Closed-loop identification of plant and disturbance models based on generalized minimum variance evaluation

A closed-loop identification method based on generalized minimum variance (GMV) evaluation is proposed. Parameters of plant and disturbance models can be estimated simultaneously based on the data-driven control technique that minimizes the variance of the generalized output. This feature distinguishes the proposed method from conventional closed-loop identification methods. In particular, an advantage of the proposed method is ease of use like data-driven controller design methods. For the purpose of identification, we introduce a new variance criterion, which is derived from a set of input-output data generated by stochastic disturbance. The paper proves that the optimization of the proposed criterion results in the unique optimal solution which corresponds to the true plant and disturbance model parameters. The proposed method is applied to datasets obtained from a continuous stirred tank reactor (CSTR), which is operated around an unstable steady state. The result illustrates the effectiveness of the proposed closed-loop identification method.