Variable-Pole Induction Machines Drives for Electric Vehicles
PhD student Elie Libbos with Advisor A. Banerjee
Induction machines (IM) are an attractive permanent magnet-free solution for electric vehicle drivetrains with distinct advantages such as low cost, reliability, and capability to withstand high short-term overload. Induction machines can operate at a much higher temperature when compared to permanent-magnet-based machines without the risk of demagnetization. Our proposed system consists of a toroidally-wound variable-pole IM (VPIM) drive with an integrated modular 18-phase converter, as shown in Figure 15. The toroidal winding configuration provides advantages for thermal performance over conventional distributed windings due to increased winding copper surface area without reducing the slot-fill factor, allowing for a higher stator density. Our developed variable-pole induction motor has an estimated power density of 40 kW/L. Our proposed power electronics and machine co-design framework shows system-level efficiency improvement, especially at partial torque and high speed, as shown in Figure 15(b), machine volume and loss reduction, 62% dc link capacitance reduction, and improvement in thermal performance compared to a conventional three-leg fixed-pole drive. Currently, we are working on integrating the thermal and electromagnetic models and pole-transition transients.
Figure 16(a) shows our experimental setup. An 18-leg GaN-based inverter is used to drive a 36-slot toroidally wound IM with access to each of its individual slots. This setup allows us to reconfigure the IM to various pole counts (2-, 4-, 6-poles). A Zedboard field-programmable gate array (FPGA) is used to control the converter. Figure 16(b) shows that the 18-leg drive significantly reduces drives losses, especially at partial load, due to variable-pole operation. These thermal improvements are mainly achieved because of the higher efficiency of the 18-leg converter and better loss-sharing among more phases.
This research is supported by Power Optimization of Electro-Thermal Systems (POETS) NSF Engineering Research Center at the University of Illinois and the Grainger Center for Electric Machinery and Electromechanics.