Torque ripple modeling and minimization for PMSM drives with consideration of magnet temperature variation

Abstract

The spatial harmonic of magnet flux is a major cause of torque ripple in permanent magnet synchronous machines (PMSMs), and it is temperature-dependent. Thus, this paper investigates torque ripple modeling and minimization for PMSMs considering magnet temperature variations. Firstly, experimental studies are conducted to demonstrate that the torque ripple is magnet temperature dependent. Then, based on extensive experimental tests, a novel linear model is proposed to model the relationship between the dc and harmonic components of magnet flux in the dq reference frame, which provides a way to estimate the magnet flux harmonic. Based on this model, the torque ripple model considering magnet temperature variation is proposed and validated with simulations. Afterwards, a novel adaptive current optimization approach is proposed for torque ripple minimization, which consists of two parts: the magnet flux harmonics estimation using the proposed linear magnet flux model, and the current optimization using the proposed torque ripple model. In our approach, the stator current is adaptively optimized with respect to the magnet temperature. However, the proposed approach is not necessary to run in real-time, because temperature variation has a large time constant. Our approach is validated through both numerical and experimental studies.