Torque Ripple Reduction Based on Adaptive-Linear-Neuron Algorithm Caused by Offset Errors of Current Measurement

The current offset errors has negative influence on the accurate current control-the output electromagnetic torque will oscillate at the fundamental frequency. Therefore, a method based on adaptive-linear- neuron algorithm is proposed to extract the useful dc component of the feedback and neglect the noise signal. Two current harmonic decomposers are utilized to obtain the current of d axis and q axis separately. The proposed method covers no additional hardware circuits and does not rely on the motor parameters. The experimental results are provided to verify the effectiveness of the proposed method.