Bayesian Estimator for High-Dimensional Systems With Block-Dependent Compensation

This article proposes a Bayesian estimation scheme with block-dependent compensation for linear high-dimensional systems. The state vector is divided into blocks by considering the relationship of state variables in time series and physical interpretation to reduce the underlying computational cost. However, blocking states damages the original system and introduces systematic errors. Consequently, the divided state blocks are compensated by introducing distinct compensation coefficients to improve state estimation accuracy. Considering the difference and uniqueness of state blocks, the distinct compensation coefficients are adaptively adjusted and updated over time to ensure higher accuracy of state estimation. Simulation results of numerical and practice-oriented examples are used to evaluate the performance of the proposed algorithm. It shows that the proposed algorithm possesses moderate computational cost and higher estimation accuracy compared with the existing methods.