Tracking inherent saliencies of standard induction machines for zero speed sensorless control using different signal processing methods

For mechanical sensor-less control of inverter fed induction machines, a satisfactory performance at low speed down to zero fundamental frequency can so far only be achieved by using signal injection methods and exploiting non-fundamental wave effects. The voltage pulse excitation technique is applied in this paper. The main problem in signal injection based sensorless control a round zero fundamental frequency is the separation of the different signal components related to the saliencies. The focus of this paper is using fast Fourier transform (FFT) to determine the amplitude and angle of the signal components induced by saturation and trained artificial neural network (ANN) for on-line compensation of these components at all loads. In addition a reduction of the remaining amplitude modulation (AM) is proposed. The output after the disturbance elimination is thus the extracted slotting signal that is used as control signal for calculating the rotor position using phase locked loop (PLL). This position is then used for calculating the rotor flux position in combination with the current model (rotor-equation) for field oriented control (FOC). The performance of the proposed structure is verified on a standard induction machine with un-skewed rotor.