Design of fractional order PID controller based on minimum variance control and application of dynamic data reconciliation for improving control performance

Abstract

Fractional order PID (FOPID) has received much attention in recent years and has been applied in different fields. However, the parameter tuning of FOPID is a tough problem. In order to design an optimal FOPID controller, a parameter tuning method based on minimum variance control (MVC) rule is proposed in both SISO and MIMO control systems Considering the influence of measurement noise with Gaussian and non-Gaussian distribution, which is generally ignored in the controller design based on MVC rule, this paper applies dynamic data reconciliation (DDR) technology to suppress its influence and improve the control performance of the designed FOPID. In the case of considering Gaussian and non-Gaussian distributed measurement noise, SISO and MIMO control systems are employed to verify the effectiveness of the proposed FOPID parameter tuning and DDR method. The results show that the designed FOPID improves the performance of the control system, and the DDR technology suppresses the negative effect of measurement noise and improves the control performance of the designed FOPID.