Renewable Energy Forecasting Research Articles

Prospective Methodologies in Hybrid Renewable Energy Systems for Energy Prediction Using Artificial Neural Networks

This paper presents a comprehensive review of machine learning (ML) based approaches, especially artificial neural networks (ANNs) in time series data prediction problems. According to literature, around 80% of the world’s total energy demand is supplied either through fuel-based sources such as oil, gas, and coal or through nuclear-based sources. Literature also shows that a shortage of fossil fuels is inevitable and the world will face this problem sooner or later. Moreover, the remote and rural areas that suffer from not being able to reach traditional grid power electricity need alternative sources of energy. A “hybrid-renewable-energy system” (HRES) involving different renewable resources can be used to supply sustainable power in these areas. The uncertain nature of renewable energy resources and the intelligent ability of the neural network approach to process complex time series inputs have inspired the use of ANN methods in renewable energy forecasting. Thus, this study aims to study the different data driven models of ANN approaches that can provide accurate predictions of renewable energy, like solar, wind, or hydro-power generation. Various refinement architectures of neural networks, such as “multi-layer perception” (MLP), “recurrent-neural network” (RNN), and “convolutional-neural network” (CNN), as well as “long-short-term memory” (LSTM) models, have been offered in the applications of renewable energy forecasting. These models are able to perform short-term time-series prediction in renewable energy sources and to use prior information that influences its value in future prediction.

Robust capacity optimization methods for integrated energy systems considering demand response and thermal comfort

Integrated Energy System can realize the cascade utilization of energy, which improves the utilization of energy efficiently and reduce the carbon emission. Taking into account the uncertainty of multi-energy load and renewable energy forecasting, this paper presents a bi-level robust optimization model with demand response and thermal comfort, for the capacity planning and operation problem of Integrated Energy System. The outer level optimization is planning to find the optimal integrated energy system configuration to minimize economic investment, while the inner level is to robustly optimize the system scheduling to simultaneously reduce carbon emissions and dissatisfaction of residents’ participation in demand response. In the simulation, the Pareto front of the multi-objective problem is obtained via NSGA-II algorithm and Gurobi solver, and the best design plans on the Pareto front selected by Topsis method are analyzed and discussed. The results from three modes are compared in simulation, which illustrates the economic and environmental benefits from demand response and thermal comfort. In addition, the impact of different thermal time scales and forecast uncertainties on integrated energy system planning are also discussed. Finally the sensitivity of the influence of the low carbon grid constrains on the optimization is analyzed.

A renewable energy forecasting and control approach to secured edge-level efficiency in a distributed micro-grid

Energy forecasting using Renewable energy sources (RESs) is gradually gaining weight in the research field due to the benefits it presents to the modern-day environment. Not only does energy forecasting using renewable energy sources help mitigate the greenhouse effect, it also helps to conserve energy for future use. Over the years, several methods for energy forecasting have been proposed, all of which were more concerned with the accuracy of the prediction models with little or no considerations to the operating environment. This research, however, proposes the uses of Deep Neural Network (DNN) for energy forecasting on mobile devices at the edge of the network. This ensures low latency and communication overhead for all energy forecasting operations since they are carried out at the network periphery. Nevertheless, the cloud would be used as a support for the mobile devices by providing permanent storage for the locally generated data and a platform for offloading resource-intensive computations that exceed the capabilities of the local mobile devices as well as security for them. Electrical network topology was proposed which allows seamless incorporation of multiple RESs into the distributed renewable energy source (D-RES) network. Moreover, a novel grid control algorithm that uses the forecasting model to administer a well-coordinated and effective control for renewable energy sources (RESs) in the electrical network is designed. The electrical network was simulated with two RESs and a DNN model was used to create a forecasting model for the simulated network. The model was trained using a dataset from a solar power generation company in Belgium (elis) and was experimented with a different number of layers to determine the optimum architecture for performing the forecasting operations. The performance of each architecture was evaluated using the mean square error (MSE) and the r-square.

The Impact of Renewable Energy Forecasts on Intraday Electricity Prices

In this paper we study the impact of errors in wind and solar power forecasts on intraday electricity prices. We develop a novel econometric model which is based on day-ahead wholesale auction curves data and errors in wind and solar power forecasts. The model shifts day-ahead supply curves to calculate intraday prices. We apply our model to the German EPEX SPOT SE data. Our model outperforms both linear and non-linear benchmarks. Our study allows us to conclude that errors in renewable energy forecasts exert a non-linear impact on intraday prices. We demonstrate that additional wind and solar power capacities induce non-linear changes in the intraday price volatility. Finally, we comment on economical and policy implications of our findings.

Deep learning for renewable energy forecasting: A review

Deep learning for renewable energy forecasting: A review

Multi-aspect renewable energy forecasting

The increasing presence of renewable energy plants has created new challenges such as grid integration, load balancing and energy trading, making it fundamental to provide effective prediction models. Recent approaches in the literature have shown that exploiting spatio-temporal autocorrelation in data coming from multiple plants can lead to better predictions. Although tensor models and techniques are suitable to deal with spatio-temporal data, they have received little attention in the energy domain. In this paper, we propose a new method based on the Tucker tensor decomposition, capable of extracting a new feature space for the learning task. For evaluation purposes, we have investigated the performance of predictive clustering trees with the new feature space, compared to the original feature space, in three renewable energy datasets. The results are favorable for the proposed method, also when compared with state-of-the-art algorithms.

Feature-Driven Improvement of Renewable Energy Forecasting and Trading

Inspired from recent insights into the common ground of machine learning, optimization and decision-making, this paper proposes an easy-to-implement, but effective procedure to enhance both the quality of renewable energy forecasts and the competitive edge of renewable energy producers in electricity markets with a dual-price settlement of imbalances. The quality and economic gains brought by the proposed procedure essentially stem from the utilization of valuable predictors (also known as features ) in a data-driven newsvendor model that renders a computationally inexpensive linear program. We illustrate the proposed procedure and numerically assess its benefits on a realistic case study that considers the aggregate wind power production in the Danish DK1 bidding zone as the variable to be predicted and traded. Within this context, our procedure leverages, among others, spatial information in the form of wind power forecasts issued by transmission system operators (TSO) in surrounding bidding zones and publicly available in online platforms. We show that our method is able to improve the quality of the wind power forecast issued by the Danish TSO by several percentage points (when measured in terms of the mean absolute or the root mean square error) and to significantly reduce the balancing costs incurred by the wind power producer.

A deep learning method for forecasting residual market curves

Forecasts of residual demand curves (RDCs) are valuable information for price-maker market agents since it enables an assessment of their bidding strategy in the market-clearing price. This paper describes the application of deep learning techniques, namely long short-term memory (LSTM) network that combines past RDCs and exogenous variables (e.g., renewable energy forecasts). The main contribution is to build up on the idea of transforming the temporal sequence of RDCs into a sequence of images, avoiding any feature reduction and exploiting the capability of LSTM in handling image data. The proposed method was tested with data from the Iberian day-ahead electricity market and outperformed machine learning models with an improvement of above 35% in both root mean square error and Frèchet distance.

Smart Distributed Energy Storage Controller (smartDESC)

While the storage properties and the anticipation potential of many classes of power system loads (such as thermal loads) can be exploited to mitigate renewable sources variability, the challenge to do so in an optimal and coherent manner is significant. This is due to the sheer number and dynamic diversity of the loads that can be involved in any large-scale application. The smartDESC concept is a control architecture that was developed for this purpose. It builds on the more pervasive communication means currently available (such as Advanced Metering Infrastructures), as well as the mathematical tools of (i) aggregate load modeling, (ii) renewable energy forecasting, (iii) optimization theory, deterministic or stochastic, and (iv) some recent developments in control of large-scale systems based on game theory, and so-called mean-field (MF) control theory, which allow a scalable yet optimal approach to the decentralized control of large pools of loads. This paper presents the building blocks of the smartDESC architecture, together with an associated simulator and simulation results.

Towards Data Markets in Renewable Energy Forecasting

Geographically distributed wind turbines, photovoltaic panels and sensors (e.g., pyranometers) produce large volumes of data that can be used to improve renewable energy sources (RES) forecasting skill. However, data owners may be unwilling to share their data, even if privacy is ensured, due to a form of prisoner's dilemma: all could benefit from data sharing, but in practice no one is willing to do do. Our proposal hence consists of a data marketplace, to incentivize collaboration between different data owners through the monetization of data. We adapt here an existing auction mechanism to the case of RES forecasting data. It accommodates the temporal nature of the data, i.e., lagged time-series act as covariates and models are updated continuously using a sliding window. A test case with wind energy data is presented to illustrate and assess the effectiveness of such data markets. All agents (or data owners) are shown to benefit in terms of higher revenue resulting from the combination of electricity and data markets. The results support the idea that data markets can be a viable solution to promote data exchange between RES agents and contribute to reducing system imbalance costs.

Combining Artificial Intelligence with Physics-Based Methods for Probabilistic Renewable Energy Forecasting

A modern renewable energy forecasting system blends physical models with artificial intelligence to aid in system operation and grid integration. This paper describes such a system being developed for the Shagaya Renewable Energy Park, which is being developed by the State of Kuwait. The park contains wind turbines, photovoltaic panels, and concentrated solar renewable energy technologies with storage capabilities. The fully operational Kuwait Renewable Energy Prediction System (KREPS) employs artificial intelligence (AI) in multiple portions of the forecasting structure and processes, both for short-range forecasting (i.e., the next six hours) as well as for forecasts several days out. These AI methods work synergistically with the dynamical/physical models employed. This paper briefly describes the methodology used for each of the AI methods, how they are blended, and provides a preliminary assessment of their relative value to the prediction system. Each operational AI component adds value to the system. KREPS is an example of a fully integrated state-of-the-science forecasting system for renewable energy.

Valuing intra-day coordination of electric power and natural gas system operations

Increasing renewable energy penetrations and low natural gas prices have increased the reliance of U.S. power systems on natural gas-fired generation. Increased reliance has sparked concerns over and resulted in natural gas deliverability constraints. To mitigate these constraints, research and regulatory reform have tried to improve coordination between power and gas systems. The U.S. Federal Energy Regulatory Commission (FERC) issued Order 809 in 2015 to improve day-ahead and intra-day coordination of power and gas systems. Given little research on intra-day coordination and FERC's recent action, we quantify the value of improved intra-day coordination between gas and electric power systems. To do so, we co-simulate coordinated day-ahead, intra-day, and real-time operations of an interconnected power and natural gas test system using a dynamic natural gas simulation model and a power system optimization model. We find intra-day coordination reduces total power system production costs and natural gas deliverability constraints, yielding cost and reliability benefits. Sensitivity analysis indicates improved intra-day renewable energy forecasts and higher renewable energy capacities increase intra-day coordination benefits for gas network congestion. Our results indicate FERC Order 809 and other policies aimed at enhancing intra-day coordination between power and gas systems will likely yield cost and reliability benefits.

Research on DC Cross-regional Accommodation Model Based on Scene Analysis

In order to cope with the impact of uncertainty of the renewable energy forecasting in two regions on the cross-regional accommodation of renewable energy, a DC cross-regional accommodation model based on scene analysis is established. Firstly, the ARMA model, scene reduction and scene combination concept are used to construct the scene set describing the uncertainty of the renewable energy forecasting in two regions. Then, the model is used to obtain the unit start-stop plan in two regions and the DC line transmission plan with the goal of smallest renewable energy power abandonment statistically. Finally, based on the measured data of renewable energy power generation, the unit power generation schedule with the goal of smallest renewable energy power abandonment in the actual operation of the power grid is obtained. The validity of the model is verified by a cross-regional system example of actual data reconstruction in a province.

A Deep Neural Network-Assisted Approach to Enhance Short-Term Optimal Operational Scheduling of a Microgrid

The inherent variability of large-scale renewable energy generation leads to significant difficulties in microgrid energy management. Likewise, the effects of human behaviors in response to the changes in electricity tariffs as well as seasons result in changes in electricity consumption. Thus, proper scheduling and planning of power system operations require accurate load demand and renewable energy generation estimation studies, especially for short-term periods (hour-ahead, day-ahead). The time-sequence variation in aggregated electrical load and bulk photovoltaic power output are considered in this study to promote the supply-demand balance in the short-term optimal operational scheduling framework of a reconfigurable microgrid by integrating the forecasting results. A bi-directional long short-term memory units based deep recurrent neural network model, DRNN Bi-LSTM, is designed to provide accurate aggregated electrical load demand and the bulk photovoltaic power generation forecasting results. The real-world data set is utilized to test the proposed forecasting model, and based on the results, the DRNN Bi-LSTM model performs better in comparison with other methods in the surveyed literature. Meanwhile, the optimal operational scheduling framework is studied by simultaneously making a day-ahead optimal reconfiguration plan and optimal dispatching of controllable distributed generation units which are considered as optimal operation solutions. A combined approach of basic and selective particle swarm optimization methods, PSO&SPSO, is utilized for that combinatorial, non-linear, non-deterministic polynomial-time-hard (NP-hard), complex optimization study by aiming minimization of the aggregated real power losses of the microgrid subject to diverse equality and inequality constraints. A reconfigurable microgrid test system that includes photovoltaic power and diesel distributed generators is used for the optimal operational scheduling framework. As a whole, this study contributes to the optimal operational scheduling of reconfigurable microgrid with electrical energy demand and renewable energy forecasting by way of the developed DRNN Bi-LSTM model. The results indicate that optimal operational scheduling of reconfigurable microgrid with deep learning assisted approach could not only reduce real power losses but also improve system in an economic way.

FPGA based substantial power evolution controlling strategy for solar and wind forecasting grid connected system

The solar and wind are both the most promising renewable and clean energy sources, the solar stable energy progress and environmental protection have been increasingly noticeable. In this regard, an accurate solar and wind energy prediction is extremely important to avoid large voltage changes to the power grid and to provide a mechanism for the system to optimally manage the generated energy. Wind energy forecasting is widely practiced among modest power systems for high levels of windmills. This paper aims to develop a new hybrid system for wind and solar energy prediction. The proposed hybrid (wind & solar) energy prediction model is based on a Substantial Power Evolution Strategy (SPES) dedicated to short-term forecasting. The proposed forecasting system SPES is implemented using MATLAB. This paper implements the short-term and hybrid power forecasting using Substantial Power Evolution Strategy based on Prediction Intervals (PIs). This feature is one of the major innovations in the proposed hybrid renewable energy forecasting system. The accuracy of the proposed system will be revealed by comparing the results of the corresponding values of the independent forecasting models called persistence models. The designed device presents a real-time application of predicting daily total solar and wind power using any geographic location and environmental conditions using FPGA. Finally, fully developed system packages can be commercialized and/or utilized for further research projects, and researchers can analyze, validate and visualize their models for related fields.

Machine learning models for renewable energy forecasting

As the world unanimously works towards utilizing non-conventional energy for powering industries, households, vehicles, etc., one of the key limiting factors is the fluctuation of energy availability by non-conventional energy sources. The power generated by wind turbines or photovoltaic cells is dependent upon factors that can’t be controlled such as wind speed, humidity, solar irradiance. In such a situation the integration of renewable energy into grids becomes difficult. Eventually, maintaining an equilibrium between the energy supply and demand can be erratic. This calls for forecasting the amount of energy available via non-conventional energy sources like wind and sun so that the transition to renewable energy can be done highly efficiently without destabilizing the power grid. As the renewable power industry has abundant data that can be exploited in renewable energy forecasting, machine learning techniques can revolutionize the way we deal with renewable energy. This paper describes the efficiency of Linear Regression, Neural Networks Regression, Random Forest Regression, and Extra Tree Regression models for forecasting solar irradiation available on Earth’s surface.

Renewable Energy Generation Assessment in Terms of Small-Signal Stability

The popularity and role of renewable energy in the power grid are increasing nowadays as countries are shifting to cleaner forms of energy. This brings new challenges in maintaining a secure and stable power system, as renewable energy is known to be intermittent in nature and may introduce stability issues to the grid. In this paper, a screening framework of renewable energy generation scenarios is proposed to determine which power system conditions and scenarios will make the system unstable. The scenario screening framework is based on a sensitivity analysis of the system eigenvalues with respect to the renewable energy penetration to the system. The average scheduled renewable energy output, forecasting error standard deviation, average forecasting error, and bus location of the renewable energy source were used to define a renewable energy generation scenario. Depending on the amount and variability of renewable energy, there is a possibility for a critical eigenvalue to cross the imaginary axis. The estimated eigenvalue location resulting from the penetration of variable renewable energy is computed by adding the computed eigenvalue sensitivity to the initial operating point. If any of the estimated system eigenvalues cross the imaginary axis, the power system might be unstable in this scenario, so it requires more detailed simulations and countermeasures. Renewable energy forecasting was done using the long short-term memory model, and the proposed method was simulated using the IEEE 39-bus New England test system. The results of the proposed method were verified by comparing the simulation results to the eigenanalysis solution. The obtained results have shown that the proposed method can determine whether the renewable energy generation scenario is critical in power system operation.

Optimal Allocation of Spinning Reserves in Interconnected Energy Systems with Demand Response Using a Bivariate Wind Prediction Model

The proposed study presents a novel probabilistic method for optimal allocation of spinning reserves taking into consideration load, wind and solar forecast errors, inter-zonal spinning reserve trading, and demand response provided by consumers as a single framework. The model considers the system contingencies due to random generator outages as well as the uncertainties caused by load and renewable energy forecast errors. The study utilizes a novel approach to model wind speed and its direction using the bivariate parametric model. The proposed model is applied to the IEEE two-area reliability test system (RTS) to analyze the influence of inter-zonal power generation and demand response (DR) on the adequacy and economic efficiency of energy systems. In addition, the study analyzed the effect of the bivariate wind prediction model on obtained results. The results demonstrate that the presence of inter-zonal capacity in ancillary service markets reduce the total expected energy not supplied (EENS) by 81.66%, while inclusion of a DR program results in an additional 1.76% reduction of EENS. Finally, the proposed bivariate wind prediction model showed a 0.27% reduction in spinning reserve requirements, compared to the univariate model.

The impact of renewable energy forecast errors on imbalance volumes and electricity spot prices

This paper contributes to the general consideration of whether a policy of incentivising system operators to improve the quality and market availability of forecasts for renewable energy outputs would be beneficial. Using data from the German electricity market, we investigate the effect of wind and solar energy forecasts errors on imbalance volumes and subsequent spot electricity prices. We use ordinary least squares regression, quantile regression and autoregressive moving averages to identify these relationships using variables that have a quarter-hourly data granularity. The results show that higher wind and solar forecast errors increase the absolute values of imbalance volumes and that these can pass through into higher spot prices. We find that wind forecast errors in Germany impact spot prices more than solar forecasting errors. Policy incentives to improve the accuracy and availability of renewable energy forecasts by the system operators should therefore be encouraged.

Deep Learning Neural Networks Trained with MODIS Satellite-Derived Predictors for Long-Term Global Solar Radiation Prediction

Solar energy predictive models designed to emulate the long-term (e.g., monthly) global solar radiation (GSR) trained with satellite-derived predictors can be employed as decision tenets in the exploration, installation and management of solar energy production systems in remote and inaccessible solar-powered sites. In spite of a plethora of models designed for GSR prediction, deep learning, representing a state-of-the-art intelligent tool, remains an attractive approach for renewable energy exploration, monitoring and forecasting. In this paper, algorithms based on deep belief networks and deep neural networks are designed to predict long-term GSR. Deep learning algorithms trained with publicly-accessible Moderate Resolution Imaging Spectroradiometer (MODIS) satellite data are tested in Australia’s solar cities to predict the monthly GSR: single hidden layer and ensemble models. The monthly-scale MODIS-derived predictors (2003–2018) are adopted, with 15 diverse feature selection approaches including a Gaussian Emulation Machine for sensitivity analysis used to select optimal MODIS-predictor variables to simulate GSR against ground-truth values. Several statistical score metrics are adopted to comprehensively verify surface GSR simulations to ascertain the practicality of deep belief and deep neural networks. In the testing phase, deep learning models generate significantly lower absolute percentage bias (≤3%) and high Kling–Gupta efficiency (≥97.5%) values compared to the single hidden layer and ensemble model. This study ascertains that the optimal MODIS input variables employed in GSR prediction for solar energy applications can be relatively different for diverse sites, advocating a need for feature selection prior to the modelling of GSR. The proposed deep learning approach can be adopted to identify solar energy potential proactively in locations where it is impossible to install an environmental monitoring data acquisition instrument. Hence, MODIS and other related satellite-derived predictors can be incorporated for solar energy prediction as a strategy for long-term renewable energy exploration.