Distributed State Estimation for Large-Scale Systems with State Equality Constraints

This paper investigates the design of distributed Kalman filter for large-scale systems with state equality constraints in a Gaussian environment. Firstly, we design a unconstrained distributed estimator for a large-scale systems using the local information of different subsystems. The optimal gain is obtained based on the minimum mean square error estimation criterion. Secondly, we study distributed Kalman filter for large-scale systems that integrate equality constraints. In this approach, the solution from the unconstrained Kalman filter is projected onto the state-constrained surface at each time step to enhance the prediction accuracy of the filter in large-scale systems. Finally, the simulation experiments demonstrate that the effectiveness of the distributed Kalman filter for large-scale systems with state equality constraints outperforms unconstrained Kalman filter.