MODIFICATION AND APPLICATION OF AUTOTUNING PID CONTROLLER ∗ NEDJELJKO PERIĆ, IVAN BRANICA, IVAN PETROVIĆ Department of Automatic Control and Computer Engineering in Automation, Faculty of Electrical Engineering and Computing, University of Zagreb, Zagreb, Croatia, nedjeljko.peric@fer.hr, ivan.branica@fer.hr, ivan.petrovic@fer.hr Abstract. This contribution presents a modified autotuning algorithm of the PID controller. The motivation for the modification of the basic autotuning algorithm is to enlarge the class of processes to which it can be applied. The basic autotuning algorithm introduced by Åstrom and Hägglund is extended by the preliminary identification procedure and through the usage of the dead time compensating controller. These modifications are detailed through the description of the algorithms’ functioning. The proposed algorithm has been implemented in the programmable logic controller (PLC) Siemens SIMATIC S7-300. The experimental results confirm the good robustness properties of the proposed algorithm, which were demonstrated in a simulation study. Key Words. Autotuning, PID control, relay experiment, PLC. ∗ Croatian Ministry of Science has supported this work under grant 036-006. Laboratory was equipped in collaboration with Siemens–Croatia and Pliva–Croatia. 1. INTRODUCTION Control systems are designed to minimise the effects of process variations and environmental influences on the quality of process control. Sometimes these variations and influences are significant, so that the conventional linear controllers with constant parameters are unable to control the processes successfully. The main tasks of the controller in such situations are retaining the stability and maintaining the desired performance of the control system. Two control engineering approaches have been devised to achieve these demanding goals: robust control and adaptive control. In the robust control approach a fixed controller is designed, which satisfies the requirements for all models belonging to a certain class. If this class of models encompasses all expectable process and environment variations, the designed controller solves the problem. On the other hand, an adaptive controller adjusts itself to the changes during operation. It recognises variations in the process and in the environment, and adapts the structure and the parameters of the controller accordingly. The adaptation mechanism further automates the control of the process by performing tasks usually made by the control engineer, and thus, extends the idea of the feedback loop. It acts in the adaptation loop of an adaptive control system. Research into control paradigms resulted in the development of the following adaptive control techniques [1]: gain scheduling, model reference adaptive systems (MRAS), self-tuning regulator (STR) and autotuning. As known from the literature [2], the autotuning procedure is performed on the demand of the user, or colloquially after a ‘button push’. Thus, it is not performed continuously in the adaptation loop, but rather when the need for tuning or re-tuning arises. This technique reiterates the design steps which the control engineer performs during the design of the controller. Firstly, a simple experiment is performed which determines some characteristics of the process. After that, using the data obtained, the controller parameters are calculated, and the designed controller Proceedings of the 8th IEEE Mediterranean Conference on Control and Automation (MED 2000) Rio Patras, GREECE 17-19 July 2000