A Current Harmonic Suppression Method for PMSM Based on Harmonic Prediction Adaptive Notch Filter

The adaptive filters (ANF) have been commonly used in current harmonic suppression. However, the digital delay will narrow the stability region and its harmonic suppression performance will be deteriorated at high-speed region. To overcome this problem, this paper proposed two modified ANF. Combining with the deadbeat current (DBC) controller, the effect of digital delay on ANF is analyzed. A simple one step harmonic prediction method (OSPANF) is proposed to replace the traditional steepest gradient method (SGM). Based on the OSPANF, the current can be controlled stably in the high-speed area and the harmonic suppression effect is better than the traditional notch filter. Through bode diagram analysis, OSPANF will have a harmonic amplification peak near the set notch frequency, which may bring additional disturbances. Thus, a two steps prediction adaptive notch filter (TSPANF) is implemented. Through the Experimental results, it can be proved that the proposed TSPANF method has better harmonic suppression effect in the high-speed area than OSPANF and the stability region of the proposed two methods is larger than the traditional method.