Room 31-213
Telephone 253-6047
Experimental application of control theory, especially in propulsion and fluid systems. This research generally spans three interrelated topics. Modeling complex, distributed systems in the context of control requires techniques that merge control theory with physical modeling methods from various disciplines (fluid dynamics, acoustics, etc.). Implementation of controllers in new environments requires procedures for optimizing the control configuration (sensor and actuator design and placement, for instance), as well as understanding hardware implementation of control. Finally, conducting meaningful experiments requires new or hybrid methods for data acquisition, signal processing, system identification, and robustness management.
Active Stabilization of Rotating Stall and Surge in Aeroengines: Practical implementation of active stabilization systems for rotating stall and surge, which are performance- limiting fluid mechanical instabilities in gas turbine engines. Experimental evaluation of fluid modeling, actuation, and control strategies. Research areas include: large-amplitude disturbance rejection through nonlinear control, reducing bandwidth requirements using advanced linear control theory, control-theoretic modeling and identification of distributed (PDE) fluid dynamics, and demonstration on high-speed compressors.
Active Control of Parallel-Flow Instabilities: Parallel-flow instabilities present new challenges to control theory because they are fundamentally infinite-dimensional, containing eigenvalue continua rather than discrete eigenvalues. To motivate and demonstrate active control techniques in this context, a water tunnel facility has been constructed which allows control of Kelvin-Helmholtz instability. Modeling, estimation and control concepts are under development for this system. Applications include shear layer instability as a source of jet noise and laminar-turbulent transition.
In-Flight Robustness Evaluation: NASA Dryden's flight test programs generally have an 'envelope expansion' stage, because the aircraft being tested are often operating for the first time (X-29), or are operating in new flight regimes (F-18 HAARV). Control system robustness must be verified as part of envelope expansion. New multivariable measures are being developed for thorough robustness evaluation of advanced flight controllers. In-flight determination of real structured singular values is the current focus. Methodologies for modifying control laws based on flight test results are envisioned.
Experimental System Identification: System identification in noisy, spatially distributed, uncertain environments. Integration of physical modeling, signal processing, and system identification. Feed- forward adaptive filtering to remove noise that corrupts the identification process. Identification of unstable dynamics during closed-loop operation through modified instrumental variables. Extensions include multivariable identification, real-time (recursive) identification in the face of noise correlated to an external source, and integration of flow visualization and ensemble averaging techniques into the identification process. Applications range from unsteady aerodynamics driven by vortex shedding to tail buffet alleviation using feedforward control.