Correlated measurement fusion Kalman estimators and their global optimality
相关观测融合Kalman估值器及其全局最优性

For the multi-sensor linear discrete time-varying stochastic control systems with correlated measurement noises and different measurement matrices, two weighted measurement fusion Kalman estimators are developed by using the weighted least squares (WLS) method. They include the state filtering, state prediction and state smoothing. Based on the Kalman filter in the information filter form, it is proved that under the same initial values, they are numerically identical to the corresponding centralized measurement fusion Kalman estimators, so that they have the global optimality. However, they can obviously reduce the computational burden. A numerical simulation example verifies their functional equivalence to the centralized measurement fusion Kalman estimator.