Towards Sustainable Mobility: Optimal Flux Control of Drive Cycles for Variable Flux Motors via Deep Reinforcement Learning

In this paper, we present a deep reinforcement learning (DRL) approach for optimizing the efficiency of variable flux motors (VFM) during a simulated drive cycle through flux control by adjusting the rotor extension. A virtual environment is constructed for the DRL algorithm to determine flux adjustments based on simulated efficiency maps using full-wave finite element methods, along with drive cycle data derived from parameters of the vehicle. By comparing the DRL-optimized VFM with a fixed-flux motor at each time instant throughout the drive cycle, our results demonstrate that the DRL-based flux control method significantly enhances VFM performance by achieving efficiency gains, contributing to advancements in sustainable mobility.