An Levenberg–Marquardt trained feed-forward back-propagation based intelligent pitch angle controller for wind generation system. (December 2017)

• Record Type: Journal Article
• Authors: Sitharthan, R.; Devabalaji, K.R.; Jees, Arun
• Is Part Of: Renewable energy focus. Volume 22/23(2017)
• Page Start: 24
• Page End: 32
• Publication Date: 2017-12
• Subjects: Renewable energy sources -- Periodicals; Solar energy -- Periodicals; 333.79405
• DOI: 10.1016/j.ref.2017.10.003
• Languages: English
• ISSNs: 1755-0084

Abstract

Highlights: An adaptive FFBP based intelligent pitch angle controller is proposed for DFIG-Wind Turbine. The FFBP based pitch controller is trained using Levenberg–Marquardt algorithm. The proposed controller is tested for different wind speed condition and results prove the effective smoothening of wind turbine output power. Abstract : The frequent variation in wind speed affects the wind turbine (WT) to produce fluctuating output power and this may negatively collide the entire power system. This paper proposes a Feed Forward Back Propagation Neural Network (FFBP-NN) based pitch angle controller to mitigate the output power fluctuation in a grid connected wind generation system. The outstanding aspect of the proposed controller is that the optimal power of the WT is tracked in such a way that the output power is smoothed, when the wind speed flows below rated speed. Consequently, during above rated speed; the power is smoothed by traditional power regulating method. Further, the FFBP-NN controller is trained online using Levenberg–Marquardt (LM) algorithm and connecting weights of the neurons are updated means of LM algorithm using back propagation methodology. The effectiveness of the proposed FFBP based pitch controller is analyzed through the simulation study carried out in MATLAB/Simulink environment.