Iterative-learning-based torque-ripple compensation in a transverse flux motor

Transverse flux motors (TFMs) are playing an increasingly important role in various drive systems owing to their high power densities, high efficiency, high torque at low speeds and reduced weight. The TFM is suitable for a variety of direct-drive applications where high efficiency and low maintenance are required. However, one of the main drawbacks of the TFM is its high parasitic torque pulsations, which severely limit its use in high-end applications where torque smoothness is very important. To overcome this limitation, an iterative learning control scheme that eliminates periodic torque-ripple effects is presented. An indirect, D-type, frequency-domain approach in a stand-alone configuration is proposed. Simulations and an experimental verification were performed, and the results obtained demonstrate the effectiveness of the proposed approach.