PID contoller autotuning based on nonlinear tuning rules

Controller tuning is one of the most important tasks of control loop design. Various control algorithms have the tuning mechanism built-in themselves. However, the most of used controllers are controllers of the PID type that contains no selftuning features; it is necessary to tune the controller manually or the selftuning features need to be added externally. Despite of the fact that many PID controller tuning algorithms have been developed since Ziegler and Nichols tuning procedures were published in 1942 their experimental method remains one of the really used methods up to now. It has several advantages in comparison with the newer tuning algorithms: it does not need any mathematical model of controlled plant and it is usable without any deep control theory background. One of the approaches trying to suit these practical limits is a newly developed method presented in this paper. It is based on an experimentally performed evaluation of excited frequency responses with the aim to achieve recommended values of one or more control quality indicators known from the course of the Nyquist plot, e.g. With regard to the experimental way of obtaining the indicators, they can be evaluated in control loops involving nonlinearities even in the controller. In this sense, the method has a philosophy similar to that of the popular Ziegler and Nichols method. The main features making model-free design attractive for practice are preserved: no mathematical model and theory is required for in controller setting, all processing is carried out by a program added to the control algorithm, i.e. no additional instrumentation is necessary.