Hybrid artificial neural network for induction motor parameter estimation

Three-phase induction motor electric parameter estimation has been widely used to improve induction motor control performance. A precise match between electrical parameter values and estimated ones is imperative. A value deviation can make induction motor misbehave, which can cause motor overheating even instability. Parameter estimation can be achieved on-line or off-line way with a large number of methods developed to calculate magnetic flux, motor speed, rotor resistance and rotor time constant. These methods include observers, adaptive systems, spectral analysis and artificial intelligence such as neural networks and fuzzy logic. This paper is focused on a hybrid neural network proposed to obtain rotor resistance and speed values, using Texas Instrument development tools to improve a sensorless vector control scheme an improve motor performance.