According to Aditya Kumar from GE Vernova, the needs for control system innovations for Nuclear fusion include:

  • Modeling for control:
    • Need real-time capable transient models for model-based plant control
    • Need fast, robust transient model and model validation
    • Need to be modular, configurable, physics-based, and parametrizable
  • Sensing needs:
    • Harsh environments
    • Includes virtual sensors
  • Digital twins, virtual sensing, controls
    • Tracking of plant health degredation of key components
  • Marry physics-based and data-based techniques for the control of components whose physics is too complex
  • AI/ML for health monitoring and fault diagnostics (including fault isolation)
  • Cyber-physical society