading memory Kalman filtering algorithm based on exponential weighting was proposed. The result of standard Kalman filtering might be divergent due to impre- cise model, and the traditional fading memory Kalman filtering could solve this problem whereas might bring the tracking accuracy deficiency because of its constant fading factor. Therefore, an exponential fading factor ob- tained from adaptive computation was designed to improve the traditional algorithm, and the improved method was applied in the IMM algorithm. Simulation results showed that IFMKF algorithm could effectively track AWACS and reduce tracking error and remarkably increase tracking accuracy, which were better than SKF and FMKF.