摘要：AbstractIn this paper, we present a distributed approach to simultaneously estimate state and parameter for a class of nonlinear systems, for which the augmented models comprising both the states and parameters are not fully observable. Specifically, we first discuss how sensitivity analysis (SA) can be used to select the best subset of states and parameters for estimation. Then, the entire system is decomposed into several interconnected subsystems as the basis for distributed estimation. Subsequently, a local moving horizon estimator (MHE) is developed based on the corresponding subsystem model, and the local estimators communicate with each other to exchange their estimates. Finally, an SA-based distributed MHE scheme is proposed. The effectiveness of the proposed approach is illustrated using a chemical process consisting of four connected reactors.
关键词：KeywordsSubsystem decompositionsensitivity analysisdistributed estimationmoving horizon estimationnonlinear systems
