Maximum Power Point Tracking Algorithm for Laser Power Beaming Based on Neural Networks

With the high voltage intelligent substation developing at a pretty high speed, more and more artificial intelligent techniques have been incorporated into the power devices to meet the automation needs. For the sake of the line maintenance staff's safety, the high voltage isolating switch draws great attention among the most important power devices because of its capability of connecting and disconnecting the high voltage circuit. However, due to high-voltage isolating switch's working environment, the power supply system of the surveillance devices could suffer from great electromagnetic interference. Laser power beaming exhibits its merits in such situation because it can provide steady power from a distance despite the day or the night. Then the energy conversion efficiency arises as a new concern. To make full use of the laser power, this paper mainly focuses on extracting the maximum power from the photovoltaic (PV) panel. And a neural network based algorithm is proposed which combines both the intrinsic and the extrinsic features of the PV panel with the proportion of the voltage of the maximum power point (MPP) to the open circuit voltage of the PV panel. Simulations and experiments are carried out to verify the validness and feasibility of the algorithm.

[1]  Mansour Souissi,et al.  Maximum Power Point Tracking Control Using Neural Networks for Stand-Alone Photovoltaic Systems , 2014 .

[2]  Jordin T. Kare,et al.  Laser power beaming for defense and security applications , 2011, Defense + Commercial Sensing.

[3]  N. H. Helwa,et al.  Maximum power point traking controller for PV systems using neural networks , 2005 .

[4]  Hans S. Rauschenbach,et al.  Solar Cell Array Design Handbook: The Principles and Technology of Photovoltaic Energy Conversion , 1980 .

[5]  F.A.S. Neves,et al.  Evaluation of maximum power point tracking methods for grid connected photovoltaic systems , 2008, 2008 IEEE Power Electronics Specialists Conference.

[6]  Hiroaki Hayashi,et al.  Improving Stochastic Gradient Descent with Feedback , 2016, ArXiv.

[7]  Mohammad Ali Badamchizadeh,et al.  PV Maximum Power-Point Tracking by Using Artificial Neural Network , 2012 .

[8]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[9]  Josep M. Guerrero,et al.  A Novel Improved Variable Step-Size Incremental-Resistance MPPT Method for PV Systems , 2011, IEEE Transactions on Industrial Electronics.

[10]  G. Petrone,et al.  Predictive & Adaptive MPPT Perturb and Observe Method , 2007, IEEE Transactions on Aerospace and Electronic Systems.

[11]  Hayaru Shouno,et al.  Analysis of function of rectified linear unit used in deep learning , 2015, 2015 International Joint Conference on Neural Networks (IJCNN).

[12]  Marcelo Gradella Villalva,et al.  Comprehensive Approach to Modeling and Simulation of Photovoltaic Arrays , 2009, IEEE Transactions on Power Electronics.

[13]  Yoshua Bengio,et al.  Understanding the difficulty of training deep feedforward neural networks , 2010, AISTATS.

[14]  Yong Kang,et al.  A Variable Step Size INC MPPT Method for PV Systems , 2008, IEEE Transactions on Industrial Electronics.

[15]  W. Beckman,et al.  Solar Engineering of Thermal Processes , 1985 .