自校正解耦融合Kalman预报器及其收敛性

Self-tuning decoupled fusion Kalman predictor and its convergence

对于带有未知模型参数和噪声方差的多传感器系统,通过系统辨识方法,得到模型参数和噪声方差的信息融合估计,将其代入到最优分量按标量加权融合Kalman预报器中,得到自校正信息融合Kalman预报器,实现了状态分量的解耦。通过动态误差系统分析(DESA)方法严格证明了提出的自校正Kalman预报器按一个实现收敛于最优融合Kalman预报器,因此它有渐近最优性。应用信号处理的仿真例子验证了其有效性。

For the multi-sensor system with unknown model parameters and noise variances, the information fusion estimation of model parameters and noise variances can be obtained by using the system identification method, and then is introduced into the optimal fusion Kalman predictor weighted by scalars for components. A self-tuning decoupled fusion Kalrnan predictor is presented which realizes the self-tuning decoupled fusion Kalman filter for the state components. The dynamic error system analysis （DESA） method verifies that the self-tuning fusion Kalman predictor converges to the optimal fusion Kalman predictor in a realization, which has asymptotic optimality. A simulation example applied in signal processing shows its effectiveness.