A model based control system for the Scirocco Plasma Wind Tunnel

The paper deals with the modeling and control of the Scirocco Plasma Wind Tunnel under construction at CIRA (Italian Aerospace Research Center). This is a modern, arc heated, plasma wind tunnel capable to reproduce, on suitable test articles, the thermal stress conditions to which some parts of capsules or space vehicles are subjected during the re-entry in atmosphere. The proposed model of the plant is essentially oriented to the control system design and development, to the determination of the preprogrammed input commands needed for the execution of dynamic tests during the operating life of the facility and to the testing of the control system by means of real-time simulation of the plant. The complex aerothermodynamic processes occurring during the wind tunnel operations, usually described by the Navier-Stokes equations are here modeled by means of a mixed ordinary differential equations-neural network model. The proposed control scheme consists of two terms: a feedforward action, calculated via a receding horizon technique using the above plant model, which guarantees the tracking of the desired output trajectory in absence of modeling errors, and a gain scheduled discrete time output feedback control action to counteract possible misalignments between the desired trajectory and the actual one. Numerical simulations, carried out using the proposed simulation model, show the effectiveness of the control scheme.