Optimizing multi-step wind power forecasting: Integrating advanced deep neural networks with stacking-based probabilistic learning

Applied Energy, Год журнала: 2024, Номер 369, С. 123487 - 123487

Опубликована: Май 30, 2024

Язык: Английский

---

Probabilistic wind speed forecasting via Bayesian DLMs and its application in green hydrogen production

Applied Energy, Год журнала: 2025, Номер 382, С. 125286 - 125286

Опубликована: Янв. 13, 2025

Язык: Английский

---

A Solar and Wind Energy Evaluation Methodology Using Artificial Intelligence Technologies

Energies, Год журнала: 2024, Номер 17(2), С. 416 - 416

Опубликована: Янв. 15, 2024

The use of renewable energy sources is becoming increasingly widespread around the world due to various factors, most relevant which high environmental friendliness these types resources. However, large-scale involvement green leads creation distributed networks that combine several different generation methods, each has its own specific features, and as a result, data collection processing necessary optimize operation such systems become more relevant. Development new technologies for optimal RES one main tasks modern research in field energy, where an important place assigned based on artificial intelligence, allowing researchers significantly increase efficiency all within systems. This paper proposes consider methodology application approaches assessment amount obtained from intelligence technologies, used optimization control processes operating with integration sources. relevance work lies formation general approach applied evaluation solar wind technologies. As verification considered by authors, number models predicting power using photovoltaic panels have been implemented, machine-learning methods used. result testing quality accuracy, best results were hybrid forecasting model, combines joint random forest model at stage normalization input data, exponential smoothing LSTM model.

Язык: Английский

---

A novel wind power deterministic and interval prediction framework based on the critic weight method, improved northern goshawk optimization, and kernel density estimation

Renewable Energy, Год журнала: 2024, Номер 226, С. 120360 - 120360

Опубликована: Март 18, 2024

Язык: Английский

---

Designing and prototyping the architecture of a digital twin for wind turbine

International Journal of Thermofluids, Год журнала: 2024, Номер 22, С. 100622 - 100622

Опубликована: Март 5, 2024

This paper outlines the key components necessary to develop a digital twin (DT) for wind turbine, aiming provide detailed methodology and guidelines building this system, which facilitates optimization during operation helps prevent system failures. It presents four major systems required construct DT: physical, digital, connection, service systems. study also critical design, measured, calculated parameters of are essential development DT. The physical turbine is examined, components, including rotor, blades, shaft, generator, tower, nacelle, discussed in detail. explores DT, data storage, models, mathematical modelling. problems that may occur were presented addition possible solutions must suggest. According project's needs requirements, it was found DT can employ various connection such as supervisory control acquisition, wireless sensor networks, smart grids, Internet Things, cloud-based

Язык: Английский

---

A machine learning approach for wind turbine power forecasting for maintenance planning

Energy Informatics, Год журнала: 2025, Номер 8(1)

Опубликована: Янв. 6, 2025

Abstract Integrating power forecasting with wind turbine maintenance planning enables an innovative, data-driven approach that maximizes energy output by predicting periods low production and aligning them schedules, improving operational efficiency. Recently, many countries have met renewable targets, primarily using solar, to promote sustainable growth reduce emissions. Forecasting is crucial for maintaining a stable reliable grid. As integration increases, precise electricity demand becomes essential at every system level. This study presents compares nine machine learning (ML) methods forecasting, Interpretable ML, Explainable Blackbox model. The interpretable ML includes Linear Regression (LR), K-Nearest Neighbors (KNN), eXtreme Gradient Boosting (XGBoost), Random Forest (RF); the explainable consists of graphical Neural network (GNN); blackbox model Multi-layer Perceptron (MLP), Recurrent Network (RNN), Gated Unit (GRU), Long Short-Term Memory (LSTM). These are applied EDP datasets three causal variable types: including temporal information, metrological curtailment information. Computational results show GNN-based outperforms other benchmark regarding accuracy. However, when considering computational resources such as memory processing time, XGBoost provides optimal results, offering faster reduced usage. Furthermore, we present various time windows horizons, ranging from 10 minutes day.

Язык: Английский

---

Clustering and dynamic recognition based auto-reservoir neural network: A wait-and-see approach for short-term park power load forecasting

iScience, Год журнала: 2023, Номер 26(8), С. 107456 - 107456

Опубликована: Июль 22, 2023

This paper proposes a novel clustering and dynamic recognition–based auto-reservoir neural network (CDbARNN) for short-term load forecasting (STLF) of industrial park microgrids. In CDbARNN, the available sets are first decomposed into several clusters via K-means clustering. Then, by extracting characteristic information series input to CDbARNN curves belonging each cluster center, recognition technology is developed identify which belongs to. After that, it constitute high-dimensional matrix entered reservoir CDbARNN. Finally, node numbers used match different optimized. Numerical experiments conducted on STLF an actual microgrid indicate dominating performance proposed approach through cases comparisons with other well-known deep learning methods.

Язык: Английский

---

Review of Estimating and Predicting Models of the Wind Energy Amount

Energies, Год журнала: 2023, Номер 16(16), С. 5926 - 5926

Опубликована: Авг. 10, 2023

Obtaining wind energy for the production of electric plays a key role in overcoming problems associated with climate change and dwindling reserves traditional types resources. The purpose this work is to analyze current methods estimation forecasting, consider main classifications forecasts used their construction review mathematical distributions calculate speed power flow, depending on specific geographical conditions. In recent years, there has been an increase capacity modern generators, which significantly improved efficiency parks. initial stage determining feasibility involving particular source overall system region preliminary assessment potential, allowing one determine possible percentage substitution energy. To solve such problem, it necessary use models supply. Evaluation as resource creates certain difficulties modeling because stochastic variable. regard, paper proposes various estimating can be classified into empirical based application intelligent data analysis technologies. presents existing amount energy, designed most optimal configuration different conversion technologies relevant case under study, also serves basis creating digital twins model optimize operation projected complex.

Язык: Английский

---

Advancing Renewable Energy Forecasting: A Comprehensive Review of Renewable Energy Forecasting Methods

Energies, Год журнала: 2024, Номер 17(14), С. 3480 - 3480

Опубликована: Июль 15, 2024

Socioeconomic growth and population increase are driving a constant global demand for energy. Renewable energy is emerging as leading solution to minimise the use of fossil fuels. However, renewable resources characterised by significant intermittency unpredictability, which impact their production integration into power grid. Forecasting models increasingly being developed address these challenges have become crucial sources integrated in systems. In this paper, comparative analysis forecasting methods developed, focusing on photovoltaic wind power. A review state-of-the-art techniques conducted synthesise categorise different models, taking account climatic variables, optimisation algorithms, pre-processing techniques, various horizons. By integrating diverse such algorithms carefully selecting forecast horizon, it possible highlight accuracy stability forecasts. Overall, ongoing development refinement achieve sustainable reliable future.

Язык: Английский

---

Non-crossing quantile probabilistic forecasting of cluster wind power considering spatio-temporal correlation

Applied Energy, Год журнала: 2024, Номер 377, С. 124356 - 124356

Опубликована: Сен. 9, 2024

Язык: Английский

---