论文部分内容阅读
对于带自回归滑动平均(ARMA)有色观测噪声的多传感器广义离散随机线性系统,应用奇异值分解,提出了广义系统多传感器信息融合状态平滑问题。基于Kalman滤波方法,在线性最小方差信息融合准则下,给出了按矩阵加权融合降阶稳态广义Kalman平滑器。为了计算最优加权,提出了局部平滑误差协方差阵的计算公式。一个Monte Carlo仿真例子说明了所提方法的有效性。
For singular value decomposition of multisensor generalized discrete stochastic linear systems with autoregressive moving average (ARMA) colored observational noises, a generalized system multisensor information fusion state smoothing problem is proposed. Based on the Kalman filter method, the reduced-order steady-state generalized Kalman smoother with matrix-weighted fusion is given under the linear minimum variance information fusion criterion. In order to calculate the optimal weight, the formula of the local smoothing error covariance matrix is proposed. A Monte Carlo simulation example illustrates the effectiveness of the proposed method.