Control System with a Predictive PID-Controller with a First-Order Filter: Estimation of the Efficiency for Thermal Proceses

This paper considers the problem of the control system transient process quality improvement, which is very important in thermal engineering even now. The way of improvement under consideration is the use of a predictive control algorithm. The control algorithm described is a PID control law with a first order filter and a linear prediction module; the control system is also equipped with an auto-tuning module, that allows to retune the system if necessary. The auto-tuning module uses a fast auto-tuning algorithm calculating the process model parameters using the process reaction of a rectangular impulse. The PID algorithm coefficients are calculated using indirect optimality indexes. The system works with a typical thermal process with a second order transfer function with a time delay. Calculations and simulation of the system are carried out in Mathcad and Matlab/Simulink. Some examples are considered, the predictive control system performance is estimated and recommendations on the prediction time interval determination are given. These recommendations may be further used while developing this system using modern controllers and controller programming software.