Distributed Identification of Large-Scale MISO Output-Error Systems

Abstract

This paper presents a distributed identification algorithm for large-scale MISO output-error systems. The considered MISO system is decomposed into multiple SISO output-error systems with their outputs being summed up to the measurable system output, and each SISO subsystem performs recursive identification by exchanging parameter estimates with other subsystems. Using the Lyapunov theory, it is proven that the proposed distributed identification algorithm converges asymptotically. Finally, a simulation example is given to demonstrate the effectiveness of the proposed algorithm.