Flexible Fuzzy PID Controller (FFPIDC) and a Nature-Inspired Method for Its Construction
Krystian Łapa, Krzysztof Cpałka · IEEE Transactions on Industrial Informatics · 2017
In this paper, a new approach to fuzzy controller construction which combines features of proportional-integral-derivative controllers and flexible fuzzy systems is proposed. Flexibility of the fuzzy systems used makes it possible to interpret a control process and reduce the controller structure, which can make hardware implementation easier. The proposed approach uses knowledge about the controlled object and possibilities offered by nature-inspired hybrid population-based algorithms. These algorithms allow us to simultaneously select controller structure and parameters while taking into account a set of evaluation criteria. The criteria can include the quality of control, controller complexity, control process interpretability, etc. This is important in those applications where an atypical controller structure needs to be selected or an extended set of expectations has to be taken into account. This approach was tested using typical control problems.