Renewable Energy Forecasting Research Articles

Comparison of machine learning and statistical methods in the field of renewable energy power generation forecasting: a mini review

In the post-COVID-19 era, countries are paying more attention to the energy transition as well as tackling the increasingly severe climate crisis. Renewable energy has attracted much attention because of its low economic costs and environmental friendliness. However, renewable energy cannot be widely adopted due to its high intermittency and volatility, which threaten the security and stability of power grids and hinder the operation and scheduling of power systems. Therefore, research on renewable power forecasting is important for integrating renewable energy and the power grid and improving operational efficiency. In this mini-review, we compare two kinds of common renewable power forecasting methods: machine learning methods and statistical methods. Then, the advantages and disadvantages of the two methods are discussed from different perspectives. Finally, the current challenges and feasible research directions for renewable energy forecasting are listed.

Non-iterative multi-area coordinated energy and reserve allocation model for hybrid AC/DC power systems with flexible frequency operation

As the result of renewable energy natural intermittency, load shedding and renewables curtailment are inevitable to keep supply and demand balance in high penetration level power systems. Luckily, with accurate estimation and proper allocation of the reserve, power fluctuation can be smoothed appreciably. Therefore, this study presents a novel coordinated day-ahead economic dispatch model for both energy and reserve in bulk hybrid AC/DC power systems. Flexible frequency operation for each regional sub AC system is applied to estimate reserve demand, while renewable energy forecast error is described by the versatile probability distribution. Considering practical operation structure in China, a non-iterative bi-layer dispatch algorithm is proposed to share power and reserve between interconnected subsystems by applying DC tie-lines’ flexibility. This bi-layer structure is achieved through the upper layer deciding the boundary variables with aggregated generation information provided by the lower layer. Then each lower layer regional AC grid solves its own programming with the boundary variables. Simulation is carried out to verify the efficacy of the proposed dispatch model and distributed solution methodology.

Assessment of spatiotemporally coordinated cyberattacks on renewable energy forecasting in smart energy system

The employment of deep learning (DL) in renewable energy forecasting (REF) may bring novel cyberthreats. However, this fact has not drawn sufficient attention in the existing literature. To fill the gap, this article, among the first, investigates the latent cyberthreat incurred by the DL-based REF model. First, a spatiotemporally coordinated cyberattack strategy is proposed by considering attackable geo-distributed renewable energy resources (RESs) and periods. It is fulfilled by compromising meteorological data transmitted from external online weather forecast interfaces which are required for REF but also exposed to potential adversaries. In this manner, the malicious adversary is able to deteriorate the REF performance, thereby jeopardizing the operation of the smart energy system. Second, the cyberattack optimization models under the white-box and black-box scenarios are designed. In the white-box scenario, the adversary has full knowledge of these operator-own REF models to obtain false data that should be injected into the meteorological data for launching cyberattacks; in the black-box scenario, the adversary will train surrogate models to replace these operator-own models for attack guidance. Third, an iterative solving method is proposed to solve the proposed white-box and black-box cyberattack optimization models. As DL models (sealed, non-differentiable, and highly non-linear) are introduced to the optimization problem, conventional optimization solvers are unable to handle the cyberattack optimization problem. The proposed solving method utilizes the proximal gradient descent principle, and is feasible to explore the near-optimal solution to the problem. At last, the efficacy of the proposed cyberattack strategy under different RES penetration levels and attack strengths and its catastrophic impacts are comprehensively assessed on the IEEE 39-bus benchmark. The results reveal that the cyberattack would cause enormous economic losses and even collapse the smart energy system.

Data-driven prediction models of photovoltaic energy for smart grid applications

Due to the low total cost of production, photovoltaic energy is a key component of installed renewable energy worldwide. However, photovoltaic energy is volatile in nature as it depends on weather conditions. This makes the integration, control, and operation of this type of energy difficult for network operators. Therefore, renewable energy forecasting is emerging as a key solution to effectively manage renewable energy in the power system. In this paper, we present a time series based energy forecasting system using a solar home database. We worked on Australian data recorded from 2010 to 2013. Three large databases have been combined to obtain better results. The objective is to test and compare the three prediction algorithms, namely: LSTM, FFNN and GRU. We also compared the performances via RMSE, MAE and R2_score. The results showed that neural networks accurately and reliably predict the energy consumed with R2_score up to 0.92 and MSE not exceeding 0.23, with a slight difference between the three models tested in this work.

Editorial: Recent advances in renewable energy automation and energy forecasting

Editorial: Recent advances in renewable energy automation and energy forecasting

Forecasting Renewable Energy Generation with Machine Learning and Deep Learning: Current Advances and Future Prospects

This article presents a review of current advances and prospects in the field of forecasting renewable energy generation using machine learning (ML) and deep learning (DL) techniques. With the increasing penetration of renewable energy sources (RES) into the electricity grid, accurate forecasting of their generation becomes crucial for efficient grid operation and energy management. Traditional forecasting methods have limitations, and thus ML and DL algorithms have gained popularity due to their ability to learn complex relationships from data and provide accurate predictions. This paper reviews the different approaches and models that have been used for renewable energy forecasting and discusses their strengths and limitations. It also highlights the challenges and future research directions in the field, such as dealing with uncertainty and variability in renewable energy generation, data availability, and model interpretability. Finally, this paper emphasizes the importance of developing robust and accurate renewable energy forecasting models to enable the integration of RES into the electricity grid and facilitate the transition towards a sustainable energy future.

Regime‐Specific Cloud Vertical Overlap Characteristics From Radar and Lidar Observations at the ARM Sites

AbstractClimate and numerical weather prediction models require assumptions to represent the vertical distribution of subgrid‐scale clouds, which have radiative transfer implications. In this study, nearly 25 years of ground‐based radar and lidar observations of vertical cloud profiles at the Atmospheric Radiation Measurement Program (ARM) Southern Great Plains (SGP) site are utilized to derive cloud vertical overlap characteristics from the Cloud Type (CLDTYPE) data product. The cloud vertical overlap characteristics are further separated by cloud regime by considering seven cloud types (i.e., low cloud, congestus, deep convection, altocumulus, altostratus, cirrostratus, and cirrus) as well as periods of shallow cumulus. The decorrelation length scale (i.e., exponential transition from maximum to random overlap with layer separation) is found to vary by cloud regime, ranging between 0.04 km for cirrostratus paired with cirrus to 4.58 km for low cloud paired with cirrus at SGP. Cloud vertical overlap characteristics are also considered for other ARM sites including the Tropical Western Pacific (TWP), North Slope of Alaska (NSA), and Eastern North Atlantic (ENA) sites among other shorter term ARM deployments globally. The decorrelation length scale ranged globally from 1.03 km in the Arctic Ocean to 3.06 km in Manacapuru, Brazil. Globally, the decorrelation length scale by cloud regime exhibited similarities (e.g., for cirrus paired with cirrus) and differences (e.g., congestus paired with cirrus). The results could help inform development of cloud vertical overlap assumptions within operational numerical weather prediction models and potentially improve prediction of radiative fluxes for weather, climate, and renewable energy forecasting.

Multi-Timescale Optimal Dispatching Strategy for Coordinated Source-Grid-Load-Storage Interaction in Active Distribution Networks Based on Second-Order Cone Planning

In order to cope with the efficient consumption and flexible regulation of resource scarcity due to grid integration of renewable energy sources, a scheduling strategy that takes into account the coordinated interaction of source, grid, load, and storage is proposed. In order to improve the accuracy of the dispatch, a BP neural network approach modified by a genetic algorithm is used to predict renewable energy sources and loads. The non-convex, non-linear optimal dispatch model of the distribution grid is transformed into a mixed integer programming model with optimal tides based on the second-order cone relaxation, variable substitution, and segmental linearization of the Big M method. In addition, the uncertainty of distributed renewable energy output and the flexibility of load demand re-response limit optimal dispatch on a single time scale, so the frequency of renewable energy and load forecasting is increased, and an optimal dispatch model with complementary time scales is developed. Finally, the IEEE 33-node distribution system was tested to verify the effectiveness of the proposed optimal dispatching strategy. The simulation results show an 18.28% improvement in the economy of the system and a 24.39% increase in the capacity to consume renewable energy.

An integrated method with adaptive decomposition and machine learning for renewable energy power generation forecasting.

In recent years, traditional energy sources have caused a variety of negative impacts on the environment, and reducing carbon emissions is a top priority. The development of renewable energy technology is the key to transform the energy structure. Renewable energy represented by wind energy and photovoltaics has abundant reserves so they are connected to the grid system on a large scale. However, because of natural energy's randomness, renewable energy power generation poses potential risks to energy production and grid security. By making short-term forecasts of renewable energy generation power, the uncertainty of energy generation can be reduced, and it is crucial to study renewable energy forecasting techniques. This paper proposes an integrated forecasting system for renewable energy sources. Firstly, ensemble empirical mode decomposition is used for data preprocessing, and stationarity analysis is used for modal identification; then, support vector regression optimized by sparrow search algorithm and statistical methods are combined to make forecast according to different characteristics of the series respectively; finally, the feasibility of this method in renewable energy time series prediction is verified by experiments. The experiments prove that the proposed model effectively improves the accuracy and prediction performance on ultra-short-term renewable energy forecasting; and it has good applicability and competitiveness with different forecasting scenarios and characteristics, which satisfy the actual forecasting requirements in terms of operational efficiency and accuracy, thus providing a technical basis for the effective utilization of renewable energy.

Renewable Energy Forecasting with Hybrid Nonlinear Model (ANFIS): Case Study of Wind Speed in Thailand

Renewable Energy Forecasting with Hybrid Nonlinear Model (ANFIS): Case Study of Wind Speed in Thailand

AN IMPROVED ACCURACY FOR THE FORECASTING OF POWER GENERATION OVER A LONG-TERM HORIZON

Renewable energy becomes increasingly popular in the global electric energy grid, improving the accuracy of renewable energy forecasting is critical to power system planning, management, and operations. However, this is a challenging task due to the intermittent and chaotic nature of renewable energy data. To date, various methods have been developed, including physical models, statistical methods, artificial intelligence techniques, and their hybrids to improve the forecasting accuracy of renewable energy. Hence this research proposed to hybridize two strong deep learning algorithms where modeling of more complex functioning is allowed by the use of multiple layers of abstraction in order to come up with a powerful forecasting model to predict solar power generation over long term horizon. Finally, the Deep Neutral Network and Long-short Term memory Network (DNN-LSTM) method can generate predicted solar energy consumption in a fully connected hierarchy. The proposed DNN-LSTM model achieved Mean Square Error (MSE) of 0.00825 and MAE of 0.00100 respectively. This is by far the lowest value when compare against the existing model i.e MLSHM which has MSE of 0.05700 and MAPE of 0.00695, LSTM which has MSE of 0.0536 and MAE of 0.0037 and Gated Recurrent Unit (GRU) which has MSE of 0.03460 and MAE of 0.00243 respectively. Thus, the proposed DNN-LSTM have clearly enhanced the forecasting accuracy as against all the existing models that was used for the evaluation and achieved the lowest values in terms of validation of MSE and MAE.

Transformer-Based Hybrid Forecasting Model for Multivariate Renewable Energy

In recent years, the use of renewable energy has grown significantly in electricity generation. However, the output of such facilities can be uncertain, affecting their reliability. The forecast of renewable energy production is necessary to guarantee the system’s stability. Several authors have already developed deep learning techniques and hybrid systems to make predictions as accurate as possible. However, the accurate forecasting of renewable energy still is a challenging task. This work proposes a new hybrid system for renewable energy forecasting that combines the traditional linear model (Seasonal Autoregressive Integrated Moving Average—SARIMA) with a state-of-the-art Machine Learning (ML) model, Transformer neural network, using exogenous data. The proposal, named H-Transformer, is compared with other hybrid systems and single ML models, such as Long Short Term Memory (LSTM), Gated Recurrent Unit (GRU), and Recurrent Neural Networks (RNN), using five data sets of wind speed and solar energy. The proposed H-Transformer attained the best result compared to all single models in all datasets and evaluation metrics. Finally, the hybrid H-Transformer obtained the best result in most cases when compared to other hybrid approaches, showing that the proposal can be a useful tool in renewable energy forecasting.

Regional Photovoltaic Power Forecasting Using Vector Autoregression Model in South Korea

Renewable energy forecasting is a key for efficient resource use in terms of power generation and safe grid control. In this study, we investigated a short-term statistical forecasting model with 1 to 3 h horizons using photovoltaic operation data from 215 power plants throughout South Korea. A vector autoregression (VAR) model-based regional photovoltaic power forecasting system is proposed for seven clusters of power plants in South Korea. This method showed better predictability than the autoregressive integrated moving average (ARIMA) model. The normalized root-mean-square errors of hourly photovoltaic generation predictions obtained from VAR (ARIMA) were 8.5–10.9% (9.8–13.0%) and 18.5–22.8% (21.3–26.3%) for 1 h and 3 h horizon, respectively, at 215 power plants. The coefficient of determination, R2 was higher for VAR, at 4–5%, than ARIMA. The VAR model had greater accuracy than ARIMA. This will be useful for economical and efficient grid management.

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

A combination of supervised dimensionality reduction and learning methods to forecast solar radiation

Machine learning is routinely used to forecast solar radiation from inputs, which are forecasts of meteorological variables provided by numerical weather prediction (NWP) models, on a spatially distributed grid. However, the number of features resulting from these grids is usually large, especially if several vertical levels are included. Principal Components Analysis (PCA) is one of the simplest and most widely-used methods to extract features and reduce dimensionality in renewable energy forecasting, although this method has some limitations. First, it performs a global linear analysis, and second it is an unsupervised method. Locality Preserving Projection (LPP) overcomes the locality problem, and recently the Linear Optimal Low-Rank (LOL) method has extended Linear Discriminant Analysis (LDA) to be applicable when the number of features is larger than the number of samples. Supervised Nonnegative Matrix Factorization (SNMF) also achieves this goal extending the Nonnegative Matrix Factorization (NMF) framework to integrate the logistic regression loss function. In this article we try to overcome all these issues together by proposing a Supervised Local Maximum Variance Preserving (SLMVP) method, a supervised non-linear method for feature extraction and dimensionality reduction. PCA, LPP, LOL, SNMF and SLMVP have been compared on Global Horizontal Irradiance (GHI) and Direct Normal Irradiance (DNI) radiation data at two different Iberian locations: Seville and Lisbon. Results show that for both kinds of radiation (GHI and DNI) and the two locations, SLMVP produces smaller MAE errors than PCA, LPP, LOL, and SNMF, around 4.92% better for Seville and 3.12% for Lisbon. It has also been shown that, although SLMVP, PCA, and LPP benefit from using a non-linear regression method (Gradient Boosting in this work), this benefit is larger for PCA and LPP because SMLVP is able to perform non-linear transformations of inputs.

Optimization-driven uncertainty forecasting: Application to day-ahead commitment with renewable energy resources

The participation of power producers who generate electricity from renewable energy resources, e.g., wind and solar, in the electricity market has been significantly promoted by the high penetration of renewable electricity generation. In this case, the unit commitment is required to be formulated and solved to determine the optimal commitment in response to uncertainty existing in renewable electricity generation. To address this problem, a forecasting-first optimization-second approach has been practically applied in that the next-day renewable electricity generation is forecasted first using historical data, and then the unit commitment problem is solved using the forecasted renewable energy generation. However, the forecasting-first optimization-second approach has limitation that forecasting and optimization are decoupled, and thus, forecasting cannot be tuned by reflecting the effect of forecasted renewable energy generation on optimizing unit commitment. Given this context, this study proposes an optimization-driven renewable energy generation forecasting that is designed to integrate the unit commitment problem with the course of regression so that the regression can be done to minimize forecasting error while maximizing profit. Numerical experiments are conducted based on general regression models, e.g., auto-regressive and multiple linear regression models, with various parameter settings, and the results demonstrate that the proposed approach provides better renewable energy forecasting in terms of resulting unit commitment, i.e., greater profit and less penalty, without degrading forecasting accuracy significantly.

신재생발전 지역 관제를 위한 신재생 발전출력 예측 기술

This paper introduces the Local Renewable Management System(LRMS) developed by Korea Electric Power Corporation and describes the Renewable Energy Forecasting System, which is one of the main module in LRMS. The Renewable Energy Forecasting System forecasts the generation output by using meteorological data from the Korea Meteorological Administration, electrical data acquired from SCADA and DAS which are KEPCO

Short Term Renewable Energy Forecasting Based on Feed Forward Back Propagation Neural Network Strategy

The fundamental inputs used as a renewable energy source are wind speed and solar radiation. Both parameters are very nonlinear and depending on their surroundings. As a result, reliable prediction of these characteristics is required for usage in a variety of agricultural, industrial, transportation, and environmental applications since they reduce greenhouse gas emissions and are environmentally benign. In this study, we used a Feed Forward Back Propagation Neural Network (FFBPN) technique to predict proper data such as temperature, relative moisture, sun radiations, rain, and wind speed. The FFBPN will be trained in such a way that it can conduct hybrid forecasting with little changes to the programming codes, ranging from hourly (short term forecasting) to daily forecasting (medium term forecasting). This feature is one of the significant improvements, showing the suggested hybrid renewable energy forecasting system's high robustness. Because the hybrid forecasting system is a unique approach, the system's accuracy will be determined by comparing the findings to the corresponding values of the persistent model, a stand-alone forecasting model. Finally, the completely created system package could be sold and/or used in future research initiatives to help researcher’s analyses, validate, and illustrate their models across a variety of areas.

A novel fractional structural adaptive grey Chebyshev polynomial Bernoulli model and its application in forecasting renewable energy production of China

Accurate mid-to-long term China's renewable energy forecasting is becoming more and more important for integrating renewable energy systems with smart grid and energy strategic planning. For this purpose, this paper proposes a novel fractional structural adaptive grey Chebyshev polynomial Bernoulli model. The new model is based on NGBM(1,1) model to describe nonlinear phenomena. The proposed adjacent accumulation operator can balance old and new information accumulation with less fluctuation. The introduction of the Chebyshev polynomial enables the model to adaptively adjust the structure of the model, which reduces the difficulty of the modeler's operation from the point of view of the expert system. Meanwhile, an improved Grey Wolf optimization algorithm (IGWO) was used to optimize the parameters. Four real cases of renewable energy in China, including production and consumption are used for verification. In addition, Monte Carlo simulation and probability density analysis illustrate the robustness and accuracy of the proposed model. Finally, the development of the four cases in the next 5 years is predicted.

Data-Driven Optimization of Transmission Section Limits and Their Adaptability for New Power Systems With High Penetration of Renewable Energy

In traditional power systems, the transmission section limits are often set to predetermined values according to the typical peak-to-valley demand modes, which may be difficult to adapt to the complex and diverse operation modes of the new power systems with high penetration of renewable energy, and it cannot consider the coupling characteristics of AC and DC transmission sections also. In order to make better use of the efficiency of the transmission sections, a mathematical model for optimizing the adaptability of the transmission section limit is firstly proposed, in which a data-driven typical demand mode extraction method based on K-means clustering is proposed according to the load forecast, renewable energy forecast and market transaction data. Then, the adaptability assessment of the section limit and the transmission requirements are carried out. Finally, the power perturbation method is used to calculate the influence factor matrix of associated transmission section limit and stability margin of its dominant fault. And the influence factors are optimized according to section limit adaptability and stability margin influence, and the best adaptable limit of the associated transmission sections are obtained. The calculation and optimization results of the actual case of the interconnection corridor between Xinjiang and the northwest main power system of China verify the effectiveness and practicability of the proposed limit optimization model, which is able to improve the flexibility and adaptability of the actual power system operation.