High-order design of adaptive inverses for signal-dependent actuator nonlinearities

In this paper, an adaptive compensation scheme is presented for controlling signal-dependent synthetic jet actuator nonlinearities, using higher order approximation of twin approximators for effective adaptive inversion of actuator nonlinearities. Approximation of a nonlinearly parametrized actuator model by a linearly parametrized function is performed. Adaptive inversion of actuator nonlinearities is implemented by another approximator in the feedforward path. The network reconstruction error is reduced in principle with usage of a second-order approximation, compared to a first-order approximation. A nonlinear state feedback control law is designed for controlling a nonlinear dynamic system. Parameter projection based adaptive laws ensure desired closed-loop stability.