Abstract:

Considering the incomplete observability of each sensor’s measurement in the network, and a distributed state estimation algorithm is proposed based on the measurement compensation strategy moving horizon estimation (MHE). To achieve the completeness of state estimation, each sensor constructs the measurements prediction and sends it to its neighbor sensors for measurement compensation. For the consensus of the distributed estimation, each sensor adopts the MHE approach by using its local and its neighbor sensors’ measurements. Then, the distributed MHE (DMHE) optimization problem and algorithm are presented for each sensor. Simulation results show that by implementing the presented DMHE algorithm, all the incompletely measured sensors can give completed and precise object state estimation in the observable whole sensor network.