Enhanced 2-DOF PID Controller Tuning Based on an Uncertainty and Disturbance Estimator With Experimental Validation

Vicente Balaguer,Pedro Garcia,Antonio Gonzalez,Francisco Blanes

doi:10.1109/access.2021.3095057

Vicente Balaguer, Pedro Garcia + Show 2 more

Open Access

https://doi.org/10.1109/access.2021.3095057

Copy DOI

Abstract

In this paper a novel tuning procedure for Two-Degree-of-Freedom (2-DOF) PID controllers is proposed. The tuning methodology is based on an Uncertainty and Disturbance Estimator. As a result, an equivalent 2-DOF PID controller with simpler tuning rules is obtained. One advantage of the proposed method is that tracking performance and disturbance rejection can be accommodated separately in control synthesis. Moreover, the trade-off between disturbance rejection and robustness can easily be adjusted by tuning a single parameter without affecting the desired set-point tracking specifications. Numerical comparisons with well-known tuning methods are performed to illustrate the advantages of the proposed method. Finally, the effectiveness of the proposal is experimentally validated in an open-loop unstable plant consisting of a test-bed quadrotor platform.

Highlights

Proportional-Integrative-Derivative (PID) control is still one of the most widely used in many industrial processes owing to its simplicity and robustness [1]–[3]
A single-parameter 2-DOF PID tuning approach based on an Uncertainty and Disturbance Estimator (UDE) has been proposed
The key aspect of the proposed method is that the 2-DOF PID parameters can be obtained from the desired set-point settling time and tuning a single parameter, which can be obtained from the disturbance rejection index IAEd or a certain robustness margin for time-varying parametric uncertainties by means of Linear Matrix Inequalities (LMIs), without needing any complex numerical optimization

Summary

INTRODUCTION

Proportional-Integrative-Derivative (PID) control is still one of the most widely used in many industrial processes owing to its simplicity and robustness [1]–[3]. By means of One-Degree-of-Freedom (1-DOF) PID structure, set-point tracking performance and disturbance rejection requirements cannot generally be simultaneously achieved for control applications [17]. Other popular control design approaches such as the AMIGO method [25], the internal model control (IMC) [26], [27] and the SIMC method [28] offer a simple guide for tuning the PID control parameters Such methods might lead to poor performance when they are applied to integrating, unstable and non-minimum phase processes if the tuning is not properly addressed [29]. The proposed method can be extended to higher-order systems by reducing them to a lower-order, and further treating the unknown behavior as a disturbance This approach, initially raised in [43], makes it possible to obtain a 2-DOF PID with only the set-point requirement and a single-parameter to deal with the compromise between disturbance rejection performance and robustness.

PROBLEM STATEMENT

TRADE-OFF BETWEEN DISTURBANCE REJECTION AND

EXPERIMENTAL RESULTS

CONCLUSION AND PERSPECTIVES