Synthetic jet actuator based adaptive neural network control of nonlinear fixed pitch wind turbine blades

Abstract

This paper presents a neural network-based adaptive compensation scheme to cancel the effect of uncertain, highly complex and dynamic synthetic jet actuator nonlinearities. Approximation of a nonlinearly parameterized model of synthetic jet actuator characteristics by a linearly parameterized function is performed using neural network approximators. The nonlinearity function is approximated over a range of rotor rotational speed of a wind turbine blade. An adaptive inverse is employed for cancelling the effect of actuator nonlinearities, which is accomplished by use if another neural network. Adaptive update laws are also employed for estimation of blade physical dimensional parameters. A state feedback control law is designed to control the nonlinear wind turbine dynamics in presence of signal dependent actuator nonlinearities.