Alumni Shahrazed 2017

Dr. Shahrazed Elmetennani has successfully defended her PhD dissertation and will be joining the University of Texas at Austin for a postdoctoral fellow position.

5/31/2017
​​The work presented by Shahrazed addresses the design of new model based flow control algorithms of a first order hyperbolic PDE. The control problem is formulated to manage the heat transfer mechanism in distributed concentrated solar collectors. Both early and late lumping approaches have been followed to design three different control strategies endowed with three adaptation laws: a Lyapunov nonlinear controller and two PDE based output feedback strategies. A new approximation scheme for PDE model reduction has been proposed resulting in a reduced order approximate state space representatio n for system analysis and control design. The proposed approximate model compromises between the accuracy of the dynamics approximation and the simplicity of the final state space representation. It represents an accurate tool to enhance the control efficiency and reduce the computational burden. Furthermore, the proposed approximation scheme is based on a modified Gaussian interpolation which can be applied to a more general class of PDEs, including higher order systems and nonlinear equations. A new adaptive control approach has been developed resorting to Lyapunov stability theory combined with a phenomenological representation of the system. A phenomenological model has been considered to describe the system behavior taking into account the modeling uncertainties and the external disturbances in the plant. The control algorithm presents interesting results under smooth and abrupt disturbance changes with limited measurement. In addition, the adaptive control can be directly adapted to a general category of nonlinear systems.

A nonlinear output feedback scheme has been designed to control the harvested thermal energy. The controller is derived by stabilization of the distributed profile of the reference tracking error. The control design results in an explicit formulation of the control law as a function of the system measurements. We adapted the proposed controller to design an adaptive strategy based on an on-line estimator. The source term estimation error is stabilized using a PI correction feedback.

A nonlinear mapping based approach for flow control of the heat transfer mechanism using the boundary measurements has been proposed. From the reference tracking problem of the hyperbolic distributed PDE, the new control approach formulates

an algebraic equation depending on the equation source term, the transport velocity, and the boundary conditions. Consequently, this control design strategy allows both flow and internal control from boundary measurement using the source term and the transport velocity respectively without numerical approximations. We endowed the nonlinear controller with an on-line estimator of the external working conditions resulting in an adaptive nonlinear strategy. The adaptation law is also based on a PI correction feedback.


 List of selected papers:

  • S. Elmetennani, T.M. Laleg-Kirati, "Bilinear approximate model based robust Lyapunov control for parabolic distributed collectors", IEEE Transactions on Control Systems Technology, To be online soon.
  •  S. Elmetennani, T.M. Laleg-Kirati, "Bilinear reduced order approximate model of parabolic distributed solar collectors", Solar Energy, Volume 131, Jun 2016.
  • S. Elmetennani, T.M. Laleg-Kirati, "Output feedback control of heat transport mechanisms in parabolic distributed solar collectors", The 2016 American Control Conference, Boston, USA, Jul 2016.