Evaluation And Comparison Of Kalman-Filter-Based Distributed State Estimators For Large-Scale Systems

This paper presents a performance comparison of Kaiman-filter-based distributed state estimations techniques for large-scale systems. Three distributed Kaiman filter schemes are selected to be implemented in simulation, namely, distributed and decentralized Kaiman filter (DDKF), distributed Kaiman filter with state consensus (DKF-SC) and distributed Kaiman filter with diffusion strategy (DKF-DS), using the heat rod dynamical system as benchmark. The heat conduction and convection in a solid rod dynamic is used as test bed. The comparison is performed by evaluating prediction error and convergence of each scheme, considering the partition of the system into two, three and four subsystems.