Estimating photovoltaic power generation: Performance analysis of artificial neural networks, Support Vector Machine and Kalman filter

Abstract Current energy policies are encouraging the connection to the grid of power generation based on low-polluting technologies, mainly those using renewable sources with distribution networks. Photovoltaic (PV) systems have experienced a wide and high increase in their adoption as an energy source over the last years. Hence, it has become increasingly important to understand technical challenges, facing high penetration of PV systems on the grid, especially considering the effects of uncertainty and intermittency of this source on power quality, reliability and stability of the electric distribution system. On the other hand, the connections for distributed generators, by PV panels, changes the voltage profile on low voltage power systems. This fact can affect the distribution networks onto which they are attached causing overvoltage, undervoltage, frequency oscillations and changes in protection design. In order to predict these disturbances, due to this PV penetration, this article analyzes seven training algorithms used in artificial neural networks, with NARX architecture, for the generated active power estimating, and thus the state of the distribution network onto which these micro generators are connected and then compare their best statistical results with the Support Vector Machine (SVM) and the Kalman filter (KF) techniques. The results show that the best training algorithm used for the ANN learning obtained a mean absolute percentage error (MAPE) of 0.02%, while the SVM and KF techniques obtained 0.33% and 3.41%, respectively. Taking in account the other statistical analysis, we concluded that artificial neural networks are more suitable for this type of problem than SVM and KF. In addition, performing the training process with cell temperature data improves the accuracy of the resulting estimations.

[1]  E. Izgi,et al.  Short–mid-term solar power prediction by using artificial neural networks , 2012 .

[2]  Wei Qiao,et al.  Short-term solar power prediction using a support vector machine , 2013 .

[3]  Ali Naci Celik.  Artificial neural network modelling and experimental verification of the operating current of mono-crystalline photovoltaic modules , 2011 .

[4]  Jan Descheemaeker,et al.  Incentives and technical considerations related to increased voltage tolerance in low voltage distribution grids , 2014, 2014 16th International Conference on Harmonics and Quality of Power (ICHQP).

[5]  Giovanni Besio,et al.  Problems in RMSE-based wave model validations , 2013 .

[6]  Christian Riess,et al.  Continuous short-term irradiance forecasts using sky images , 2014 .

[7]  Mohamed A. A. Wahab.  Artificial neural network-based prediction technique for transformer oil breakdown voltage , 2004 .

[8]  Alagan Anpalagan,et al.  Improved short-term load forecasting using bagged neural networks , 2015 .

[9]  E. Mizutani,et al.  Neuro-Fuzzy and Soft Computing-A Computational Approach to Learning and Machine Intelligence [Book Review] , 1997, IEEE Transactions on Automatic Control.

[10]  Diego Carrion,et al.  Characterization and modeling of the efficiency of photovoltaic systems , 2015, IEEE Latin America Transactions.

[11]  Aleena Swetapadma,et al.  Directional relaying using support vector machine for double circuit transmission lines including cross-country and inter-circuit faults , 2016 .

[12]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[13]  Huseyin Eristi,et al.  Optimal feature selection for classification of the power quality events using wavelet transform and least squares support vector machines , 2013 .

[14]  Giuseppe Marco Tina,et al.  Analysis of forecast errors for irradiance on the horizontal plane , 2012 .

[15]  M. Valipour.  Optimization of neural networks for precipitation analysis in a humid region to detect drought and wet year alarms , 2016 .

[16]  Subrahmanyam Pulipaka,et al.  Modeling of soiled PV module with neural networks and regression using particle size composition , 2016 .

[17]  Zhang Qizhong.  Gene Selection and Classification Using Non-linear Kernel Support Vector Machines Based on Gene Expression Data , 2007, 2007 IEEE/ICME International Conference on Complex Medical Engineering.

[18]  Martin A. Riedmiller,et al.  A direct adaptive method for faster backpropagation learning: the RPROP algorithm , 1993, IEEE International Conference on Neural Networks.

[19]  Kanungo Barada Mohanty,et al.  Fault classification in SEIG system using Hilbert-Huang transform and least square support vector machine , 2016 .

[20]  Enen Ren,et al.  Comparative study of two uncertain support vector machines , 2012, 2012 IEEE Fifth International Conference on Advanced Computational Intelligence (ICACI).

[21]  M. Valipour,et al.  Comparison of the ARMA, ARIMA, and the autoregressive artificial neural network models in forecasting the monthly inflow of Dez dam reservoir , 2013 .

[22]  Antonio Valerio Netto.  Planning of network system for the distribution and transmission areas of electric energy , 2015 .

[23]  N. D. Kaushika,et al.  Artificial neural network model based on interrelationship of direct, diffuse and global solar radiations , 2014 .

[24]  A. Barin,et al.  Methodology for placement of Dispersed Generation Systems by analyzing its Impacts in Distribution Networks , 2012, IEEE Latin America Transactions.

[25]  M. Bolic,et al.  Comparison of Feed-Forward Neural Network training algorithms for oscillometric blood pressure estimation , 2010, 4th International Workshop on Soft Computing Applications.

[26]  Kleber Melo Silva,et al.  Modeling and Simulation of the Protection of Distribution Feeders in ATP , 2015 .

[27]  Xin Li,et al.  A Sensor Registration Method Using Improved Bayesian Regularization Algorithm , 2009, 2009 International Joint Conference on Computational Sciences and Optimization.

[28]  Fan Liu,et al.  Genetic Algorithms for MLP Neural Network parameters optimization , 2009, 2009 Chinese Control and Decision Conference.

[29]  Santosh Kumar Singh,et al.  Several variants of Kalman Filter algorithm for power system harmonic estimation , 2016 .

[30]  G. Lera,et al.  A quasi-local Levenberg-Marquardt algorithm for neural network training , 1998, 1998 IEEE International Joint Conference on Neural Networks Proceedings. IEEE World Congress on Computational Intelligence (Cat. No.98CH36227).

[31]  Lei Li,et al.  Fuzzy Multi-class Support Vector Machine Based on Binary Tree in Network Intrusion Detection , 2010, 2010 International Conference on Electrical and Control Engineering.

[32]  George J. Tsekouras,et al.  Estimation of seasonal variation of ground resistance using Artificial Neural Networks , 2013 .

[33]  Dipti Srinivasan,et al.  An improved particle swarm optimisation algorithm applied to battery sizing for stand-alone hybrid power systems , 2016 .

[34]  Mohammad Ali Gholami Sefidkouhi,et al.  Estimation of reference evapotranspiration using multivariate fractional polynomial, Bayesian regression, and robust regression models in three arid environments , 2017, Applied Water Science.

[35]  Vishal Kumar,et al.  Genetic algorithm based support vector machine for on-line voltage stability monitoring , 2015 .

[36]  Andries P. Engelbrecht,et al.  Computational Intelligence: An Introduction , 2002 .

[37]  Patrick P. K. Chan,et al.  Fuzzy support vector machines for solving two-class problems , 2003, Proceedings of the 2003 International Conference on Machine Learning and Cybernetics (IEEE Cat. No.03EX693).

[38]  Qunying Liu,et al.  Wind speed forecasting based on Time series - Adaptive Kalman filtering algorithm , 2014, 2014 IEEE Far East Forum on Nondestructive Evaluation/Testing.

[39]  E. Skoplaki,et al.  A simple correlation for the operating temperature of photovoltaic modules of arbitrary mounting , 2008 .

[40]  Zhang Xiaoling,et al.  Application of Kalman filter in agricultural economic forecasting , 1996, 1996 IEEE International Conference on Systems, Man and Cybernetics. Information Intelligence and Systems (Cat. No.96CH35929).

[41]  Farshid Keynia,et al.  Short-term wind power forecasting using ridgelet neural network , 2011 .

[42]  José L. Bernal-Agustín,et al.  Comparison of different lead–acid battery lifetime prediction models for use in simulation of stand-alone photovoltaic systems , 2014 .

[43]  Yonghua Song,et al.  Probabilistic Forecasting of Photovoltaic Generation: An Efficient Statistical Approach , 2017, IEEE Transactions on Power Systems.

[44]  Boudewijn Elsinga,et al.  An artificial neural network to assess the impact of neighbouring photovoltaic systems in power forecasting in Utrecht, the Netherlands , 2016 .

[45]  Soteris A. Kalogirou,et al.  Artificial neural network-based model for estimating the produced power of a photovoltaic module , 2013 .

[46]  Jie Yu,et al.  Short-term wind speed prediction using an unscented Kalman filter based state-space support vector regression approach , 2014 .

[47]  Kasra Mohammadi,et al.  A support vector machine–firefly algorithm-based model for global solar radiation prediction , 2015 .

[48]  Miguel Edgar Morales Udaeta,et al.  Distributed generation in Brazil: Advances and gaps in regulation , 2015 .

[49]  Xing Yan,et al.  Mid-term electricity market clearing price forecasting utilizing hybrid support vector machine and auto-regressive moving average with external input , 2014 .

[50]  O. S. Sastry,et al.  Estimation of solar radiation using a combination of Hidden Markov Model and generalized Fuzzy model , 2013 .

[51]  Ertuğrul Çam,et al.  Forecasting electricity consumption: A comparison of regression analysis, neural networks and least squares support vector machines , 2015 .

[52]  A. Mellit,et al.  A 24-h forecast of solar irradiance using artificial neural network: Application for performance prediction of a grid-connected PV plant at Trieste, Italy , 2010 .