自适应高阶容积H_∞滤波算法在目标跟踪中的应用

Application of Self-adaptive High-order Cubature H_∞ Filtering Algorithm in Target Tracking

为了改善传统容积卡尔曼滤波器(CKF)的滤波性能,将高阶容积卡尔曼滤波算法与非线性H∞鲁棒滤波算法相结合,提出一种自适应高阶容积H∞滤波算法(AHCHF)。该算法引入数值稳定性更强的奇异值分解方法(SVD)替换传统的Cholesky分解;同时将H∞鲁棒滤波的思想应用于高阶容积卡尔曼滤波;并基于新息与约束水平反比的关系,对约束水平γ值进行自适应选取,兼顾了滤波精度和系统的鲁棒性。仿真结果表明:相比于传统CKF算法和高阶CKF算法,AHCHF算法具有更高的滤波精度和鲁棒性。

For purpose of improving filter performance of the conventional cubature Kalman filter(CKF),a self-adaptive high-order cubature H∞ filtering algorithm(AHCHF)which having the high-order cubature Kalman filter(HCKF)combined with H∞ robust filtering algorithm was proposed.This algorithm adopts the singular value decomposition(SVD)with stronger numerical stability to replace traditional Cholesky decomposition,including the application of H∞ robust filter to the high-order cubature Kalman filtering.Basing on the inverse proportional relationship established between constraint level and filter information under nonlinear condition,the constraint level γ was selected adaptively and meanwhile,both filtering accuracy and system robustness were considered.The simulation results show that,as compared to the traditional CKF algorithm and the high-order CKF algorithm,this algorithm proposed has better filtering performance.