Skip to main content
APEC 2026 Main banner

Modeling and Control Techniques for Motor Drives

T10.9 - Online Reinforcement Learning Torque Controller for Electrical Motor Drives

Wednesday, March 25, 2026
11:40 AM - 12:00 PM CT
Location: 214B
This paper proposes a novel torque controller for electrical motor drives based on a Proximal Policy Optimization (PPO) reinforcement learning (RL) algorithm, able to track the torque references while aiming to minimize the norm of the stator currents. The method is based on the design of the reward function by establishing a balance between torque tracking and current minimization. The algorithm is implemented in a digital signal processor (DSP) and tested in a synchronous reluctance machine (SynRM) simulated in a hardware in-the-loop (HIL) setup. The results show that the algorithm is able to correctly track the torque reference for different points of operation, without relying on any offline data, look-up table (LUTs) or edge-computing.