Date of Award
12-2016
Access Type
Thesis - Open Access
Degree Name
Master of Science in Aerospace Engineering
Department
Graduate Studies
Committee Chair
Dr. Richard Prazenica
First Committee Member
Dr. Mark Balas
Second Committee Member
Dr. Troy Henderson
Abstract
The purpose of this study is to investigate the potential effectiveness of using a Volterra-based Model Predictive Control strategy to control a nonlinear aeroelastic system. Model Predictive Control (MPC), also known as Receding Horizon Control (RHC), entails computing optimal control inputs over a finite time horizon, applying a portion of the computed optimal control sequence, and then repeating the process over the next time horizon. The Volterra series provides input-output models of a dynamical system in terms of a series of integral operators of increasing order, where the first-order Volterra operator models the linear dynamics and the higher-order operators model the nonlinear dynamics. In this thesis, Volterra-based Model Predictive Control is applied to simulated linear and nonlinear pitch-plunge aeroelastic systems. A linear MPC controller based on a first-order Volterra model is used to control the linear aeroelastic system, and the results are compared to those obtained using a standard LQR controller and a LQR-based MPC strategy. The controller is implemented for regulator and tracking cases for a free-stream velocity of 6 m/s, a condition for which the open-loop linear system is stable, and a free-stream velocity of 12.5 m/s, which corresponds to an unstable flutter condition. Nonlinear MPC controllers, using second- and third-order Volterra models, are then used to control the nonlinear aeroelastic system for regulator and tracking cases at the stable flight condition. The stability and performance of the linear and nonlinear Volterra-based MPC strategies are discussed, and a detailed analysis of the effect of different parameters such as the optimization horizon, control horizon and control discretization, is provided. The results show that the linear MPC controller is able to successfully track a reference input for the stable condition and stabilizes the system at the unstable flutter condition. It is also shown that the incorporation of the second- and third-order Volterra kernels in the nonlinear MPC controller provides superior performance on the nonlinear aeroelastic system compared to the results obtained using only a linear model.
Scholarly Commons Citation
Law, Wai Leuk, "Model Predictive Control of a Nonlinear Aeroelastic System Using Volterra Series Representations" (2016). Doctoral Dissertations and Master's Theses. 304.
https://commons.erau.edu/edt/304
