PMSM for high demands on low torque ripple using optimized stator phase currents controlled by an iterative learning control algorithm

Brushless permanent magnet machines (BPMMs) gained more importance due to the high power density ensured by permanent magnets with high energy densities. Therefore, a more compact design can be achieved compared to other types of electrical machines. However, the ferromagnetic components are typically driven near or above the saturation flux density at least at nominal load operation. This leads to non-sinusoidal linked fluxes in the stator coils and, hence, creates disruptive torque ripple during operation when the motor is fed with sinusoidal currents. Additionally, BPMMs often show a considerable cogging torque and, hence, torque ripple can even be noticed at no-load operation. In this article, a motor design with a certain number of significant flux harmonics is presented. Non-sinusoidal current waveforms were optimized in order to fulfill the operating performance requirements of the electric drive regarding efficiency and torque ripple. For impressing the desired current waveforms including higher harmonics related to the electrical frequency of the motor, the specification of the used power electronics has to be taken into account. Furthermore, the structure and design of the current control has a big impact on the possibility to impress specific current waveforms even at higher rotor speeds. This article is about a current control structure using an iterative learning control algorithm (ilc). In the introduction, different current control strategies are given that were published in the past. This will be followed by the presentation of the considered motor design. The control structure and the hardware used for experimental results are subsequently described. This will be followed by measurement results of the achieved currents and load torques and finally a conclusion is derived.