Distributed Fusion Estimation Based on Pseudo-measurement for Multi-sensor System

Abstract

By considering the relation between local fusion estimation and fusion center, in this paper the estimation from local fusion nodes is regarded as a pseudo-measurement. Then the distributed estimation algorithm is turned to be two-level centralized fusion estimation and the new optimal distributed fusion estimation algorithm is obtained with Kalman filtering form, which in general only centralized estimation method has. Simulations show the developed algorithm has the excellent estimation performance. By the developed algorithm, the distributed multisensor system can be unified with centralized system and make it possible that applying the abundant research result of centralized system to distributed multisensor system.