Multi-model modeling methods based on novel clustering strategy and comparative study: Application to induction machines

This paper is a comparative study of three doubly fed induction motor (DFIM) speed modeling strategies through multi-model approach based on three clustering algorithms; subtractive, C-means and K-means clustering. The comparison leads to a novel clustering strategy compound of the three clustering algorithms. The novel clustering strategy is applied to modeling the speed of the doubly fed induction motor then validated experimentally on a 1kw induction motor. The experimental study is held with the help of MATLAB/SIMULINK and a dSpace system with DS1104 controller board based on digital signal processor (DSP) TMS320F240. Simulation and experimental results approve the efficiency of the proposed approach.