Low-order linear parameter-varying model development and robust control design for automotive and manufacturing applications

This dissertation proposes low-order model development and controller design for various automotive and manufacturing applications. The first task models a single wafer rapid thermal process (RTP) using first-principles modeling. Then, making use of the linear parameter-varying (LPV) state-space representation, we convert the nonlinear RTP model into an affine LPV model. For controller design purposes, we reduce the number of scheduling variables and the order of the model. Using this reduced-order model, we design a gain-scheduled H_infinity controller for reference tracking. In the second task, we reconfigure the model of a cooperative adaptive cruise control platoon in order to account for the uncertain and time-varying parameters of the system dynamics and communication delay. Using this model, we design a robust controller and reduce its order. The reduced-order controller remains robust to the model. We validate the controller design by performing experiments on the test bed to show the need for robust control. The third task uses model order reduction techniques to design a low-order robust controller in a parabolic convection-diffusion equation application.