Memory machine (MM) was regarded as a promising candidate for wide-speed-range applications [1]–[6]. Since the low coercive force (LCF) PM is employed in MM, the air-gap flux can be flexibly adjusted by changing the magnetization state (MS) of PM with a temporary current pulses. The speed range can be subsequently extended with high efficiency maintained. According to the magnetizing current pattern, MMs can be generally categorized into AC- [1]–[3] and DC-magnetized [4]–[6] types. The former MM normally employs vector-control algorithm to apply d-axis current pulse in stator windings to vary the magnetization level of LCF PMs, while the latter one adopts auxiliary DC magnetizing coils to facilitate the magnetization control. Besides, the rotor-PM topologies of MMs have structures similar to conventional interior PM machines, which shows comparable torque density with those conventional surface-mounted PM machines . Nonetheless, LCF magnets on the rotor generally suffer from to armature reaction demagnetization effect, and the integration of armature and magnetizing functions in the stator winding results in complicated online FW control. On the other hand, the stator-PM MMs have the advantages of easy online magnetization control, and a simple and robust salient rotor as well as easy thermal management can be obtained. Nevertheless, the fact that two kinds of PMs, i.e., NdFeB and LCF PMs, as well as two sets of windings are located on stator leads to excessively crowded stator space and low torque density. Therefore, this paper attempts to propose a novel dual-sided PM memory machine (DSPM-MM) by combing the distinct advantages of “high torque density” of rotor-PM MM and “simple online PM flux control” of stator-PM MM. In the proposed design, the consequent-pole NdFeB PMs are placed in the rotor, while LCF PMs are mounted between the adjacent stator teeth to enable flexible air-gap flux adjustment. The machine topology and operating principle are introduced and addressed, respectively. Then, the electromagnetic performance is analyzed, which confirms the feasibility of the proposed design.
Read full abstract