Chapter 4 - Multivariable time-delay system identification

Abstract

This chapter first considers the identification problem of the state space model with d-step delay for multivariable systems and presents a state estimation-based recursive least squares parameter identification algorithm by using the hierarchical identification principle. Combining the linear transformation and the property of the shift operator, a state space system is transformed into an equivalent canonical state space model, and its identification model is derived. Then, it focuses on presenting a new identification algorithm to estimate the parameters and state variables for two-input two-output systems with time delay based on canonical state space models. The related input–output equation is obtained and transferred into an identification model, which does not involve the unmeasurable states, and then, a residual-based least squares identification algorithm is given. After the parameters being estimated, the system states are estimated by using the estimated parameters.