On-line fuzzy relational identification of a permanent-magnet drive system

Fuzzy relations can model nonlinear dynamic systems performing nonlinear input-output mappings. A new online fuzzy model identification algorithm builds a linguistic model, establishing a time-variant relational matrix which takes into account the input-output relations. The identification algorithm is applied to an electromechanical system consisting of a current-controlled inverter and a permanent-magnet motor. The identified model has input-output variables (stator-reference current and speed) treated as fuzzy sets, whereas the relations existing between them are described by means of a fuzzy relation matrix R defining an online linguistic model of a permanent-magnet drive system