Artificial neural network–based sensorless control of wind energy conversion system driving a permanent magnet synchronous generator

Abstract

Due to the wind characteristic, wind speed measure requires more than one sensor. However, to track the maximum power point of the wind, knowing the wind speed or mechanical speed is necessary. So, the solution is to use a sensorless control. This article is mainly focused on a sensorless control of a wind energy conversion system that employs an artificial neural network observer. The detailed mathematical model of the studied system is presented. It includes a permanent magnet synchronous generator. The contribution of the studied wind energy conversion system is to integrate a three-cell DC–DC converter. For the generation of maximum power from the wind, an algorithm to track the maximum power is developed. Then, to avoid the disadvantages of using sensors, an artificial neural network observer is implemented. The capabilities and contributions of the proposed control scheme are demonstrated by simulation results using MATLAB/Simulink.