Plug and play partition-based state estimation based on Kalman filter

In the last years, the distributed state estimation issue has gained great importance in the framework of distributed monitoring and control of large-scale systems. It consists of estimating, through a network of sensors endowed with computational capabilities, the state of a large scale system, characterized by the interconnection of a number of subsystems. In this paper we focus on partition-based distributed estimation and we propose a novel scheme for non-overlapping subsystems based on Kalman filter. The online implementation of the proposed estimation scheme is scalable, as far as both the computation requirements and communication effort are concerned, and convergence results are provided. A simulation example is finally shown, to test the performance of the distributed Kalman filter proposed.