| Definition | gym-electric-motor (GEM) is a toolbox for simulating and controlling electric motor systems, supporting both classical control and reinforcement learning applications. |
| No of environments | 27 |
The gym-electric-motor (GEM) package is a versatile Python toolbox designed for simulating and controlling various electric motor systems, facilitating both classical control simulations and reinforcement learning experiments. Built upon the Farama Gymnasium Environments, GEM enables the construction of detailed drive train models composed of essential components such as supply voltages, converters, electric motors, and load models. This package not only provides a closed-loop simulation of these physical systems but also integrates a rich interface for implementing various decision-making algorithms, ranging from traditional linear feedback control to advanced methods like Deep Deterministic Policy Gradient (DDPG) agents.
Getting started with GEM is streamlined through interactive notebooks available in Google Colaboratory, showcasing key features and application demonstrations that serve as valuable resources for engineers in both industry and academia. Environments, such as the discretely controlled Permanent Magnet Synchronous Motor (PMSM), can be instantiated easily, allowing for experimentation with various control actions. The library supports multiple models of DC and AC motors, including synchronous reluctance motors and squirrel cage induction motors, alongside flexible converter options driven by duty cycles or switching commands. Overall, GEM serves as a powerful tool for exploring the dynamics of electric drive systems and testing and refining control strategies.
| Example observations | Motor current, velocity, torque, converter’s state. |
| Example actions | Apply voltage. Definine the converter’s switching state. |
| Example reward | Distance between the momentary and desired operation point. Limit values are surpassed. |