容积Rauch-Tung-Striebel平滑器

针对离散非线性系统的状态平滑问题,基于Rauch-Tung-Striebel(RTS)理论设计了一种容积卡尔曼平滑器(Cubature Kalman Smoother,CKS),即容积Rauch-Tung-Striebel平滑器(RTSCKS)。首先,基于经典贝叶斯状态估计理论框架,推导了状态概率密度分布形式的非线性系统最优平滑算法;其次,基于Rauch-Tung-Striebel理论,建立了相应的最优平滑递推算法;然后,将其与容积卡尔曼滤波算法相结合,建立了递推形式的RTS-CKS平滑器;最后,通过典型的纯方位跟踪模型验证了该平滑器的可行性和有效性。该平滑器为非线性系统的状态估计提供了新的估计算法。

A cubature Rauch-Tung-Striebel smoother

In view of the state smoothing problem of nonlinear discrete-time system,a cubature Kalman smoother is derived based on the Rauch-Tung-Striebel theory,namely,the cubature Rauch-Tung-Striebel smoother(RTS-CKS).Firstly,based on the classical Bayesian state estimation framework,the optimal smoothing algorithm of the nonlinear system is derived under the state probability density distribution form.Secondly,the corresponding optimal smoothing recursion algorithm is established based on the Rauch-Tung-Striebel theory.Then,the recursion type form of RTS-CKS smoother is derived through the combinations of the cubature Kalman filter and the optimal smoothing recursion algorithm above.Finally,the simulation shows the feasibility and effectiveness of the proposed smoother through classical bearings only tracking model.The proposed smoother provides a novel estimation algorithm for state estimation of nonlinear system.