Iterative Feedback Tuning for regulatory control systems with measurable disturbances

Abstract

Abstract The study proposes Iterative Feedback Tuning (IFT) with measurable disturbances for disturbance attenuation using regulatory control data. The control system has a feedforward controller from the measurable disturbances as well as a feedback controller. The variance of the closed-loop output is used for the cost criterion. The gradient of the cost criterion in terms of both feedback and feedforward controller parameters are estimated via spectral analysis for single set of regulatory control data. The proposed method has a feature that it makes full use of measurable disturbances. In addition, it has a feature that it does not require a specific experiment for estimating the gradient of the cost criterion. These features facilitate the application of IFT to industrial processes. That is because they usually use an optimal set point value from the aspect of economical evaluation, and require to keep the set point a constant value. A numerical simulation shows the proposed method effectively tune both feedback and feedforward controller parameters.