Automatic Input/output Modeling of a Squirrel-cage Induction Motor Drive System Using Neural Network

This paper presents an alternative input/output modelling of a squirrel-cage induction motor drive system. Usually these systems are modelled using state variables like rotor currents or linking fluxes, some of them extremely difficult to obtain. A neural network based learning through examples algorithm is used to represent the motor dynamics, using only input/output information. This is very important when implementing modern control techniques which require a precise modelling of the drive. The learning algorithm performance is verified through experimental results.