Photovoltaic Power Data Research Articles

A method for configuring hybrid electrolyzers based on joint wind and photovoltaic power generation modeling using copula functions

Considering the specific wind and photovoltaic power characteristics of a certain region, this study investigates the optimal ratio of Alkaline Electrolysis Cells (AEL) to Proton Exchange Membrane (PEM) electrolyzers in a hybrid electrolysis system for hydrogen production. A flexible model for configuring the hybrid electrolysis system is proposed, based on a copula function for joint wind and solar power modeling. This model generates wind and photovoltaic power generation scenarios using the copula function, incorporating a selection mechanism to ensure that the output scenarios are more representative of the actual data characteristics of wind and photovoltaic power output. Consequently, considering both the fluctuation and amplitude, the wind and photovoltaic power data are decomposed using the Ensemble empirical mode decomposition method. The decomposed components are then allocated to the two types of electrolyzers. Furthermore, the optimal configuration of the hybrid electrolysis system is determined by minimizing the costs associated with wasted power, electricity purchases, and other expenses. Finally, a case study of a 100 MW wind farm and a 50 MW photovoltaic power station in Northwest China is presented, concluding that the optimal configuration ratio of AEL to PEM electrolyzers is 2:1. In a Matlab/Simulink platform, the performance metrics of the hybrid electrolysis system were validated. It was found that the hydrogen production rate of the hybrid electrolyzer is comparable to that of the PEM electrolyzer, but with a lower required cost. Additionally, the hydrogen production rate and volume of the optimal configuration for the hybrid electrolyzer determined by the model proposed in this paper are higher than those obtained through the optimization algorithm's optimal configuration.

Distributed photovoltaic power forecasting based on personalized federated adversarial learning

Existing distributed photovoltaic (PV) power forecasting methods fail to address the impact of sample scarcity and heterogeneity in PV power data. Moreover, training a single prediction model proves challenging to meet the personalized forecasting needs of different PV stations in distributed environments. This paper proposes a personalized federated generative adversarial network (PFedGAN)-based DPV power forecasting method. Leveraging the federated learning (FL) framework, it achieves collaborative training of prediction models among DPV stations while preserving data privacy. y introducing generative adversarial networks (GAN) and personalized strategy optimization into the FL training process, it alleviates data scarcity issues and reduces the impact of data heterogeneity. A TimesNet-DeepAR (TNE-DeepAR) power prediction model is designed, where the TimesNet module extracts correlations between PV power data from different time periods, and the DeepAR module facilitates PV power prediction, mitigating the effects of meteorological factors' multi-periodic variations on PV power. Experimental results show that the proposed hybrid prediction model reduces the average mean absolute percentage error (MAPE) by 30–40 % compared to single models. The proposed approach reduces the MAPE by 9.79 % compared to traditional methods and by 49.62 % for PV stations with scarce data.

Multi-scale RWKV with 2-dimensional temporal convolutional network for short-term photovoltaic power forecasting

Improving the accuracy of Photovoltaic (PV) power forecasting is crucial for optimizing the schedule of power stations and maintaining the grid stability. However, PV power generation exhibits complex periodicity and is significantly influenced by weather conditions, introducing instability, intermittency, and randomness, making accurate PV power forecasting a challenging task. Therefore, this study proposes a multi-scale Receptance Weighted Key Value with 2-Dimensional Temporal Convolutional Network (MSRWKV-2DTCN) for PV power forecasting, which can learn periodicity and interdependencies of data and improve forecasting accuracy. Firstly, the proposed model identifies multi-periodicity of PV power data with the Fast Fourier Transform (FFT). Subsequently, we combine these identified periods with the canonical time mixing block of Receptance Weighted Key Value (RWKV) and introduce a multi-scale time mixing block to learn periodicity of data. Finally, to explore complex interdependencies of historical data, we replace the channel-mixing block of RWKV with a multi-scale 2-Dimensional Temporal Convolutional Network (2D TCN). Experiments were conducted on real-world datasets collected from Yulara solar system in Australia to validate the performance of the proposed model. Comparisons with other PV power forecasting models and ablation studies confirm that the MSRWKV-2DTCN achieves higher accuracy in short-term PV power forecasting.

Overview of Wind and Photovoltaic Data Stream Classification and Data Drift Issues

The development in the fields of clean energy, particularly wind and photovoltaic power, generates a large amount of data streams, and how to mine valuable information from these data to improve the efficiency of power generation has become a hot spot of current research. Traditional classification algorithms cannot cope with dynamically changing data streams, so data stream classification techniques are particularly important. The current data stream classification techniques mainly include decision trees, neural networks, Bayesian networks, and other methods, which have been applied to wind power and photovoltaic power data processing in existing research. However, the data drift problem is gradually highlighted due to the dynamic change in data, which significantly impacts the performance of classification algorithms. This paper reviews the latest research on data stream classification technology in wind power and photovoltaic applications. It provides a detailed introduction to the data drift problem in machine learning, which significantly affects algorithm performance. The discussion covers covariate drift, prior probability drift, and concept drift, analyzing their potential impact on the practical deployment of data stream classification methods in wind and photovoltaic power sectors. Finally, by analyzing examples for addressing data drift in energy-system data stream classification, the article highlights the future prospects of data drift research in this field and suggests areas for improvement. Combined with the systematic knowledge of data stream classification techniques and data drift handling presented, it offers valuable insights for future research.

Ultra-short-term photovoltaic power prediction based on similar day clustering and temporal convolutional network with bidirectional long short-term memory model: A case study using DKASC data

Due to its strong dependence on weather conditions, photovoltaic (PV) power is highly intermittent in nature. In light of this, this paper introduces a PV power prediction model based on similar-day clustering and temporal convolutional network (TCN) with bidirectional long short-term memory (BiLSTM) model. Firstly, in the data preprocessing stage, improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN) method is used to decompose and denoise the original PV power data to obtain subsequences of different frequencies. Subsequently, the sample entropy (SE) method is incorporated to reconstruct the subsequences and generate low-frequency and high-frequency subsequences with more pronounced temporal features. In addition, a self-organizing map (SOM) with the K-means algorithm is used to categorize historical days into four weather types: sunny, cloudy, rainy, and intermittent. Through grey relational analysis (GRA), meteorological factors significantly affecting PV power prediction are identified to construct a multidimensional feature set. During the model prediction phase, historical PV power data and related meteorological impact factors are input into the TCN-BiLSTM model to perform multi-data-driven PV power prediction. Finally, the predicted results of each subsequence are linearly combined to obtain the ultimate PV power prediction. The proposed model is compared with convolutional neural network (CNN)-LSTM, LSTM, and TCN models, achieving a reduction in the root mean square error (RMSE) of 0.129 kW, 0.238 kW, and 0.257 kW, respectively. This demonstrates that the proposed model exhibits superior overall performance in time series modeling and information capture, enhancing the understanding of seasonal, periodic, and irregular patterns in PV power generation data.

Prediction of rainy-day photovoltaic power generation based on Generative Adversarial Networks and enhanced Sparrow Search Algorithm

Accurate and timely photovoltaic (PV) power forecasting is crucial for the stable operation of power systems. To address the issue of sparse PV power data on rainy days, this paper proposes the use of Wasserstein Generative Adversarial Networks with Gradient Penalty (WGAN-GP) to augment rainy-day data. A progressive gradient penalty strategy is introduced to avoid gradient vanishing. A Bidirectional Long Short-Term Memory network is employed for power forecasting, with adaptive hyperparameter optimization achieved through an improved Sparrow Search Algorithm (LSSA). The optimization capability is enhanced using a chaotic mapping strategy. Validation on data from a 1 MW PV plant demonstrates that the LSSA-optimized model achieves the best forecasting results. After augmenting the rainy-day dataset, the Mean Absolute Error and Root Mean Square Error decreased by 15.3 % and 6.15 %, respectively, indicating that WGAN-GP significantly improves forecasting accuracy on rainy days.

Parameter Estimation and Preliminary Fault Diagnosis for Photovoltaic Modules Using a Three-Diode Model

Accurate estimation of photovoltaic (PV) power generation can ensure the stability of regional voltage control, provide a smooth PV output voltage and reduce the impact on power systems with many PV units. The internal parameters of solar cells that affect their PV power output may change over a period of operation and must be re-estimated to produce a power output close to the actual value. To accurately estimate the power output for PV modules, a three-diode model is used to simulate the PV power generation. The three-diode model is more accurate but more complex than single-diode and two-diode models. Different from the traditional methods, the 9 parameters of the three-diode model are transformed into 16 parameters to further provide more refined estimates. To accurately estimate the 16 parameters in the model, an optimization tool that combines enhanced swarm intelligence (ESI) algorithms and the dynamic crowing distance (DCD) index is used based on actual historical PV power data and the associated weather information. When the 16 parameters for a three-diode model are accurately estimated, the I–V (current-voltage) curves for different solar irradiances are plotted, and the possible failures of PV modules can be predicted at an early stage. The proposed method is verified using a 200 kWp PV power generation system. Three different diode models that are optimized using different ESI algorithms are compared for different weather conditions. The results affirm the reliability of the proposed ESI algorithms and the value of creating more refined estimation models with more parameters. Preliminary fault diagnosis results based on the differences between the actual and estimated I–V curves are provided to operators for early maintenance reference.

Short-Term Prediction of Rural Photovoltaic Power Generation Based on Improved Dung Beetle Optimization Algorithm

Addressing the challenges of randomness, volatility, and low prediction accuracy in rural low-carbon photovoltaic (PV) power generation, along with its unique characteristics, is crucial for the sustainable development of rural energy. This paper presents a forecasting model that combines variational mode decomposition (VMD) and an improved dung beetle optimization algorithm (IDBO) with the kernel extreme learning machine (KELM). Initially, a Gaussian mixture model (GMM) is used to categorize PV power data, separating analogous samples during different weather conditions. Afterwards, VMD is applied to stabilize the initial power sequence and extract numerous consistent subsequences. These subsequences are then employed to develop individual KELM prediction models, with their nuclear and regularization parameters optimized by IDBO. Finally, the predictions from the various subsequences are aggregated to produce the overall forecast. Empirical evidence via a case study indicates that the proposed VMD-IDBO-KELM model achieves commendable prediction accuracy across diverse weather conditions, surpassing existing models and affirming its efficacy and superiority. Compared with traditional VMD-DBO-KELM algorithms, the mean absolute percentage error of the VMD-IDBO-KELM model forecasting on sunny days, cloudy days and rainy days is reduced by 2.66%, 1.98% and 6.46%, respectively.

Application of artificial neuronal networks for the prediction of monthly photovoltaic power using tilt angle as input in Rabat city, Morocco

The use of artificial neural networks (ANNs) is attracting interest in the field of renewable energies, particularly for predicting operational parameters of photovoltaic (PV) systems. Thus, the present study aims to develop an ANN model for predicting the output power of a photovoltaic module. In this context, two ANN models with two and three hidden layers, respectively, were trained and validated to predict the average monthly photovoltaic power Pp produced by a 0. 325 kW photovoltaic module by varying the tilt angle from 0° to 90° of Rabat city, Morocco. The training data used in this study include longitude, latitude, calculated photovoltaic power, tilt angle, and months as inputs, while the predicted Pp is the target. These two proposed models are trained using the Levenberg-Marquardt (LM) optimiser. Furthermore, to confirm the reliability of each of these two proposed ANN models, existing photovoltaic power data were used to validate the two proposed models by predicting the average monthly power of the photovoltaic module. For the tilt angle of 34°, which corresponds to the latitude of the city of Rabat, the ANN model with three hidden layers performed better with an R2 value of 0.9997.

A novel learning approach for short-term photovoltaic power forecasting - A review and case studies

Integrating photovoltaic power into the power system can offer significant economic and environmental benefits. However, the intermittent and random nature of photovoltaic power generation poses a challenge to the current power system's planning and operation. Accurate photovoltaic power generation forecasting is crucial for delivering high-quality electric energy to consumers and increasing system reliability.In this study, a multi-stage approach is proposed. In the first stage, three decomposition techniques (Iterative Filtering decomposition, Variational Mode Decomposition, and Wavelet Packet Decomposition) are employed for time series decomposition. Then, for each Intrinsic Mode Function (IMF) component resulting from the decomposition block, five machine learning and three deep learning algorithms are utilized, serving as local forecasting models. In the final forecast phase, the best forecasting result for each regressor is selected during the reconstruction phase. Two years of photovoltaic power data recorded in three grid-connected photovoltaic systems installed in South Algeria were utilized for training and testing the proposed forecasting models. Upon comprehensive analysis and examination of the outcomes, the proposed method exhibits the lowest normalized Root Mean Square Error values across all forecast horizons and monitoring stations. Particularly, for forecasting steps at time intervals +1, +3, and +5, the proposed method attains an average normalized Root Mean Square Error, showcasing its efficacy: 0.709%, 2.097%, and 3.241% for station 1; 1.147%, 3.546%, and 5.347% for station 2; and 0.922%, 2.158%, and 4.539% for station 3. The experimental results underscore the superiority of our approach over conventional regression algorithms, thus substantiating its prowess in delivering robust and competitive performance outcomes.

Short-term PV power data prediction based on improved FCM with WTEEMD and adaptive weather weights

Photovoltaic (PV) systems are commonly used in zero energy buildings(ZEBs) due to their high efficiency and convenience. However, PV systems are affected by meteorological factors with nonlinearities and stochasticity in the power generation process, which leads to inaccurate prediction of PV power, affects the relationship between energy supply and user demand, and then has a significant impact on the stability of PV grid connection. To address these challenges, this paper proposes a WTEEMD-FCM-IGWO-LSTM method for the numerical prediction of PV power. Firstly, for the noise interference in the meteorological data of PV power generation collected from PV power stations, this paper proposes an Ensemble Empirical Mode Decomposition (EEMD) method based on the wavelet threshold algorithm improvement (WTEEMD) for noise reduction and reconstruction of meteorological data to enhance the signal-to-noise ratio in the data. Secondly, considering the stochastic nature of unstable meteorological factors, an improved fuzzy C-mean clustering (FCM) method is employed to classify the PV power process dataset and enhance the correlation between meteorological factors and power data. Subsequently, prediction models for the power values of the PV power generation process under each of the three meteorological conditions are developed using Long Short-Term Memory (LSTM) based on the classified sample data. Finally, the proposed model is evaluated through simulation using historical Australian PV data. The experimental results show that it is possible to accurately predict the power of PV power generation, improve the utilization of clean energy, and support the sustainable development of ZEBs.

Hourly photovoltaic power prediction based on signal decomposition and deep learning

Accurate photovoltaic (PV) power prediction is important for the utilization of solar energy resources. However, PV power is non-stationary due to the variable influence of meteorological factors, which poses a challenge for accurate forecasting. In this paper, a hybrid method based on signal decomposition and a deep learning model is proposed. The hybrid model integrates complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and the Informer model. The CEEMDAN algorithm is used to separate different modes from the photovoltaic power sequence, enhancing its predictability. The deep learning model, the Informer, is employed to capture the complex relationship between photovoltaic power data and its historical data as well as external meteorological factors, ultimately enabling multi-step forecasting of photovoltaic power data. In hourly PV power forecasting experiments using a public dataset, the model exhibits significant performance improvements when compared to benchmark models such as LSTM, GRU, and Transformer. Specifically, the RMSE reduces by 6.07%-34.74% and the MAE reduces by 7.07%-37.5%. The results demonstrate that the hybrid model exhibits accurate predictive performance in the task of hourly photovoltaic power forecasting.

Short-term prediction of photovoltaic power generation based on ICEEMDAN-SE-GAPSO-LSTM

This study proposes an improved adaptive noise-complete ensemble empirical mode decomposition (ICEEMDAN), sample entropy (SE), and genetic particle swarm optimization (GAPSO) algorithm-based short-term solar power forecast model. The photovoltaic power data is first decomposed using the ICEEMDAN algorithm to produce a number of eigenmode components with various properties. The sample entropy of each component is then calculated, and the components with similar sample entropy values are combined to create new modal components. These new modal components are then input into the GAPSO-optimized LSTM model for prediction, and the prediction results of each component are summed and reconstructed to produce the final prediction. The results completely validate the efficacy and dependability of the suggested method, using the measured data from a photovoltaic power station in Xinjiang, China as an example.

Photovoltaic Power Prediction Based on Hybrid Deep Learning Networks and Meteorological Data.

Conventional point prediction methods encounter challenges in accurately capturing the inherent uncertainty associated with photovoltaic power due to its stochastic and volatile nature. To address this challenge, we developed a robust prediction model called QRKDDN (quantile regression and kernel density estimation deep learning network) by leveraging historical meteorological data in conjunction with photovoltaic power data. Our aim is to enhance the accuracy of deterministic predictions, interval predictions, and probabilistic predictions by incorporating quantile regression (QR) and kernel density estimation (KDE) techniques. The proposed method utilizes the Pearson correlation coefficient for selecting relevant meteorological factors, employs a Gaussian Mixture Model (GMM) for clustering similar days, and constructs a deep learning prediction model based on a convolutional neural network (CNN) combined with a bidirectional gated recurrent unit (BiGRU) and attention mechanism. The experimental results obtained using the dataset from the Australian DKASC Research Centre unequivocally demonstrate the exceptional performance of QRKDDN in deterministic, interval, and probabilistic predictions for photovoltaic (PV) power generation. The effectiveness of QRKDDN was further validated through ablation experiments and comparisons with classical machine learning models.

Short-Term Photovoltaic Output Prediction Based on Decomposition and Reconstruction and XGBoost under Two Base Learners

Photovoltaic power generation prediction constitutes a significant research area within the realm of power system artificial intelligence. Accurate prediction of future photovoltaic output is imperative for the optimal dispatchment and secure operation of the power grid. This study introduces a photovoltaic prediction model, termed ICEEMDAN-Bagging-XGBoost, aimed at enhancing the accuracy of photovoltaic power generation predictions. In this paper, the original photovoltaic power data initially undergo decomposition utilizing the Improved Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (ICEEMDAN) algorithm, with each intrinsic mode function (IMF) derived from this decomposition subsequently reconstructed into high-frequency, medium-frequency, and low-frequency components. Targeting the high-frequency and medium-frequency components of photovoltaic power, a limiting gradient boosting tree (XGBoost) is employed as the foundational learner in the Bagging parallel ensemble learning method, with the incorporation of a sparrow search algorithm (SSA) to refine the hyperparameters of XGBoost, thereby facilitating more nuanced tracking of the changes in the photovoltaic power’s high-frequency and medium-frequency components. Regarding the low-frequency components, XGBoost-Linear is utilized to enable rapid and precise prediction. In contrast with the conventional superposition reconstruction approach, this study employs XGBoost for the reconstruction of the prediction output’s high-frequency, intermediate-frequency, and low-frequency components. Ultimately, the efficacy of the proposed methodology is substantiated by the empirical operation data from a photovoltaic power station in Hebei Province, China. Relative to integrated and traditional single models, this paper’s model exhibits a markedly enhanced prediction accuracy, thereby offering greater applicational value in scenarios involving short-term photovoltaic power prediction.

PV output estimation method of power distribution station area based on federated learning framework and improved transformer neural network

In contemporary distribution networks (DNs), characterized by extensive integration of distributed energy, the photovoltaic (PV) power output data from the distribution station areas become crucial for system planning and operational optimization. Since many PVs are installed behind-the-meter (BTM), it is difficult to directly obtain PV power data through measurement devices. Therefore, it is important to estimate the BTM PV power from the aggregating data that can be directly obtained. However, the existing estimation methods usually require centralized large-scale data training, which brings certain privacy leakage risks. In order to solve these problems, we propose a federated learning-based improved Transformer Neural Network strategy to estimate BTM PV generation at the community level with data privacy protection. Initially, enhanced Transformer neural networks, employing a fused-attention mechanism, are deployed to precisely delineate the solar power generation pattern. Subsequently, federated learning principles facilitate the sharing of specific parameters among multiple edge endpoints and a central server. This model bifurcates into two layers: an individual layer, where parameters are retained locally, and an exchange layer, where parameters are collectively shared and conveyed through momentum aggregation. This dual-layer structure effectively synchronizes the capture of both unique and common characteristics. The test on the Australian residential load dataset verifies the effectiveness of the proposed method.

Ultra-short-term forecast of distributed photovoltaic power based on satellite cloud image and LSTM model

Photovoltaic (PV) power generation, with its volatile, intermittent, and random characteristics, and large-scale PV access pose a threat to grid stability. For this reason, predicting the photovoltaic output will help keep the grid safe and stable. On the basis of the influence of cloud groups on solar radiation, a very short-term forecast of distributed PV energy will be made using satellite cloud picture information to improve the forecast accuracy of PV energy production. The paper presents a method to predict distributed PV power at very short notice based on satellite clouds and a network model with Long Short-Term Memory (LSTM). First, extract a subset of meteorological and PV power data from the forecast area as training samples., and the abnormal part of the samples is cleaned by an isolated forest algorithm. Secondly, the occlusion feature is extracted from the satellite cloud image in the same period. Finally, the measured solar irradiance, meteorological information, and obscuration features are input into the LSTM network for prediction, and the photovoltaic power prediction results in the next 4 hours are obtained. The measured PV power of Jinghai Guangfu Power Station in Hefei, Anhui province on the 5th day was the training sample for the prediction of PV power on the 6th day. The prediction results show that the prediction error is 2.73% when a satellite cloud image is added, and 16.15% when a satellite cloud image is not added, and the prediction error is reduced by 13.42%.

Forecast-Based Energy Management for Optimal Energy Dispatch in a Microgrid

This article describes the development of an optimal and predictive energy management system (EMS) for a microgrid with a high photovoltaic (PV) power contribution. The EMS utilizes a predictive long-short-term memory (LSTM) neural network trained on real PV power and consumption data. Optimal EMS decisions focus on managing the state of charge (SoC) of the battery energy storage system (BESS) within defined limits and determining the optimal power contributions from the microgrid components. The simulation utilizes MATLAB R2023a to solve a mixed-integer optimization problem and HOMER Pro 3.14 to simulate the microgrid. The EMS solves this optimization problem for the current sampling time (t) and the immediate sampling time (t+1), which implies a prediction of one hour in advance. An upper-layer decision algorithm determines the operating state of the BESS, that is, to charge or discharge the batteries. An economic and technical impact analysis of our approach compared to two EMSs based on a pure economic optimization approach and a peak-shaving algorithm reveals superior BESS integration, achieving 59% in demand satisfaction without compromising the life of the equipment, avoiding inexpedient power delivery, and preventing significant increases in operating costs.

Multiphase Solar Photovoltaic Prediction Model Based on Season, Hierarchical k‐Means Clustering, GRA‐PCC, SVM, and Neural Network

Solar photovoltaic (PV) has accounted for the highest percentage of power generation capacity among other renewables. However, solar PV power generation is highly variable because of different factors; therefore, accurate forecasting is critical for reliable integration into the power system. This paper proposes a multiphase solar PV prediction model that includes grouping, clustering, linking, classifying, and predicting using historical solar PV power and weather data. Seasonal variation is considered in the grouping phase, followed by hybrid hierarchical k‐means clustering to enhance data division in the clustering phase. A hybrid gray relational analysis‐Pearson correlation coefficient identifies significant weather factors impacting solar PV power in the linking phase. The classification phase employs a support vector machine to establish the relationship between the clusters and the relevant weather factors. Lastly, a neural network (NN) is trained to predict solar PV power. The solar PV power profiles are presented to show the variability in season and time. The simulation results of the proposed model showed relatively accurate forecasting results, including MAE of 0.408 MW, MSE of 460.51 MW, RMSE of 0.679 MW, nRMSE of 4.345%, and MRE of 2.266%. These results represent that the uncertainties of the proposed model are 6 and 12 times lower than those of the conventional methods (i.e., conventional NN and ARMAX). These results assure that the proposed model can provide more accurate solar PV power profiles for reliable power system integration.

Zero-Shot Photovoltaic Power Forecasting Scheme Based on a Deep Learning Model and Correlation Coefficient

Accurate photovoltaic (PV) power forecasting is essential for the stable and reliable operation of PV power generation systems. Recently, various deep learning- (DL-) based forecasting models have been proposed for accurate forecasting, but newly built systems cannot benefit from them due to the absence of PV power data. Although zero-shot methods based on single site can be used for PV power forecasting, they suffer from performance degradation problems when the characteristics of the source data and target data are different. To address this issue, we propose a novel zero-shot PV power forecasting scheme that leverages historical data from multiple PV generation systems at different sites. The proposed scheme constructs an individual forecasting model using historical data from each PV generation system. Then, two correlation coefficients are calculated for each forecasting model: one based on the correlation between the input variables of the source data and target data and the other on the correlation between the input variables and output variables of the source data. Lastly, the final forecasting value is calculated as a weighted sum of the predicted values of the constructed forecasting models for the input variables of the target data. In the extensive experiments for diverse DL models for forecasting, correlation coefficient types for weights, and data time intervals, the combination of recurrent neural network, Pearson’s correlation coefficient, and solar-noon time yielded the best prediction performance, with an improvement of up to 34.47% in mean absolute error and up to 15.94% in root mean square error compared to the best single-site zero-shot prediction. In addition, in experiments on PV power data from 9 cities in Korea using this combination, the proposed scheme achieved the best predictive performance in almost all cases and the second-best performance with a very narrow margin only in a few cases.