Torque Allocation Algorithm and Modular SiC MOSFET Dual-Inverter Hardware Design for Distributed Multi-Motor Electric Vehicle Powertrains

Electric vehicles (EVs) have been developed over the last decades towards higher rated power levels to replace traditional ICE vehicles as a green energy alternative. The EV powertrains are changing from a mechanical differential with a central high-speed motor to distributed multi-motor direct-drive topologies to increase modularity, flexibility, controllability of the EV powertrains, and resulting in right-sizing solutions with reduced Total Cost of Ownership. For a multi-drive EV, the torque vectoring technique produces the reference total torque value based on vehicle handling performance. The torque allocation technique allocates the total torque between the motor drives of the vehicle, including front-to-rear and side-to-side distribution. This research presents a hybrid side-to-side torque allocation algorithm. The designed approach combines energy efficiency and efficient handling performance modes. The simulations are performed using an experimentally validated real-time losses calculation model of the vehicle axle with two motor drives. The dual 200 kW (100 kW $\times 2$ ) SiC MOSFET inverter for a vehicle axle is designed for hardware validation. The dual-inverter setup includes two SiC-based 3-phase inverter modules combined with a shared DC link. The first experimental setup includes the designed dual-inverter setup connected to two direct-drive 1600 Nm 1500 rpm PMSM electric machines loaded with 400 Nm eddy current brakes. The second experimental setup is built with in-wheel motors loaded with identical motors to investigate the regeneration operation mode in a two-motor vehicle axle. The experimental results confirm the advantages of the proposed hybrid technique compared to the baseline torque allocation algorithms and show $\approx 20$ % less total energy losses resulting in $\approx 0.4$ percentage points higher average efficiency for proposed improved torque allocation algorithm over the tested scenario. At the beginning of algorithm application, for small steering angles, the losses of the proposed algorithm are almost halved. Then, by the end of the application period, for higher steering angles, the losses of the proposed algorithm reach the level of SOTA energy efficiency mode.