Data-Driven Day-Ahead PV Estimation Using Autoencoder-LSTM and Persistence Model

Inherent variability in photovoltaic (PV) and associated impacts on power systems is a challenging problem for both the PV owners and the grid operators. Existing statistical and machine learning algorithms typically work well for weather conditions similar to historical data. However, uncertain weather conditions pose a great challenge to the estimation accuracy of the estimation models. With the enhanced integration of intelligent electronic devices and the realization of associated automation in the power grid, renewable energy data are becoming more accessible, which can be utilized by deep learning models and improve the PV power generation estimation accuracy. In this article, a hybrid deep learning model driven by external weather data is proposed to do day-ahead PV output forecasting at 15-min interval. The proposed model is motivated by the recent advancement of long-short-term-memory networks and autoencoder, which estimates uncertainties in sequence while making the prediction for complex weather conditions. Meanwhile, the persistence model is used to predict continuous sunny weather conditions. The forecasting result is validated with data from multiple locations.

[1]  Mohamed Abdel-Nasser,et al.  Accurate photovoltaic power forecasting models using deep LSTM-RNN , 2017, Neural Computing and Applications.

[2]  Yoshua Bengio,et al.  Learning long-term dependencies with gradient descent is difficult , 1994, IEEE Trans. Neural Networks.

[3]  Ignacio J. Ramirez-Rosado,et al.  Short-Term Power Forecasting Model for Photovoltaic Plants Based on Historical Similarity , 2013 .

[4]  Kara Clark,et al.  How Do High Levels of Wind and Solar Impact the Grid? The Western Wind and Solar Integration Study , 2010 .

[5]  Tao Ding,et al.  Hybrid method for short‐term photovoltaic power forecasting based on deep convolutional neural network , 2018, IET Generation, Transmission & Distribution.

[6]  A. G. Bakirtzis,et al.  Application of time series and artificial neural network models in short-term forecasting of PV power generation , 2013, 2013 48th International Universities' Power Engineering Conference (UPEC).

[7]  Guochang Wang,et al.  One-day-ahead daily power forecasting of photovoltaic systems based on partial functional linear regression models , 2016 .

[8]  T. Logenthiran,et al.  Forecasting of photovoltaic power using deep belief network , 2017, TENCON 2017 - 2017 IEEE Region 10 Conference.

[9]  J. Rogers,et al.  Review of Variable Generation Integration Charges , 2013 .

[10]  M. Omar Faruque,et al.  Forecasting of PV plant output using hybrid wavelet‐based LSTM‐DNN structure model , 2019, IET Renewable Power Generation.

[11]  Xiaoxia Qi,et al.  A comparison of day-ahead photovoltaic power forecasting models based on deep learning neural network , 2019, Applied Energy.

[12]  Robert Gall,et al.  Scientific Documentation for the NMM Solver , 2010 .

[13]  Kwanho Kim,et al.  Recurrent Neural Network-Based Hourly Prediction of Photovoltaic Power Output Using Meteorological Information , 2019, Energies.

[14]  Yan Su,et al.  Forecasting the daily power output of a grid-connected photovoltaic system based on multivariate adaptive regression splines , 2016 .

[15]  Hao Dong,et al.  Short-term output power forecasting of photovoltaic systems based on the deep belief net , 2017 .

[16]  Ratnesh K. Sharma,et al.  Data-Driven Day-Ahead PV Estimation Using Hybrid Deep Learning , 2019, 2019 IEEE Industry Applications Society Annual Meeting.

[17]  Mathieu David,et al.  Combining solar irradiance measurements, satellite-derived data and a numerical weather prediction model to improve intra-day solar forecasting , 2016 .

[18]  Reinhard Madlener,et al.  Economic merits of a state-of-the-art concentrating solar power forecasting system for participation in the Spanish electricity market , 2013 .

[19]  Bernhard Sick,et al.  Deep Learning for solar power forecasting — An approach using AutoEncoder and LSTM Neural Networks , 2016, 2016 IEEE International Conference on Systems, Man, and Cybernetics (SMC).

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

[21]  Shanlin Yang,et al.  A hybrid deep learning model for short-term PV power forecasting , 2020 .

[22]  Yue Zhang,et al.  Day-Ahead Power Output Forecasting for Small-Scale Solar Photovoltaic Electricity Generators , 2015, IEEE Transactions on Smart Grid.

[23]  Luca Massidda,et al.  Use of Multilinear Adaptive Regression Splines and numerical weather prediction to forecast the power output of a PV plant in Borkum, Germany , 2017 .

[24]  Badia Amrouche,et al.  Artificial neural network based daily local forecasting for global solar radiation , 2014 .

[25]  Peng Wang,et al.  Forecasting Power Output of Photovoltaic Systems Based on Weather Classification and Support Vector Machines , 2011, IEEE Transactions on Industry Applications.

[26]  João P. S. Catalão,et al.  Neural Network Based Irradiance Mapping Model of Solar PV Power Forecasting Using Sky Image , 2018, 2018 IEEE Industry Applications Society Annual Meeting (IAS).

[27]  N. Rahim,et al.  Solar photovoltaic generation forecasting methods: A review , 2018 .