Neural Inverse Control of Wind Energy Conversion Systems

Abstract

Induction generators with double inputs have a high efficiency over a wide range of velocities caused by their capability of operation at variable speeds, so their application is increasing progressively. These generators along with wind turbines construct a suitable wind energy conversion system. Due to the intensive nonlinear and variant time characteristics of wind turbines and generators, there are some difficulties in conventional controlling methods. For this reason, usage of an adaptive controller is required. In the present paper, a predictive inverse neural model has been employed in order to control a wind system. In order to design this controller, input-output data set for wind energy conversion systems is required. In this paper, since real data for system were not available, modeling of the considered system has been performed. Afterward, two controlling structures including direct structure and adaptive scheme, which are based on multilayer neural networks, have been introduced for our modeled system. Furthermore, in order to study the ability of proposed controllers, several situations have been considered including the application of instant disturbance, the application of noise on the system as well as parameters variations and uncertainties of the system