Dynamic model of Torus motor drive using ANN and FSE methods for the back EMF estimation

In this article, two dynamic models of the Torus machine are discussed. The first model uses the Fourier series expansion for the back emf and is implemented with PSpice. The second model uses the neural network estimation of the back emf and is implemented with Matlab/Simulink. The models may be used to study the steady-state as well as the transient performances of the machine operating as a motor or as a generator. Experimental results are compared with Matlab/Simulink and PSpice predictions for a number of operating conditions, including the complete starting transient. Excellent correlation between the experimental and simulated results is obtained which validates the Torus machine modeling approach.