Plasma Control Simulation
Control simulation codes model the plasma response to actuators so shape, position, and profile controllers can be designed and tested before operation.
Closing the loop
A tokamak is an unstable dynamical system that must be actively controlled: its vertical position, shape, current, and profiles are held by feedback from diagnostics to actuators like coils and heating systems. Control simulation codes model this closed loop so controllers can be designed and validated in software.
The plant model
Control design needs a model of how the plasma responds to actuator changes, the plant. This is often a linearized model around an operating point: how the equilibrium, position, and profiles shift when coil currents or heating change. Free-boundary evolution codes and system-identification from data both produce such models.
Controllers
- Vertical-stabilization loops that counter the fast vertical instability
- Shape and position controllers acting on the poloidal-field coils
- Current and profile controllers using heating and current drive
- Event-handling logic for off-normal conditions
Flight simulators
A control-oriented simulator, sometimes called a plasma flight simulator, runs the plant model in closed loop with the real control algorithms and realistic diagnostic noise and actuator limits. It lets operators rehearse scenarios and lets engineers find failure modes before they occur on hardware.
Real-time constraints
Controllers must compute within the control cycle, so the models embedded in real-time systems are deliberately simplified versions of the offline physics codes. Part of control simulation is verifying that these fast models remain faithful enough across the operating space to keep the plasma safe.
Designing the control approach for a machine like the Hyperion breeder, with its demanding shape, is done through such simulation long before the machine exists.