A nonlinear vibration control design with a neural network realization

A novel type of nonlinear dynamic controller design is explored that was originally motivated by certain issues in the problem of vibration suppression in flexible structures. Because of the particular form of quadratic nonlinearities, the controller provides extremely efficient energy exchange mechanisms capable of rapidly draining energy away from the structural plant. In addition the controller is shown to consist of the interconnection of simple repetitive modules and its efficiency increases with the number of such modules. These features motivate the implementation of the nonlinear controller via a specified neural network architecture