This paper describes a new technique which has been implemented on the JET tokamak to investigate integrated real-time control of several plasma profiles simultaneously (such as current, temperature and pressure) and reports the results of the first experimental tests. The profiles are handled through their projection on a suitable basis of functions according to the Galerkin scheme. Their response to three actuators (heating and current drive powers injected in the plasma) is linearized in an experimentally deduced multi-input multi-output model. The singular value decomposition of this model operator allows us to design a distributed-parameter real-time controller which maximizes the steady state decoupling of the multiple feedback loops. It enables us to control several coupled profiles simultaneously, with some degree of fuzziness to let the plasma evolve towards an accessible non-linear state which is the closest to the requested one, despite a limited number of actuators.The first experiments using these techniques show that different current and electron temperature profiles can be obtained and sustained by the controller during a closed-loop operation time window. Future improvements and perspectives are briefly mentioned.
First-principles-driven model-based current profile control for the DIII-D tokamak via LQI optimal control