<p>In this paper, a speed fluctuation suppression method is proposed, where active disturbance rejection controller (ADRC) combined with iterative learning algorism is newly adopted, and a magnetic field modulation permanent magnet (MFM-PM) hub motor is selected as a control example. Based on the derivation and analysis of the system disturbance transfer function of the traditional, active disturbance rejection controller the intrinsic mechanism of the limitation of traditional active disturbance rejection controller in suppressing periodic interference is investigated in detail. Based on this, the iterative learning is introduced and integrated into the extended state observer (ESO) to extract the periodic disturbance information in the speed observation error, so as to enhance the observation capability for periodic disturbances and enabling effective compensation. In addition, the control performances of the proposed active disturbance rejection controller are investigated, including speed ripple, disturbance observation capability, and so on. Finally, the experimental results confirm the validity of the speed fluctuation suppression method with proposed active disturbance rejection controller algorism.</p>