Adaptation of Predictive Acoustic Noise Control of IM Drive to Variable Operating Conditions

Finite control set MPC of IM drive is known to produce a spread spectrum of the stator currents. Spectrum of the acoustic noise also exhibits higher spread than that of the PWM which may be perceived as more pleasant sound by humans. Predictive acoustic noise control has been recently proposed using FCS-MPC with additive term maximizing spectrum flatness of the emitted acoustic noise. It has been shown that the acoustic noise generated by this control has better properties than the conventional random PWM. In this contribution, we study variability of the acoustic model with respect to operating conditions. We show that constant penalization factors are not optimal. We design a measure of performance of the acoustic noise controller and propose a scheduling mechanism for adjustment of the penalization coefficient to optimize this measure. Since we require noise measurements, the procedure is expected to be used during commissioning of the drive.