Research On The Control Methods For Processes With Time Delay

Processes with time delay widely exist in the process control industries. The system analysis and control design become more difficult because of time delay. So, great efforts have been made to develop effective control strategies for the processes with time delay. In this paper, the PID controller, the IMC and the Smith predictor, which have been commonly adopted in practice, are mainly studied. New control methods have been provided based on the existed academic achievements. The main contents of the research include:1. Some of the most used PID tuning methods, such as Z-N, R-Z-N, ?str?m, PM, SPAM and so on, are discussed and their performance is compared by simulations.2. A new method tuning the PID parameters is presented by the normalized dead time of process. The tuning formulae are obtained from the curve fitting in the least-squares sense.3. The relationship between the IMC and the single-loop control is argued in detail and the IMC can be exploited to tune PID controllers in single-loop configuration. Furthermore, the IMC controller can also be viewed as an extension of the Smith predictor.4. The Smith predictor is sensitive to model error and cannot be used for unstable processes or integrating processes since a constant load disturbance will result in a steady-state error. In this paper, analytical approaches and modified methods for Smith predictor are analyzed and summarized.5. A new approach is presented to find the controller parameters for the modified Smith predictor given in Majhi method by IMC. The controller for setpoint tracking is derived by IMC principles and ISE criterion. To make the closed loop system stable under load disturbance, the optimum phase margin criterion or the Nyquist stability criterion is used to design the load-disturbance controller.6. Simulations are included to demonstrate excellent time-domain performance and robustness of the proposed methods compared with some existing ones.