Control of a Class of Industrial Processes With Time Delay Based on a Modified Uncertainty and Disturbance Estimator

Abstract

The control algorithm based on the uncertainty and disturbance estimator (UDE) is considered as an efficient control strategy that has attracted much attention recently. However, it cannot be directly applied to the widely existing industrial processes with time delay. To extend its applicability, the UDE control algorithm is modified so that it can be applied to stable, integrating, and unstable processes with time delay. The stability condition of the modified UDE (MUDE) is established. It is shown that the design of tracking and regulation can be separated. The tradeoff between the performance and robustness is achieved by parameter tuning. Compared with the recent strategies evolving from the Smith predictor, simulation, and laboratory experimental results are presented to show that the MUDE-based control gives significant improvement in 1) robustness; 2) control for unstable systems; 3) eliminating steady-state error; and 4) structure simplicity, which makes it easy to configure in industrial distributed control systems. Finally, the promising prospect of the proposed strategy in industry is exemplified via a field test in an in-service 1000-MW power plant.