Distributed Estimation for Markov Jump Systems via Diffusion Strategies
Wenling Li, Yingmin Jia · IEEE Transactions on Aerospace and Electronic Systems · 2017
We consider the problem of distributed estimation for Markov jump systems. A distributed interacting multiple model Kalman filter is developed based on the diffusion strategy, where the local measurements, the mode-conditioned estimates, and the likelihoods are exchanged between neighboring nodes. The proposed filter leads to stable estimates for all nodes as long as at least one node is stable in a connected network. Simulation results show that the proposed approach outperforms the existing techniques.