LPV Modeling and Robust Sampled-Data $h_{\infty }$ Control of a Tailless Flapping Wing Microaerial Vehicle with Parameter Uncertainties
Qingcheng Guo,Zhicheng Deng,Chaofeng Wu,Feng Cui,Qing-Hao Zhang,Jun-Guo Lu
DOI: https://doi.org/10.1109/tmech.2024.3425310
2024-01-01
Abstract:To enhance the tracking performance of tailless flapping wing microair vehicles in the presence of parameter uncertainties, this work proposes a linear parameter varying (LPV) model of the flapping wing micro aerial vehicle (FWMAV) and a robust H-infinity controller. Initially, a polytopic LPV model for the FWMAV is established, considering variations in parameters, such as aerodynamic coefficients, mass distribution, and wing morphological parameters. Based on this LPV model, a robust H-infinity controller is designed for parameter uncertainties, ensuring robust performance in adverse conditions. Two system features are considered for practical applicability. The first is to avoid divergence and performance degradation caused by controller discretization, which is handled by utilizing sampled-data signals in controller design. The second is the low computing power of the hardware of the control unit, which is resolved by constraining the controller to be a low-order proportional-integral (PI) output feedback structure. Within these two features, the robust H-infinity controller design is solved by an optimization problem with linear matrix inequality (LMI) constraints. A Lyapunov-Krasovskii function is employed to establish the stability and H-infinity performance of the FWMAV. Numerical simulations demonstrate the strong robustness of the proposed controller, and flight experiments verify its superior performance compared to a state-of-the-art benchmark controller.