Reinforcement learning algorithm for improving speed response of a five-phase permanent magnet synchronous motor based model predictive control
2025

Improving Speed Response of a Five-Phase Motor Using Reinforcement Learning

publication 20 minutes Evidence: high

Author Information

Author(s): Hassan Ahmed M., Ababneh Jafar, Attar Hani, Shamseldin Tamer, Abdelbaset Ahmed, Metwally Mohamed Eladly

Primary Institution: Department of Electrical Power and Machines Engineering, Faculty of Engineering, Benha University, Shoubra, Cairo, Egypt

Hypothesis

Can a reinforcement learning algorithm optimize the speed response of a five-phase interior permanent magnet synchronous motor (5ph-IPMSM) drive system?

Conclusion

The proposed reinforcement learning algorithm significantly improves the speed response of the 5ph-IPMSM compared to traditional optimization techniques.

Supporting Evidence

  • The RL-based TD3 algorithm resulted in the fastest settling time and lowest overshoot compared to other optimization techniques.
  • Simulation results were obtained using MATLAB SIMULINK to validate the proposed control methodology.
  • The study compared the RL approach with Transit Search, Honey Badger Algorithm, Dwarf Mongoose, and Dandelion-Optimizer.

Takeaway

This study shows that using a smart computer program can help a special motor work faster and better, making it useful for things like electric cars.

Methodology

The study used a reinforcement learning algorithm called TD3 to tune two PI controllers in a five-phase motor control system, comparing its performance with four recent metaheuristic optimization techniques.

Limitations

The study does not include experimental validation of the proposed methodology.

Digital Object Identifier (DOI)

10.1371/journal.pone.0316326

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication