A neural network-based approach to active structural mode suppression for flexible transport aircraft

The dynamics of the first few structural bending modes in large, flexible transport aircraft are typically characterized by low frequency and light damping. In a scenario where the frequencies of these modes are in close proximity with the rigid body modes, notch filtering yields undesirable results, and active control must be used to suppress structural excitation. This paper develops a methodology for the design of a structural mode suppression system using an ordered neural network-based approach. The resulting control system is robust over a set of linear plant models, which are used as a training set during the neural control design process for the original nonlinear system.