Toshiba AI Technology Catalog

  • Operation and Control

Automatic construction of PMSM drive logic using reinforcement learning

RL expects to achieve advanced control using a data-driven approach.


  • RL will obtain control logics that appropriately drive PMSM through trial and error basis without any knowledge of the target.

Applications



  • Various types of control; e.g., motor control and crane control

Benchmarks, strengths, and track record



  • Research is currently ongoing to achieve differentiation through more advanced controls, leveraging the strength of Toshiba’s devices

Inquiries



Please include the title “AI Technology Catalog: Automatic construction of PMSM drive logic using reinforcement learning” or the URL in the inquiry text.
Please note that because this technology is currently the subject of R&D activities, immediate responses to inquiries may not be possible.

References:

  • Toshiya Takano, Ryosuke Saito; “Simulation of Reinforcement Learning Approach to a Motor Speed Control”; 2019 Institute of Electrical Engineers of Japan Electronics, Information and Systems Conference, p. 1210-1214
  • Toshiya Takano, Tomoaki Shigeta; “Experiment of Motor Speed Control by Reinforcement Learning Approach”, 2021 Institute of Electrical Engineers of Japan Electronics, Information and Systems Conference, p. 974-979
  • Naoya Matsumoto,Toshiya Takano, “Experiment of Motor Speed Control by Reinforcement Learning Approach”; 2021 Institute of Electrical Engineers of Japan Electronics, Information and Systems Conference, p. 1081-1085
  • “Automated generation of PMSM speed control model by reinforcement learning”; Toshiba Review technology results, Vol. 75, No. 2, p. 74
  • Tomoaki Shigeta, Toshiya Takano, “Method to Automatically Generate PMSM Speed Control Model Using Reinforcement Learning”; Toshiba Review, Vol. 76, No. 6, General Papers, p. 29-33