Maximum Power point Tracking Using P&O Control Optimized by a Neural Network Approach: A Good Compromise between Accuracy and Complexity

Abstract In the field of power optimization of photovoltaic panels (PV), there exist many maximum power point tracking (MPPT) control algorithms, such as: the perturb and observe (P&O) one, the algorithms based on fuzzy logic and the ones using a neural network approaches. Among these MPPT control algorithms, P&O is one of the most widely used due to its simplicity of implementation. However, the major drawback of this kind of algorithm is the lack of accuracy due to oscillations around the PPM. Conversely, MPPT control using neural networks have shown to be a very efficient solution in term of accuracy. However, this approach remains complex. In this paper we propose an original optimization of the P&O MPPT control with a neural network algorithm leading to a significant reduction of the computational cost required to train it, ensuring a good compromise between accuracy and complexity. The algorithm has been applied to the models of two different types of solar panels, which have been experimentally validated.

[1]  F. Olivie,et al.  Conception et modélisation d’un système photovoltaïque adapté par une commande MPPT analogique , 2007 .

[2]  M. Georgiopoulos,et al.  Feed-forward neural networks , 1994, IEEE Potentials.

[3]  L. Coelho,et al.  Nonlinear identification using a B-spline neural network and chaotic immune approaches , 2009 .

[4]  Bidyadhar Subudhi,et al.  A differential evolution based neural network approach to nonlinear system identification , 2011, Appl. Soft Comput..

[5]  Niels Kjølstad Poulsen,et al.  Neural Networks for Modelling and Control of Dynamic Systems: A Practitioner’s Handbook , 2000 .

[6]  Ali Nazari,et al.  Artificial neural network to predict the effect of heat treatments on Vickers microhardness of low-carbon Nb microalloyed steels , 2013, Neural Computing and Applications.

[7]  Hector M. Romero Ugalde,et al.  Neural network design and model reduction approach for black box nonlinear system identification with reduced number of parameters , 2013, Neurocomputing.

[8]  Wei Wu,et al.  Boundedness and convergence of batch back-propagation algorithm with penalty for feedforward neural networks , 2012, Neurocomputing.

[9]  C. Larbes,et al.  Maximum power point tracking using a fuzzy logic control scheme , 2007 .

[10]  Marzuki Khalid,et al.  Structure optimization of neural network for dynamic system modeling using multi-objective genetic algorithm , 2011, Neural Computing and Applications.

[11]  Andrzej Cichocki,et al.  Neural networks for optimization and signal processing , 1993 .

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

[13]  Peter J. Gawthrop,et al.  Neural Networks for Modelling and Control , 1997 .

[14]  Mehran Habibi,et al.  New MPPT Controller Design for PV Arrays Using Neural Networks (Zanjan City Case Study) , 2009, ISNN.

[15]  Yanchun Liang,et al.  Identification and control of nonlinear systems by a dissimilation particle swarm optimization-based Elman neural network , 2008 .

[16]  Ruey-Maw Chen.  Reducing network and computation complexities in neural based real-time scheduling scheme , 2011, Appl. Math. Comput..

[17]  I. Colda,et al.  Influence d’un ombrage partiel sur la production électrique d’un module photovoltaïque en silicium monocristallin , 2012, Journal of Renewable Energies.

[18]  C. Cabal,et al.  Optimisation énergétique de l'étage d'adaptation électronique dédié à la conversion photovoltaïque , 2008 .

[19]  Xiangping Zeng,et al.  Low-Complexity Nonlinear Adaptive Filter Based on a Pipelined Bilinear Recurrent Neural Network , 2011, IEEE Transactions on Neural Networks.

[20]  Feng Qian,et al.  Identification and control of nonlinear systems by a time-delay recurrent neural network , 2009, Neurocomputing.