Probabilistic net load forecasting framework for application in distributed integrated renewable energy systems DOI Creative Commons
Jan-Simon Telle, Ajay Upadhaya, Patrik Schönfeldt

et al.

Energy Reports, Journal Year: 2024, Volume and Issue: 11, P. 2535 - 2553

Published: Feb. 16, 2024

Integrating various sectors enhances resilience in distributed sector-integrated energy systems. Forecasting is vital for unlocking full potential and enabling well-informed decisions management. Given the inherent variability generation demand prediction, quantification of uncertainty crucial. Therefore, probabilistic forecasting becoming imperative compared to deterministic forecasting, as it ensures a more comprehensive depiction uncertainty. This paper introduces net load framework (PNLFF), non-blackbox approach that robust, non-parametric, computational data inexpensive, adaptable across sectors. It utilizes personalized standard profile forecasts, integrates quantile regression generate forecast. The cumulative distribution function approximated from quantiles forecast using piecewise cubic hermite interpolating polynomial, then derived probability density (PDF). Then was obtained by convolution PDFs electricity demand, heat PV generation. A case study demonstrates its application operational optimization system logistics facility. In first stage PNLFF, results profiles clearly show they can be applied all outperform their respective benchmarks. second stage, expansion regression, also performs promisingly sectors, with best being achieved particular small training set 30 days. With extension interpolation, demonstrated how PDF without prior knowledge data. result demonstrate PNL, an aggregated different convolution, used decision making under uncertainty, e.g. planning flexible loads.

Language: Английский

Forecasting Solar Power Generation Utilizing Machine Learning Models in Lubbock DOI Creative Commons
Afshin Balal, Yaser Pakzad Jafarabadi, Ayda Demir

et al.

Emerging Science Journal, Journal Year: 2023, Volume and Issue: 7(4), P. 1052 - 1062

Published: July 12, 2023

Solar energy is a widely accessible, clean, and sustainable source. power harvesting in order to generate electricity on smart grids essential light of the present global crisis. However, highly variable nature solar radiation poses unique challenges for accurately predicting photovoltaic (PV) generation. Factors such as cloud cover, atmospheric conditions, seasonal variations significantly impact amount available conversion into electricity. Therefore, it precisely estimate output assess potential grids. This paper presents study that utilizes various machine learning models predict generation Lubbock, Texas. Mean Squared Error (MSE) R² metrics are utilized demonstrate performance each model. The results show Random Forest Regression (RFR) Long Short-Term Memory (LSTM) outperformed other models, with MSE 2.06% 2.23% values 0.977 0.975, respectively. In addition, RFR LSTM their capability capture intricate patterns complex relationships inherent data. developed can aid PV investors streamlining processes improving planning production energy. Doi: 10.28991/ESJ-2023-07-04-02 Full Text: PDF

Language: Английский

Citations

50
A review on enhancing energy efficiency and adaptability through system integration for smart buildings DOI

Um-e-Habiba,

Ijaz Ahmed, Mohammad Asif

et al.

Journal of Building Engineering, Journal Year: 2024, Volume and Issue: 89, P. 109354 - 109354

Published: April 18, 2024

Language: Английский

Citations

32
A comprehensive analysis of the emerging modern trends in research on photovoltaic systems and desalination in the era of artificial intelligence and machine learning DOI Creative Commons
Laxmikant D. Jathar, Keval Chandrakant Nikam,

Umesh V. Awasarmol

et al.

Heliyon, Journal Year: 2024, Volume and Issue: 10(3), P. e25407 - e25407

Published: Feb. 1, 2024

Integration of photovoltaic (PV) systems, desalination technologies, and Artificial Intelligence (AI) combined with Machine Learning (ML) has introduced a new era remarkable research innovation. This review article thoroughly examines the recent advancements in field, focusing on interplay between PV systems water within framework AI ML applications, along it analyses current to identify significant patterns, obstacles, prospects this interdisciplinary field. Furthermore, incorporation methods improving performance systems. includes raising their efficiency, implementing predictive maintenance strategies, enabling real-time monitoring. It also explores transformative influence intelligent algorithms techniques, specifically addressing concerns pertaining energy usage, scalability, environmental sustainability. provides thorough analysis literature, identifying areas where is lacking suggesting potential future avenues for investigation. These have resulted increased decreased expenses, improved sustainability system. By utilizing artificial intelligence freshwater productivity can increase by 10 % efficiency. offers informative perspectives researchers, engineers, policymakers involved renewable technology. sheds light latest desalination, which are facilitated ML. The aims guide towards more sustainable technologically advanced future.

Language: Английский

Citations

31
Hybrid deep learning models for time series forecasting of solar power DOI Creative Commons
Diaa Salman, Cem Direkoğlu, Mehmet Kuşaf

et al.

Neural Computing and Applications, Journal Year: 2024, Volume and Issue: 36(16), P. 9095 - 9112

Published: Feb. 22, 2024

Abstract Forecasting solar power production accurately is critical for effectively planning and managing renewable energy systems. This paper introduces investigates novel hybrid deep learning models forecasting using time series data. The research analyzes the efficacy of various capturing complex patterns present in In this study, all possible combinations convolutional neural network (CNN), long short-term memory (LSTM), transformer (TF) are experimented. These also compared with single CNN, LSTM TF respect to different kinds optimizers. Three evaluation metrics employed performance analysis. Results show that CNN–LSTM–TF model outperforms other models, a mean absolute error (MAE) 0.551% when Nadam optimizer. However, TF–LSTM has relatively low performance, an MAE 16.17%, highlighting difficulties making reliable predictions power. result provides valuable insights optimizing systems, significance selecting appropriate optimizers accurate forecasting. first such comprehensive work presented involves networks

Language: Английский

Citations

18
Machine learning based on reliable and sustainable electricity supply from renewable energy sources in the agriculture sector DOI
Ahmed I. Taloba, Alanazi Rayan

Journal of Radiation Research and Applied Sciences, Journal Year: 2025, Volume and Issue: 18(1), P. 101282 - 101282

Published: Jan. 5, 2025

Language: Английский

Citations

3
A Comprehensive Review on Deep Learning Applications in Advancing Biodiesel Feedstock Selection and Production Processes DOI Creative Commons
Olugbenga Akande, Jude A. Okolie, Richard Kimera

et al.

Green Energy and Intelligent Transportation, Journal Year: 2025, Volume and Issue: unknown, P. 100260 - 100260

Published: Jan. 1, 2025

Language: Английский

Citations

2
Robust operating strategy for voltage and frequency control in a non-linear hybrid renewable energy-based power system using communication time delay DOI

Rasmia Irfan,

Muhammad Majid Gulzar, Adnan Shakoor

et al.

Computers & Electrical Engineering, Journal Year: 2025, Volume and Issue: 123, P. 110119 - 110119

Published: Jan. 30, 2025

Language: Английский

Citations

2
ENHANCING ENERGY EFFICIENCY WITH AI: A REVIEW OF MACHINE LEARNING MODELS IN ELECTRICITY DEMAND FORECASTING DOI Creative Commons

Adebayo Olusegun Aderibigbe,

Emmanuel Chigozie Ani,

Peter Efosa Ohenhen

et al.

Engineering Science & Technology Journal, Journal Year: 2023, Volume and Issue: 4(6), P. 341 - 356

Published: Dec. 7, 2023

This study presents a comprehensive review of the impact artificial intelligence (AI) and machine learning (ML) on enhancing energy efficiency, particularly in context electricity demand forecasting. The systematically explores paradigm shift brought about by emergence AI focusing role forecasting historical evolution techniques. A critical analysis various ML models is conducted, examining their theoretical underpinnings, selection criteria, performance diverse scenarios. Key insights reveal that models, especially those incorporating deep big data analytics, significantly outperform traditional methods accuracy adaptability. These are adept at handling complex, nonlinear relationships large datasets, making them effective dynamic increasingly renewable-focused markets. also highlights importance selecting appropriate based criteria such as accuracy, adaptability to periods, capabilities, environmental considerations. further delves into technological, economic, impacts efficiency. It underscores potential drive innov4eations forecasting, contributing more sustainable efficient management. However, challenges privacy, cybersecurity, need for skilled professionals identified areas requiring attention. Strategic recommendations provided practitioners policymakers, emphasizing investment training, development supportive regulatory frameworks, fostering collaborations across sectors. concludes with future outlook, suggesting directions research developing robust scalable can integrate renewable sources smart grid technologies. serves valuable resource researchers, practitioners, policymakers engaged field efficiency AI-driven forecasting. Keywords: Machine Learning, Energy Efficiency, Demand Forecasting, Artificial Intelligence.

Language: Английский

Citations

33
A comprehensive review of artificial intelligence and machine learning applications in energy consumption and production DOI Creative Commons
Asif Raihan

Journal of Technology Innovations and Energy, Journal Year: 2023, Volume and Issue: 2(4), P. 1 - 26

Published: Oct. 19, 2023

The energy industry worldwide is today confronted with several challenges, including heightened levels of consumption and inefficiency, volatile patterns in demand supply, a dearth crucial data necessary for effective management. Developing countries face significant challenges due to the widespread occurrence unauthorized connections electricity grid, resulting substantial amounts unmeasured unpaid consumption. Nevertheless, implementation artificial intelligence (AI) machine learning (ML) technologies has potential improve management, efficiency, sustainability. Therefore, this study aims evaluate influence AI ML on progress industry. present employed systematic literature review methodology examine arising from frequent power outages limited accessibility various developing nations. results indicate that possess domains, predictive maintenance turbines, optimization consumption, management grids, prediction prices, assessment efficiency residential buildings. This concluded discussion measures enable nations harness advantages sector.

Language: Английский

Citations

23
Renewable energy sources integration via machine learning modelling: A systematic literature review DOI Creative Commons

Talal Alazemi,

Mohamed Darwish, Mohammed Radi

et al.

Heliyon, Journal Year: 2024, Volume and Issue: 10(4), P. e26088 - e26088

Published: Feb. 1, 2024

The use of renewable energy sources (RESs) at the distribution level has become increasingly appealing in terms costs and technology, expecting a massive diffusion near future placing several challenges to power grid. Since RESs depend on stochastic —solar radiation, temperature wind speed, among others— they introduce high uncertainty grid, leading imbalance deteriorating network stability. In this scenario, managing forecasting RES is vital successfully integrate them into grids. Traditionally, physical- statistical-based models have been used predict outputs. Nevertheless, former are computationally expensive since rely solving complex mathematical atmospheric dynamics, whereas latter usually consider linear models, preventing from addressing challenging scenarios. recent years, advances machine learning techniques, which can learn historical data, allowing analysis large-scale datasets either under non-uniform characteristics or noisy provided researchers with powerful data-driven tools that outperform traditional methods. paper, systematic literature review conducted identify most widely learning-based approaches forecast results show deep artificial neural networks, especially long-short term memory accurately model autoregressive nature output, ensemble strategies, allow handling large amounts highly fluctuating best suited ones. addition, promising integrating forecasted output decision-making problems, such as unit commitment, address economic, operational managerial grid discussed, solid directions for research provided.

Language: Английский

Citations

14