Enhancing performance of permanent magnet motor for electric vehicle drives using model predictive control

Abstract

In the electric three-wheeler vehicular segment, permanent magnet (PM) motors are strong competitors for traction applications. The vehicular environment faces sudden load variations, where traditional control like field-oriented control (FOC) over-excites the machine. It also degrades the machine’s overall performance due to dynamic changes in the motor states with limited bandwidth of the cascaded proportional-integral (PI) controllers; this makes the overall motor operation inefficient. The paper proposes a model predictive control (MPC) based on torque ripple and over-current excitation minimization for permanent magnet motors to overcome these challenges. The proposed controller is tailored to improve the overall electrical efficiency of permanent magnet motors while considering the real back EMF. The performance of the proposed controller is tested in simulation, considering an electric three-wheeler as a load along with a European driving cycle. The feasibility of the proposed method is tested by deployment on (1) STM32F546ZG micro-controller and (2) Xilinx’s ZYNQ-7000 SoC ZC706 FPGA and validated the implementation results with hardware-in-the-loop(HIL) co-simulation. The implementation results show that the proposed MPC outperforms the conventionally used FOC in energy consumption, torque ripples minimization, and torque disturbance handling while tracking the desired speed accurately with load variations.