拖曳阵阵形估计的自适应Kalman滤波算法

柔性拖曳阵在水下拖动时受拖船拖动及海流等的扰动,因此拖曳阵的阵形估计问题是个存在未知输入的系统状态估计问题。文中采用了一个自适应的KALMAN滤波算法来解决这一问题。自适应Kalman滤波器包括两部分:一部分是没有输入的Kalman滤波,另一部分是自适应加权的Kalman滤波用于估计快时变的余量偏差。在迭代的每一步,均利用M-估计器和Huber函数相结合构造作为更新偏差函数的遗忘因子。数值仿真与海试结果表明,该方法比传统的状态估计方法估计效果好。

Towed array shape estimation using adaptive Kalman filters

The dynamical behavior of a thin flexible array towed through the water is perturbed by the towing ship and ocean flow.Estimation of array shape is to estimate the unknown input of the system.The input estimation approach includes two parts: a Kalman filter without an input term.A recursive least squares estimator weighted by adaptive forgetting factor is proposed to robustly extract the unknowns.The M-estimator combined with the Huber function is used to con-struct the weighting forgetting factor as a reciprocal function of biased innovation at each time step.Numerical experiments and sea trial data verify the proposed algorithm,and are shown to be an improvement over the conventional input estimation.