Search

Ultra-short-term photovoltaic (PV) power forecasting can support the real-time dispatching of the power grid. However, PV power has great fluctuations due to various meteorological factors, which increase energy prices and cause difficulties in managing the grid. This article proposes an ultra-short-term PV power forecasting model based on the optimal frequency-domain decomposition and deep learning. First, the optimal frequency demarcation points for decomposition components are obtained through frequency-domain analysis. Then, the PV power is decomposed into the low-frequency and high-frequency components, which supports the rationality of decomposition results and solves the problem that the current decomposition model only uses the direct decomposition method and the decomposition components are not physical. Then, a convolutional neural network (CNN) is used to forecast the low-frequency and high-frequency components, and the final forecasting result is obtained by addition reconstruction. Based on the actual PV data in heavy rain days, the mean absolute percentage error (MAPE) of the proposed forecasting model is decreased by 52.97%, 64.07%, and 31.21%, compared with discrete wavelet transform, variational mode decomposition, and direct prediction models. In addition, compared with recurrent neural network and long–short-term memory model, the MAPE of the CNN forecasting model is decreased by 23.64% and 46.22%, and the training efficiency of the CNN forecasting model is improved by 85.63% and 87.68%. The results fully show that the proposed model in this article can improve both forecasting accuracy and time efficiency significantly. [ABSTRACT FROM AUTHOR]

The precise minute time scale forecasting of an individual PV power station output relies on accurate prediction of cloud distribution, which can lead to dramatic fluctuation of PV power generation. Precise cloud distribution information is mainly achieved by ground-based total sky imager, then the future cloud distribution can also be achieved by sky image prediction. In previous studies, traditional digital image processing technology (DIPT) has been widely used in predicting sky images. However, DIPT has two deficiencies: relatively limited input spatiotemporal information and linear extrapolation of images. The first deficiency makes the input spatiotemporal information not rich enough, while the second creates the prediction error from the beginning. To avoid these two deficiencies, convolutional autoencoder (CAE) based sky image prediction models are proposed due to the spatiotemporal feature extraction ability of two-dimensional (2-D) CAEs and 3-D CAEs. For 2-D CAEs and 3-D CAEs, four architectures are given respectively. To verify the effectiveness of the proposed models, two typical DIPT methods, including particle image velocimetry and Fourier phase correlation theory are introduced to build the benchmark models. Besides, five different scenarios are also set and the results show that the proposed models outperform the benchmark models in all scenarios. [ABSTRACT FROM AUTHOR]

With the increase of solar photovoltaic (PV) penetration in power system, the impact of random fluctuation of PV power on the secure operation of power grid becomes more and more serious. High-precision PV power forecasting can effectively promote the grid's accommodation of PV power generation. Cloud is the most important factor affecting the surface irradiance and PV power. For the ultra-short-term solar PV power forecast considering the influence of cloud movement, it is necessary to be able to obtain the surface irradiance according to the sky cloud observation data. Therefore, in order to accurately achieve the real-time mapping relationship between sky image and surface irradiance, a hybrid mapping model based on deep learning applied for solar PV power forecasting is proposed in this article. First, the sky image data are clustered based on the feature extraction of convolutional autoencoder and K-means clustering algorithm after preprocess stage. Second, a hybrid mapping model based on deep learning methods are established for surface irradiance. Finally, the simulation results are compared and evaluated with different deep learning methods (CNN, LSTM, and ANN). The results show that the proposed model in this article has higher accuracy and can maintain robustness under different weather conditions. [ABSTRACT FROM AUTHOR]

The motion of cloud over a photovoltaic (PV) power station will directly cause the change of solar irradiance, which indirectly affects the prediction of minute-level PV power. Therefore, the calculation of cloud motion speed is very crucial for PV power forecasting. However, due to the influence of complex cloud motion process, it is very difficult to achieve accurate result using a single traditional algorithm. In order to improve the computation accuracy, a pattern classification and particle swarm optimization optimal weights based sky images cloud motion speed calculation method for solar PV power forecasting (PCPOW) is proposed. The method consists of two parts. First, we use a k-means clustering method and texture features based on a gray-level co-occurrence matrix to classify the clouds. Second, for different cloud classes, we build the corresponding combined calculation model to obtain cloud motion speed. Real data recorded at Yunnan Electric Power Research Institute are used for simulation; the results show that the cloud classification and optimal combination model are effective, and the PCPOW can improve the accuracy of displacement calculation. [ABSTRACT FROM AUTHOR]

Solar photovoltaic (PV) power forecasting has become an important issue with regard to the power grid in terms of the effective integration of large-scale PV plants. As the main influence factor of PV power generation, solar irradiance and its accurate forecasting are the prerequisite for solar PV power forecasting. However, previous forecasting approaches using manual feature extraction (MFE), traditional modeling and single deep learning (DL) models could not satisfy the performance requirements in partial scenarios with complex fluctuations. Therefore, an improved DL model based on wavelet decomposition (WD), the Convolutional Neural Network (CNN), and Long Short-Term Memory (LSTM) is proposed for day-ahead solar irradiance forecasting. Given the high dependency of solar irradiance on weather status, the proposed model is individually established under four general weather type (i.e., sunny, cloudy, rainy and heavy rainy). For certain weather types, the raw solar irradiance sequence is decomposed into several subsequences via discrete wavelet transformation. Then each subsequence is fed into the CNN based local feature extractor to automatically learn the abstract feature representation from the raw subsequence data. Since the extracted features of each subsequence are also time series data, they are individually transported to LSTM to construct the subsequence forecasting model. In the end, the final solar irradiance forecasting results under certain weather types are obtained via the wavelet reconstruction of these forecasted subsequences. This case study further verifies the enhanced forecasting accuracy of our proposed method via a comparison with traditional and single DL models. [ABSTRACT FROM AUTHOR]

For minute time scale solar photovoltaic (PV) power forecasting, the motion of clouds over PV power plant mainly contribute to the fluctuant and intermittent nature of solar PV power output. Therefore, research on cloud motion displacement (CMD) calculation to realize cloud motion prediction is a key sub-process for minute time scale solar PV power forecasting approaches. Fourier phase correlation theory (FPCT) is widely applied in CMD calculation for its superiority of simplicity and less computation, then an improved algorithm based on image-phase-shift-invariance (IPSI) is proposed to reduce the outlier probability of CMD results. However, at present, the current IPSI algorithm still has limitations and cannot avoid the occurrence of outliers altogether. In this paper, we presented a novel method, termed IPSI based multi-transform-fusion (MTF) method, to further improve the effectiveness compared with traditional FPCT and affine transform based IPSI method. First, three image transform methods satisfying IPSI condition, respectively wavelet transform (WT), affine transform (AT), and convolution transform (CT), are explored. Then the information increment of the transformed sky images using the above three methods is analyzed, respectively. Second, we determine the suitable image transform method for IPSI algorithm under specific cloud condition according to the corresponding information increment. Third, an IPSI based MTF method for CMD calculation in sky images is proposed. The original sky images are transformed through WT, AT, and CT to generate multiple images that maintain the same object motion information, then calculate the CMDs in each generated image. Finally, we apply Gaussian distribution to fit the multiple CMD values and taking its mathematical expectation as final CMD result. Various experimental results in 4 different scenarios show that the performance of the proposed approach is better than FPCT, AT based IPSI, and OF method, by reducing plenty of CMD outliers, thus delivering greater accuracy and robustness. [ABSTRACT FROM AUTHOR]

• The temporal distribution of irradiance is a mapping of cloud spatial distribution. • Irradiance data can be achieved using cloud position and RGB values in sky images. • Sky image-irradiance mapping model is updated according to current cloud conditions. • Minutely irradiance forecasting is achieved by a sky image-irradiance mapping model. Accurate minutely solar irradiance forecasting is the basis of minute-level photovoltaic (PV) power forecasting. In this paper, a minutely solar irradiance forecasting method based on real-time surface irradiance mapping model is proposed, which is beneficial to achieve higher accuracy in solar power forecasting. First, we extract the red–green–blue (RGB) values and position information of pixels in sky images after background elimination and distortion rectification, to explore the mapping relationship between sky image and solar irradiance. Then a real-time sky image-irradiance mapping model is built, trained, and updated according to real-time sky images and solar irradiance. Finally, the future solar irradiance within the time horizons varying from 1 min to 10 min ahead are capable to be forecasted by using the latest updated surface irradiance mapping model with extracted input from the current sky image. The average measures of proposed method by using MAPE, RMSE, MBE are 22.66%, 92.72, −1.26% for blocky clouds; 20.44%, 132.15, −1.06% for thin clouds and 18.82%, 120.78, −0.98% for thick clouds, thus deliver much higher forecasting accuracy than other benchmarks. [ABSTRACT FROM AUTHOR]

• Photovoltaic power presents volatility, annual periodicity, and adjacent similarity. • Deep learning and time correlation principles are combined in a forecasting problem. • Time correlation modification improves the accuracy of LSTM-RNN forecasting model. • Partial daily pattern prediction can be used to optimize the modification process. Photovoltaic (PV) power generation is an effective means to realize solar energy utilization. Due to the natural characteristics of random fluctuations in solar energy, the applications of PV power such as grid-connected PV power plant, distributed PVs, and building integrated PVs will introduce new characteristics to the generation and load side of the power grid. Therefore, accurate day-ahead PV power forecasting is of great significance for enabling grid manager to achieve PV power output data in advance and mitigate the influence of random fluctuations. To tackle the deficiencies of conventional artificial intelligence (AI) modeling methods such as overfitting problem and insufficient generalization ability to complex nonlinear modeling, a day-ahead PV power forecasting model assembled by fusing deep learning modeling and time correlation principles under a partial daily pattern prediction (PDPP) framework is proposed. First, an independent day-ahead PV power forecasting model based on long-short-term memory recurrent neural network (LSTM-RNN) is established. Second, a modification method is proposed to update the forecasting results of LSTM-RNN model based on time correlation principles regarding different patterns of PV power in the forecasting day. Third, a partial daily pattern prediction (PDPP) framework is proposed to provide accurate daily pattern prediction information of particular days, which is used to guide the modification parameters. Simulation results show that the proposed forecasting method with time correlation modification (TCM) is more accurate than the individual LSTM-RNN model, and the performance of the forecasting model can be further improved for those days with accurate daily pattern predictions under the proposed PDPP framework. [ABSTRACT FROM AUTHOR]

Accurate ultra-short-term PV power forecasting is essential for the power system with a high proportion of renewable energy integration, which can provide power fluctuation information hours ahead and help to mitigate the interference of the random PV power output. Most of the PV power forecasting methods mainly focus on employing local ground-based observation data, ignoring the spatial and temporal distribution and correlation characteristics of solar energy and meteorological impact factors. Therefore, a novel ultra-short-term PV power forecasting method based on the satellite image data is proposed in this paper, which combines the spatio-temporal correlation between multiple plants with power and cloud information. The associated neighboring plant is first selected by spatial-temporal cross-correlation analysis. Then the global distribution information of the cloud is extracted from satellite images as additional inputs with other general meteorological and power inputs to train the forecasting model. The proposed method is compared with several benchmark methods without considering the information of neighboring plants. Results show that the proposed method outperforms the benchmark methods and achieves a higher accuracy at 4.73%, 10.54%, and 4.88%, 11.04% for two target PV plants on a four-month validation dataset, in terms of root mean squared error and mean absolute error value, respectively. • The spatio-temporal correlation between PV power plants is investigated. • Introduce how to select an appropriate neighboring plant. • Verify the mapping relationship between cloud characteristics of neighboring plants and the power of target plants. • An ultra-short-term PV power forecasting method based on cloud information from neighboring plant is proposed. [ABSTRACT FROM AUTHOR]