PID controller tuning

The article discusses autotuning of controllers and then concentrates on the use of relay autotuning and in particular the errors involved when using the describing function method to analyse the data. To show this, the exact analysis of limit cycles in relay feedback systems is considered. It is also shown how the measured data can be used in different ways for controller tuning and how additional measurements can yield better accuracy. Some limitations of `classical' PID control are illustrated by showing how significantly better performance can be obtained with a PI-PD controller.