Application of a Simplified Inverse Fuzzy Model for an Induction Motor Drive Control

Abstract

Even, after the adoption of some simplifying assumptions, the three-phase induction motor presents a complex nonlinear dynamic system of the 5th order with multiple inputs. Modern methods of artificial intelligence (fuzzy approach, neural networks) make it possible to find its much simpler and practically usable models based on the description of its important qualitative properties. The presented article deals with the construction of a simplified inverse fuzzy model of an induction motor using an estimation of critical parameters of the Kloss relation, which can be utilized for its control. To build such model, it is necessary to perform only several standard measurements – step responses on the electrical motor. No priori knowledge of motor parameters is required. The obtained simple fuzzy model of the induction motor (IM) is then used for linearization of its torque loop and for design of superior control loops (speed, position). The presented procedure also eliminates the need for transformations of motor quantities, as is usual in vector control.