Improving lambda tuning of PI controllers for load disturbance rejection

In this paper we propose a tuning rule for PI controllers that is based on the same rationale of lambda-tuning but aims to optimize the performance in the load disturbance rejection task. In particular, the maximum sensitivity obtained by lambda-tuning is used as a constraint in minimizing the integral of the absolute error, in such away to achieve the desired tradeoff between performance and robustness. The tuning rule has been obtained by finding the optimal parameters for different processes by means of genetic algorithms and then by fitting the results with appropriate functions.