Solar Power Forecasting Research Articles

Design and Development of a Short-Term Photovoltaic Power Output Forecasting Method Based on Random Forest, Deep Neural Network and LSTM Using Readily Available Weather Features

Renewable energy sources (RES) are an essential part of building a more sustainable future, with higher diversity of clean energy, reduced emissions and less dependence on finite fossil fuels such as coal, oil and natural gas. The advancements in the renewable energy sources domain bring higher hardware efficiency and lower costs, which improves the likelihood of wider RES adoption. However, integrating renewables such as photovoltaic (PV) systems in the current grid is still a major challenge. The main reason is the volatile, intermittent nature of RES, which increases the complexity of the grid management and maintenance. Having access to accurate PV power output forecasting could reduce the number of power supply disruptions, improve the planning of the available and reserve capacities and decrease the management and operational costs. In this context, this paper explores and evaluates three Artificial Intelligence (AI) methods - random forest (RF), deep neural network (DNN) and long short-term memory network (LSTM), which are applied for the task of short-term PV output power forecasting. Following a statistical forecasting approach, the selected models are trained on weather and PV output data collected in Berlin, Germany. The assembled data set contains predominantly broadly accessible weather features, which makes the proposed approach more cost efficient and easily applicable even for geographic locations without access to specialized hardware or hard-to-obtain input features. The performance achieved by two of the selected algorithms indicates that the RF and the DNN models are able to generate accurate solar power forecasts and are also able to handle sudden changes and shifts in the PV power output.

A Hybrid Machine Learning Model for Solar Power Forecasting

The paper presents a near investigation of different AI procedures for solar power forecasting. The objective of the research is to identify the most accurate and efficient machine learning algorithms for solar power forecasting. The paper also considers different parameters such as weather conditions, solar radiation, and time of day in the forecasting model. This paper proposes a hybrid machine learning model for solar power forecasting that consolidates the strengths of multiple algorithms, including support vector regression, random forest regression, and artificial neural network. However, the study also highlights the importance of incorporating domain knowledge and feature engineering in machine learning models for better forecasting accuracy.

An Empirical Analysis of Machine Learning Algorithms for Solar Power Forecasting in a High Dimensional Uncertain Environment

In the fabric of energy generation, solar power is the most promising clean energy solution as an alternative to non-renewable energy sources. However, solar power’s dependency on environmental factors adds uncertainty to energy production. In such a scenario, solar power forecasting provides an edge to mitigate this uncertainty and improves overall system stability. Recently, machine learning (ML) models have been extensively deployed for designing and forecasting solar power. However, data pre-processing, forecast horizon, and performance evaluation of ML algorithms have to be carefully evaluated to find an accurate model. This paper provides an empirical comparison of different generation ML models for solar power forecasting, which can help understand future research on which method to adopt, depending on the ML model’s strengths and weaknesses. Therefore, an effective forecasting method is designated in aspects such as performance errors, convergence time, and computational complexity. So, this work rates different ML models on error performance metrics and convergence time. Moreover, cross-fold validations and hyperparameter are also examined for the top five performing models for a comprehensive evaluation and to give more intuitive and calibrated insight into various stakeholders working in the solar power plant modeling field.

太陽光発電出力予測技術に関するレビュー

A review of photovoltaic (PV) power output forecasting technology is presented based on the latest information in the literature. In addition to short-term ahead prediction and day ahead prediction techniques using numerical weather prediction models, the challenges of forecasting techniques caused by factors other than weather forecasting (dust and snow) are discussed. Uncertainties in forecasts using ensemble forecasting are also discussed. Case studies are presented on the impact assessment of various solar power forecasts. It also introduces the available data for climate change adaptation in anticipation of future climate change. In addition, the trends of related projects to improve forecasting technology in Japan and abroad are described. The challenges of PV output forecasting technology are itemized.

Online updating of a Markovian solar forecast representation

Risk management is an essential task in power systems operations, and temporal variability of renewable energy resources requires new techniques to capture the effects of such variability on the system. We utilize use a Markovian representation of probabilistic solar power forecasts to enhance the temporal variability assessment of balancing area-wide solar power production. Using this representation, we determine the optimal amount of historical forecast and observational data to use when updating the Markov transition matrix – which captures the past variability of the observations with respect to the forecasts – for use in stochastic optimization problems. We propose two novel multivariate scoring metrics based upon the observation vector’s band depth and discuss the performance with respect to these two metrics, as well as the variogram-based score, for multiple transition matrices. The optimal amount of data to use for these online transition matrices depends upon the season and likely upon the prevailing weather regime. • Transition matrices updated by ranking recent solar observations among forecasts. • Introduced two novel transition matrix scoring metrics based on band depth. • Fewer transitions in the spring and summer months improve transition matrices. • Compared changes in reserves and Area Control Error with respect to a fixed matrix. • Using an online matrix improved the Area Control Error often substantially.

Evolution of the shape parameters of photovoltaic module as a function of temperature and irradiance: New method of performance prediction without setting reference conditions

The accurate characterization and prediction of current–voltage (I–V) characteristics of photovoltaic (PV) modules under different weather conditions are essential for solar power forecasting and ensuring grid stability. This paper proposed a novel method based on the power-law model (PLM) for estimating PV module performance under different weather conditions without setting reference conditions. The effects of solar irradiance, ambient temperature, and module types are all fully considered. The dependence of shape parameters in PLM on irradiance and temperature are thoroughly investigated and the effect of the selection of reference conditions is eliminated by modifying a set of new transforming equations. The parameters of the new transforming equations can be extracted from experimental data by an optimization algorithm. Due to the advantages of PLM, the proposed method applies to any type of PV module and the I–V characteristics can be expressed explicitly without Lambert W-function or iterative solution. The effectiveness and accuracy proposed method are verified and tested on the large datasets of eighteen PV modules in three locations. Compared with existing methods, the proposed method shows higher accuracy and better performance in the estimation of I–V characteristics and maximum power under different weather conditions.

Recent developments in multivariate wind and solar power forecasting

AbstractThe intermittency of renewable energy sources, such as wind and solar, means that they require reliable and accurate forecasts to integrate properly into energy systems. This review introduces and examines a selection of state‐of‐the‐art methods that are applied for multivariate forecasting of wind and solar power production. Methods such as conditional parametric and combined forecasting already see wide use in practice, both commercially and scientifically. In the context of multivariate forecasting, it is vital to model the dependence between forecasts correctly. In recent years, reconciliation of forecasts to ensure coherency across spatial and temporal aggregation levels has shown great promise in increasing the accuracy of renewable energy forecasts. We introduce the methodology used for forecast reconciliation and review some recent applications for wind and solar power forecasting. Many forecasters are beginning to see the benefit of the greater information provided by probabilistic forecasts. We highlight stochastic differential equations as a method for probabilistic forecasting, which can also model the dependence structure. Lastly, we discuss forecast evaluation and how choosing a proper approach to evaluation is vital to avoid misrepresenting forecasts.This article is categorized under: Climate and Environment > Net Zero Planning and Decarbonization Sustainable Energy > Solar Energy Sustainable Energy > Wind Energy

Cloud Nowcasting with Structure-Preserving Convolutional Gated Recurrent Units

Nowcasting of clouds is a challenging spatiotemporal task due to the dynamic nature of the atmosphere. In this study, the use of convolutional gated recurrent unit networks (ConvGRUs) to produce short-term cloudiness forecasts for the next 3 h over Europe is proposed, along with an optimisation criterion able to preserve image structure across the predicted sequences. This approach is compared against state-of-the-art optical flow algorithms using over two and a half years of observations from the Spinning Enhanced Visible and Infrared Imager (SEVIRI) instrument onboard the Meteosat Second Generation satellite. We show that the ConvGRU trained using our structure-preserving loss function significantly outperforms the optical flow algorithms with an average change in R2, mean absolute error and structural similarity of 12.43%, −8.75% and 9.68%, respectively, across all time steps. We also confirm that merging multiple optical flow algorithms into an ensemble yields significant short-term performance increases (<1 h), and that nowcast skill can vary significantly across different European regions. Furthermore, our results show that blurry images resulting from using globally oriented loss functions can be avoided by optimising for structural similarity when producing nowcasts. We thus showcase that deep-learning-based models using locally oriented loss functions present a powerful new way to produce accurate cloud nowcasts, with important applications to be found in solar power forecasting.

Parametric Probabilistic Forecasting of Solar Power With Fat-Tailed Distributions and Deep Neural Networks

The need of solar power uncertainty quantification in the power system has inspired probabilistic solar power forecasting. This paper proposes a novel multi-step parametric method for intra-day probabilistic solar power forecasting. First, statistical analysis on solar power distribution is done using four forecasting methods in real-world data. Fat tails are clearly found in solar power distribution, which could not be modelled by the widely-used normal distribution. In light of this discovery, two fat-tailed distributions, i.e., Laplace and two-sided power distributions, along with their generalized variants are then proposed to better model the conditional distribution of solar power output. Second, a recently proposed DeepAR model for time series probabilistic forecasting based on deep recurrent neural network is used to map various predictors into parameters of the fat-tailed distribution. Moreover, a novel loss function based on the continuous ranked probability score is proposed, and its closed-form formula over the proposed fat-tailed distributions is derived for efficient model training. Numerical results on public real-world data show that our method is very effective and the proposed model can provide intra-day probabilistic solar power forecasting with high quality and reliability.

Evaluation of artificial neural network methods to forecast short-term solar power generation: a case study in Eastern Mediterranean Region

Solar power forecasting is substantial for the utilization, planning, and designing of solar power plants. Global solar irradiation (GSI) and meteorological variables have a crucial role in solar power generation. The ever-changing meteorological variables and imprecise measurement of GSI raise difficulties for forecasting photovoltaic (PV) output power. In this context, a major motivation appears for the accurate forecast of GSI to perform effective forecasting of the short-term output power of a PV plant. The presented study comprises of four artificial neural network (ANN) methods; recurrent neural network (RNN) method, feedforward backpropagation neural network (FFBPNN) method, support vector regression (SVR) method, and long short-term memory (LSTM) for daily total GSI prediction of Tarsus by using meteorological data. Moreover, this study proposes a model that utilizes the predicted daily GSI for output power forecasting of a grid-connected PV plant. The obtained results are compared with the output power generation data of a 350 kW solar power plant. The results are evaluated with the performance indices as mean absolute percentage error (MAPE), normalized root mean squared error (NRMSE), weighted mean absolute error (WMAE), and normalized mean absolute error (NMAE). FFBPNN method is chosen with the best results of MAPE 7.066%, NMAE 3.629%, NRMSE 4.673%, and WMAE 5.256%.

Summarizing ensemble NWP forecasts for grid operators: Consistency, elicitability, and economic value

On the one hand, grid integration of solar and wind power often requires just point (as opposed to probabilistic) forecasts at the individual plant level to be submitted to grid operators. On the other hand, solar and wind power forecasting can benefit greatly from dynamical ensemble forecasts from numerical weather prediction (NWP) models. Combining these two facts, this study is concerned with drawing out point forecasts from NWP ensembles. The scoring function for penalizing bad forecasts (or equivalently, rewarding good forecasts), in most scenarios, is specified by grid operators ex ante. The optimal point forecast therefore should be an elicitable functional of the predictive distribution, for which the specified scoring function is strictly consistent. Stated differently, the optimal way to summarize a predictive distribution depends on how the point forecast is to be penalized. Using solar irradiance forecasts issued by the ECMWF’s Ensemble Prediction System, the statistical theory on consistency and elicitability is validated empirically with extensive data. The results show that the optimal point forecasts elicited from ensembles have constantly higher accuracy than the best-guess forecasts, regardless of the choice of scoring function. Surprisingly, however, the correspondence between the two types of goodness of forecasts, namely, quality and value, is neither linear nor monotone, but depends on the penalty triggers and schemes specified by grid operators. In other words, using the optimally elicited forecasts, in many scenarios, would lead to lower economic values.

Evaluation of opaque deep-learning solar power forecast models towards power-grid applications

Solar photovoltaic power plays a vital role in global renewable energy power generation, and an accurate solar power forecast can further promote applications in integrated power systems. Due to advanced artificial intelligence technologies, various deep-learning models have been developed with the benefits of improved prediction precision, but these models inevitably sacrifice their interpretability compared to linear methods. Since a 100% accurate forecast is impossible to achieve, an opaque black-box model will always raise doubts for the operators of renewable power-grids, especially when the prediction deviation may produce higher economic costs and even a system turbulence. Motivated by this, the present study summarizes the requirements of deep-learning solar power forecast models from the power-grid application perspective. Post-hoc evaluation and discussion are conducted to analyze the performances of a typical deep-learning benchmark model based on open-access dataset for solar forecasting. Based on the results, the aim of this study is to increase confidence of deep-learning-based intelligent models into the practical engineering utilization of solar power forecasting. The case studies indicate that some simple evaluation procedures can aid a better understanding of the factors that influence the performances of opaque models, and these procedures can help in the design methods for model modifications.

Prediction of the Power of the Solar Panel

The article presents the results of predicting the power at the output of the solar panel by polynomials of different degrees. The article indicates the need for solar power forecast. The article describes what factors affect the forecast of solar power at the output of the solar panel. Forecasting the amount of electricity generated by a solar power plant is primarily a prediction of the amount of solar radiation received by the solar panel, which in turn depends on environmental conditions and parameters. Data were taken from 04.05.2019 - 05.05.2019 with a discreteness of 1 minute. In order to calculate the forecast, the values of solar insolation were converted to power. The hourly curve of change of solar power with a discreteness in 1 minute is presented. A two-hour curve of the change in solar power with a resolution of 10 minutes is presented. The daily curve of change of solar power with a discreteness in 1 hour is presented. The horizon at 1 hour and 1 day was chosen for forecasting. Approximation of data by means of polynomials of various degrees is checked. The article shows graphs of changes in real and predicted values of solar power at the output of the solar panel. The graphs clearly show which method of forecasting is more accurate. The accuracy of the predicted values was assessed using the average relative error. Of all the considered methods of calculating the predicted value of the power of the solar panel, the smallest error is obtained when the data are selected for 2 hours, differ by no more than 2 times and have a discreteness of 10 minutes. The benefit of using the correction of the predicted data by the Hoyne method is checked. To predict the power of the solar panel by approximation, it is advisable to adjust the predicted data. To correct the data, it is advisable to use the method of predictor-corrector. Predictor - is the predicted value, and the corrector - is the adjusted value After calculating the power forecast at the output of solar power, an algorithm was developed with which you can calculate the predicted value of power. The developed algorithm for calculating the forecast uses the following parameters: data discreteness, the period for which the data are taken for analysis, the degree of the polynomial. First, the algorithm selects data for the selected period, selects discreteness. If you want to increase the discreteness, it averages the value. But on the basis of the selected values calculates the polynomial of the selected degree. Then, based on the calculated equation, the forecast is calculated and the predicted values are displayed in the form of a graph.

Artificial intelligence in renewable systems for transformation towards intelligent buildings

Carbon-neutrality transition in building sectors requires combinations of renewable systems and artificial intelligence (AI) for robustness, reliability, automation, and flexibility. In this study, a comprehensive review on AI applications in renewable systems, have been conducted, to report the current progress, tendency and challenges. Underlying learning mechanisms of AI-based various applications have been studied, in terms of modelling techniques on solar power forecasting, multi-level stochastic uncertainty analysis, smart controls, fault detection and diagnosis, single and multi-objective optimisations and intelligent buildings. Furthermore, AI techniques in renewable energy utilisations have been clearly reviewed, with respect to solar potential evaluation, multi-level stochastic uncertainty analysis, smart controls, fault detection and diagnosis, single and multi-objective optimisations. Results showed that, compared to physical models, the data-driven models show superiority in modelling simplification and modification, high prediction accuracy, and computational efficiency. In order to improve the system operational robustness under multi-level scenario uncertainty, a generic approach with artificial intelligence was proposed, including nonlinear data-driven model development, uncertainty-based energy performance prediction, stochastic uncertainty and statistical analysis. Furthermore, artificial intelligence can also be applied in renewable systems for security, reliability and stability. This study provides a clear roadmap on historical development, recent advances, cutting-edge techniques and progress, and future challenges of AI applications in renewable energy systems, paving path for developing smart and resilient renewable energy systems for decarbonization in buildings.

A Review on forecasting the photovoltaic power Using Machine Learning

In this review paper on different forecasting method of the solar power output for effective generation of the power grid and proper management of transfer rate of energy per unit area occurred into the solar PV system. Essential part in focusing the prediction of solar power is irradiance and temperature. The irradiance can be forecasted by many algorithm and method is applied in prediction of generation of Short-term photovoltaic power and long term solar power forecasting. And many papers describes on numerical weather forecasting and some algorithm like neural networks or support vector regression for two step approach for predicting the PV power. In this review shown that methods like Bagging Model, deep learning, genetic algorithm, random forest, gradient boosting and artificial neural network. We found that for enhancing the performance of predicting PV power many authors proposed the ensemble method that is the hybrid models of different algorithm. And I found that on this review process ensemble methods show that good results and improve the forecasting solar PV power.

Artificial neural networks for power output forecasting from bifacial solar PV system with enhanced building roof surface Albedo

The main objective of this research study is to evaluate the performance of bifacial solar PV systems (bPV) installed on flat roof buildings with controlled surface albedo and to develop forecasting models to anticipate the power output from bifacial solar PV systems due to the enhancement of surface albedo. The originality of this study and the novelty include integrating innovative bifacial solar PV and cool roof technologies to increase solar PV panel power output and anticipate electricity generation in advance to balance supply and demand. For the methodology, modeling and simulation analysis was used to test the performance of bPV system installed on a building with a surface albedo of 0.2–0.8. An Artificial Neural Network was used as a machine learning approach for bifacial solar PV power and energy forecasting. The significant findings show that an increase in surface albedo from 0.2 to 0.5 and 0.2 to 0.8 will help to boost the annual bifacial solar PV power production by 7.75% and 14.96%, respectively. The predictive models are exceptionally accurate in anticipating the bifacial solar PV power. The monthly power from bPV rose from 10.87 to 21.54%, with a 0.3 (from 0.2 to 0.5) to 0.6 (from 0.2 to 0.8) albedo increase. The coefficient of correlation R values for all of the power data ranges from 0.99145 to 0.99382 for the roof surface albedo, ranging from 0.2 to 0.8. The ANN models illustrate how effectively the model mimics the modeling and simulation data. For the research implications, the developed forecast models will assist the industry in predicting ahead of the power production, building operations and maintenance, demand-side management, and advanced energy purchases.

A novel long term solar photovoltaic power forecasting approach using LSTM with Nadam optimizer: A case study of India

AbstractSolar photovoltaic (PV) power is emerging as one of the most viable renewable energy sources. The recent enhancements in the integration of renewable energy sources into the power grid create a dire need for reliable solar power forecasting techniques. In this paper, a new long‐term solar PV power forecasting approach using long short‐term memory (LSTM) model with Nadam optimizer is presented. The LSTM model performs better with the time‐series data as it persists information of more time steps. The experimental models are realized on a 250.25 kW installed capacity solar PV power system located at MANIT Bhopal, Madhya Pradesh, India. The proposed model is compared with two time‐series models and eight neural network models using LSTM with different optimizers. The obtained results using LSTM with Nadam optimizer present a significant improvement in the forecasting accuracy of 30.56% over autoregressive integrated moving average, 47.48% over seasonal autoregressive integrated moving average, and 1.35%, 1.43%, 3.51%, 4.88%, 11.84%, 50.69%, and 58.29% over models using RMSprop, Adam, Adamax, SGD, Adagrad, Adadelta, and Ftrl optimizer, respectively. The experimental results prove that the proposed methodology is more conclusive for solar PV power forecasting and can be employed for enhanced system planning and management.

Forecasting Solar Energy Production Using Machine Learning

When it comes to large-scale renewable energy plants, the future of solar power forecasting is vital to their success. For reliable predictions of solar electricity generation, one must take into consideration changes in weather patterns over time. In this paper, a hybrid model that integrates machine learning and statistical approaches is suggested for predicting future solar energy generation. In order to improve the accuracy of the suggested model, an ensemble of machine learning models was used in this study. The results of the simulation show that the proposed method has reduced placement cost, when compared with existing methods. When comparing the performance of an ensemble model that integrates all of the combination strategies to standard individual models, the suggested ensemble model outperformed the conventional individual models. According to the findings, a hybrid model that made use of both machine learning and statistics outperformed a model that made sole use of machine learning in its performance.

Statistical learning for NWP post-processing: A benchmark for solar irradiance forecasting

The share of solar power in the global and local energy mixes has increased dramatically in the past decade. Consequently, there has been a significant rise in the interest for solar power forecasting, for different time horizons, ranging from few minutes to seasons. For day-ahead forecasts, combination of Numerical Weather Prediction (NWP) models and post-processing algorithms is the most popular approach. Many recent publications have proposed innovative NWP post-processing methods. However, because different works use different datasets, metrics, and even cross-validation methods, it is rarely possible to fairly compare results across several papers.In this work, we propose a rigorous benchmark of several solar NWP post-processing models representative of the literature. For our results to be as general as possible, the comparison is conducted with an open dataset, over 6 years and 7 locations. In addition, we propose a novel benchmarking approach, that focuses on the systematicity of the ranking of forecasting models.Our results show that, when used in combination with proper regularization, large predictor sets are systematically beneficial to NWP post-processing methods. They also demonstrate that more complex algorithms such as neural networks and gradient boosting generally have the lowest mean square error, while support vector regression, a more parsimonious algorithm, performs better in terms of mean absolute error. Lastly, the focus given to ranking systematicity reveals that no model is better in all occasions. This means that researchers should be measured when they conclude to the superiority of a model, in particular when testing data is scarce.

A New Solar Power Prediction Method Based on Feature Clustering and Hybrid-Classification-Regression Forecasting

Solar generation systems are globally extending in terms of scale and number, which highlights the increasing importance of solar power forecast. In this paper, a day-ahead solar power prediction method is proposed including 1) a novel feature selecting/clustering approach based on relevancy and redundancy criteria and 2) an innovative hybrid-classification-regression forecasting engine. The proposed feature selecting/clustering approach filters out irrelevant features and partitions relevant features to two separate subsets to decrease the redundancy of features. Each of these two subsets is separately trained by one forecasting engine and the final solar power prediction of the proposed method is obtained by a relevancy-based combination of these two forecasts. The proposed forecasting engine classifies the historical data based on the learnability of its constituent regression models and assigns each class of training samples to one regression model. Each regression model predicts the outputs of the test samples that belong to its class. The effectiveness of the proposed solar power prediction method is illustrated by testing on two real-world solar farms.