Optimized fuzzy logic controller parameters for open-loop stable or evolutive electromechanical systems

This paper describes a new and original tuning method of a fuzzy PID-like controller which is just a particular case of nonlinear controllers. As Ziegler-Nichols or Broida methods for the tuning of conventional controllers, this method is based on only one experiment: the open-loop step response of the process (which is supposed to be open-loop stable or evolutive). In this way, this method can be seen as a very simple on-site industrial tuning methodology. In this paper, we provide pre-established settings for the fuzzy logic controller parameters in terms of parameters that have been identified from the open-loop experiment. These settings have been obtained using the experimental designs methodology and a simplex algorithm. They are optimized considering an IAE-criterion on a very common industrial benchmark. The pre-established parameters are then validated on two experimental systems: a DC/DC buck converter and a permanent magnet synchronous machine