Title:Identification of Structured State-Space Models With Unknown Inputs

Speaker:Chengpu Yu

Date&Time:15:00, June 11, 2019

Location: Room C317, School of IoT Engineering

Abstract: The identification of structured state-space system models is quite popular for modeling real physical systems or networked systems. In the presence of unknown excitation or disturbance, the corresponding system identification problem turns out to be challenging and sometimes unidentifiable. In this talk, the identifiability conditions for the structured system models are provided, and a subspace identification method is presented which successively estimates the system states and the structured system matrices. The effectiveness of the identification method is demonstrated through simulation examples.