Testing of a Model-Based Predictive Control System in a Transonic Aerodynamic Test Facility

A model-based control system has been developed and tested in the Arnold Engineering Development Center 4-ft Transonic Wind Tunnel. Unlike systems which provide static feedforward values based on a table or curve fit, the 4T Predictive and Adaptive Control System uses a set of generic time-accurate models which predict the effect of facility control surface and test article movements on tunnel conditions to determine future errors and make adjustments to maintain set point at each instant in time. Therefore, the system is designed to hold conditions while the test article is moved and to follow arbitrary set point trajectories (within facility limits), rather than simply reducing the time to return to conditions or to reduce time to change between fixed set points. The generic nature of the models facilitates rapid retuning (adaptation) to changes in system behavior; for example, if a test article of different size or shape is inserted into the tunnel. While hand-tuned feed-forward control has been implemented in various facilities, this is the first time (to the authors’ knowledge) that a general Model Predictive Control-derived control system has been developed for a large-scale test facility. In this paper, the basic results of system checkout and demonstration are presented.