论文部分内容阅读
提出一种基于特征模型的强跟踪无迹卡尔曼滤波(CSUKF)算法对状态和参数进行联合估计,利用特征模型参数构造时变的二阶状态转移阵,使滤波和辨识模型简化;结合强跟踪滤波(STF)的强跟踪能力和无迹卡尔曼滤波的(UKF)的非线性高逼近性对含测量噪声的高超声速飞行器系统进行参数辨识和滤波,并将其与非线性黄金分割自适应控制律相结合,对高超声速飞行器进行姿态控制.最后,将提出的CSUKF与基于特征模型的无迹卡尔曼滤波(CUKF)和基于特征模型的普通扩展卡尔曼滤波算法(CEKF)进行比较,仿真结果说明CSUKF与非线性黄金分割自适应控制律相结合可以有效改善控制的平稳性,且具有更好的滤波精度和系统输出,从而能更好地处理含测量噪声情况下的高超声速飞行器的辨识与控制问题.
A robust tracing unscented Kalman filter (CSUKF) algorithm based on eigenmodel is proposed to jointly estimate the state and parameters, and the second order state transition matrix is constructed by using the eigenmodel parameters to simplify the filtering and identification models. Combined with strong tracking (STF) and the nonlinear high approximation of unscented Kalman filter (UKF) are used to identify and filter the hypersonic vehicle system with measurement noise, and then, it is compared with the nonlinear golden section adaptive control Then, the proposed CSUKF is compared with the feature-based unscented Kalman filter (CUKF) and the feature-based normal extended Kalman filter (CEKF). The simulation results It shows that CSUKF combined with nonlinear golden section adaptive control law can effectively improve the stability of control and has better filtering accuracy and system output so as to better deal with the identification of hypersonic vehicles with measurement noise Control the problem.