A design method for self-tuning PID controllers

The paper describes a method for designing PID adaptive regulators. Firstly, a simple proof is given for a formula presented by Sebakby and Murad (1992) for computing the quadratic performance indices and their gradients which are needed in any gradient algorithm for updating the regulator parameters. Secondly, the author demonstrates how to apply the proposed method to the design of different types of PID controllers. Two examples are given to show the feasibility of the technique in both adaptive and the nonadaptive setups.