Automatic tuning of optimal PI controllers

The PI controller is unquestionably the most commonly used control algorithm in the process control industry. PI controllers have traditionally been tuned empirically, e.g. by the method described in Ziegler and Nichols (1942). This method has the great advantage of requiring very little information about the process. There is, however, a significant disadvantage because the method inherently gives very poor damping. Hang et al. (1991) modified the Ziegler-Nichols tuning formula, which improved the performance of PI controllers. In this paper, a method for autotuning optimal PI controllers is proposed. The PI autotuner first introduces relay feedback experiments to identify a second-order plus dead time model, a PI controller is then designed based on the proposed robustness specification and integrated error optimum. Simulation examples show that the proposed PI autotuner gives satisfactory performance for different types of processes.