Performance Comparison for Different Control Systems with Setpoint Changes

Improvement of control system performance is important today. Modern microprocessor hardware allows to implement and put into practice complex control algorithms. In many cases it is necessary to know and periodically renew the control process model. The model is usually obtained during the working process, so the way of obtaining the model must not influence the process significantly. When a control system is designed, it is important in many cases to obtain a high quality of the transient process when the system setpoint is changed. This paper considers this problem for the control systems for thermal processes which often have transport delay. Their properties can be also changed significantly over time. This paper considers a single loop control system equipped with a auto-tuning module, which can renew the process model and tune the controller. This module is based on a fast auto-tuning algorithm, so the influence on the process is small. Several variants of the control system for the case of setpoint changes are considered: a system with a Smith predictor and a PID-controller, a system with a command unit and a PID-controller and a hybrid system with a maximal rapid response algorithm and a PID-controller and a system with a PID-controller and a command unit. All the systems are considered taking into consideration operating limitations such as the limitations of the control action range.