Modal Blending for Active Flutter Suppression

Abstract

This paper presents a control design methodology for active flutter suppression in parameter-varying aeroservoelastic systems. A rectangular wing exhibiting bending-torsion flutter is modeled in a linear parameter-varying framework. Modal analysis of the critical flutter pole pair shows that both associated eigenvectors are well observed by distributed inertial measurements. Leveraging this property, a least-squares optimal blending of eight accelerometer signals is computed to isolate the flutter dynamics in a two-channel virtual output. A low-order structured $H_\infty$ controller is then synthesized on this reduced interface, targeting robust stabilization up to a specified freestream velocity with prescribed gain and phase margins while minimizing control effort. Relative to an $H_2$-optimal SISO blending baseline, the proposed approach achieves increased flutter damping and reduced control effort. The architecture further supports sensor fault tolerance by precomputing blending matrices for alternative sensor configurations, enabling reconfiguration without modifying the controller dynamics. Open-source MATLAB code accompanies the paper to facilitate reproducibility and extension to more complex systems.