Optimizing deep neural network architectures for renewable energy forecasting

Discover Sustainability, Journal Year: 2024, Volume and Issue: 5(1)

Published: Nov. 12, 2024

An accurate renewable energy output forecast is essential for efficiency and power system stability. Long Short-Term Memory(LSTM), Bidirectional LSTM(BiLSTM), Gated Recurrent Unit(GRU), Convolutional Neural Network-LSTM(CNN-LSTM) Deep Network (DNN) topologies are tested solar wind production forecasting in this study. ARIMA was compared to the models. This study offers a unique architecture Networks (DNNs) that specifically tailored forecasting, optimizing accuracy by advanced hyperparameter tuning incorporation of meteorological temporal variables. The optimized LSTM model outperformed others, with MAE (0.08765), MSE (0.00876), RMSE (0.09363), MAPE (3.8765), R2 (0.99234) values. GRU, CNN-LSTM, BiLSTM models predicted well. Meteorological time-based factors enhanced accuracy. addition sun data improved its prediction. results show deep neural network can predict energy, highlighting importance carefully selecting characteristics fine-tuning model. work improves estimates promote more reliable environmentally sustainable electricity system.

Language: Английский

---

Renewable Wind Energy Implementation in South America: A Comprehensive Review and Sustainable Prospects

Sustainability, Journal Year: 2024, Volume and Issue: 16(14), P. 6082 - 6082

Published: July 16, 2024

South America is a region that stands out worldwide for its biodiversity of ecosystems, cultural heritage, and potential considering natural resources linked to renewable energies. In the global crisis due climate change, American countries have implemented actions carry progressive energy transition from fossil energies contribute planet’s sustainability. this context, are implementing green strategies investment projects wind farms move towards achieving sustainable development goals year 2030 UN agenda low-carbon economies 2050. This article studies advances in implementation America, highlighting progress experiences these issues through review scientific literature 2023. The methodology applied was carried Preferred Reporting Items Systematic Reviews Meta-Analyses (PRISMA) guidelines generation maps. As result, presents main developments, lessons learned/gaps, future prospects on road According results, infrastructure during change era. Different levels on-shore been reached each country. Also, promising exists off-shore highest potential. Finally, concludes emerging technologies like production hydrogen synthetic e-fuels looks synergetic clean solution combined with energy, which may transform into world-class territory.

Language: Английский

---

Microstructural and Phase Degradation of Monocrystalline Solar Photovoltaic Panels Under Extreme Desert Conditions: Insights from XRD and FTIR Analysis

Materials Chemistry and Physics, Journal Year: 2025, Volume and Issue: unknown, P. 130742 - 130742

Published: March 1, 2025

Language: Английский

---

Solar selective absorbers via electrophoretic deposition: A comparative and critical review of the method

Materials Today Communications, Journal Year: 2025, Volume and Issue: unknown, P. 112621 - 112621

Published: April 1, 2025

Language: Английский

---

Digital twin technology and artificial intelligence in energy transition: A comprehensive systematic review of applications

Energy Reports, Journal Year: 2025, Volume and Issue: 13, P. 5196 - 5218

Published: May 3, 2025

Language: Английский

---

Comparative analysis of deep neural network architectures for renewable energy forecasting: enhancing accuracy with meteorological and time-based features

Discover Sustainability, Journal Year: 2024, Volume and Issue: 5(1)

Published: Dec. 31, 2024

This study evaluates and differentiates five advanced machine learning models—LSTM, GRU, CNN-LSTM, Random Forest, SVR—aimed at precisely estimating solar wind power generation to enhance renewable energy forecasting. LSTM achieved a remarkable Mean Squared Error (MSE) of 0.010 R2 score 0.90, highlighting its proficiency in capturing intricate temporal relationships. GRU closely followed, demonstrating potential as viable option due combination computational efficiency accuracy (MSE = 0.015, 0.88). In datasets abundant spatial correlations, the CNN-LSTM hybrid demonstrated utility by providing novel insights into spatial–temporal patterns; nonetheless, it lagged considerably accuracy, with mean squared error 0.020 0.87. Conversely, traditional models reliable albeit less dynamic ability elucidate complexities data; for instance, Forest exhibited 0.025, while Support Vector Regression (SVR) recorded an MSE 0.030. The results affirm that deep architectures, particularly LSTM, offer transformative method forecasting, hence enhancing reliability management systems.

Language: Английский

---

Program Module for Locating Charging Stations for Electric Light Vehicles in a Settlement

Published: March 19, 2024

One of the challenges resulting from proliferation such vehicles is need to create a network infrastructure that allows charging devices used. In this regard, paper considers process implementing software module for arrangement stations electric vehicles, based on use geographic information systems used as source input data and development environment Python, with which all basic computational procedures are implemented. This plug-in QGIS, placement points small along selected route (road network). Unlike previous solution, developed aimed at working large spaces, cities towns, full-fledged standalone application does not require installation additional software.

Language: Английский

---

Optimizing Lightweight Recurrent Networks for Solar Forecasting in TinyML: Modified Metaheuristics and Legal Implications

Energies, Journal Year: 2024, Volume and Issue: 18(1), P. 105 - 105

Published: Dec. 30, 2024

The limited nature of fossil resources and their unsustainable characteristics have led to increased interest in renewable sources. However, significant work remains be carried out fully integrate these systems into existing power distribution networks, both technically legally. While reliability holds great potential for improving energy production sustainability, the dependence solar plants on weather conditions can complicate realization consistent without incurring high storage costs. Therefore, accurate prediction is vital efficient grid management trading. Machine learning models emerged as a prospective solution, they are able handle immense datasets model complex patterns within data. This explores use metaheuristic optimization techniques optimizing recurrent forecasting predict from substations. Additionally, modified optimizer introduced meet demanding requirements optimization. Simulations, along with rigid comparative analysis other contemporary metaheuristics, also conducted real-world dataset, best achieving mean squared error (MSE) just 0.000935 volts 0.007011 two datasets, suggesting viability usage. best-performing further examined applicability embedded tiny machine (TinyML) applications. discussion provided this manuscript includes legal framework forecasting, its integration, policy implications establishing decentralized cost-effective system.

Language: Английский

---

Geospatial Data in the Design of an Intelligent Information and Analytics System for Renewable Energy

Published: March 19, 2024

This paper reviews examples of some existing open source renewable energy GIS and shows two ways to integrate with such systems, using the example a module for locating charging stations electric vehicles assessing possible involvement non-conventional sources in system an individual consumer.

Language: Английский

---

Improving wind power forecast accuracy for optimal hybrid system energy management

Journal of Energy Resources Technology, Journal Year: 2024, Volume and Issue: 146(9)

Published: May 20, 2024

Abstract Due to its renewable and sustainable features, wind energy is growing around the world. However, speed fluctuation induces intermittent character of generated power. Thus, power estimation, through forecasting, very inherent ensure effective scheduling. Four predictors based on deep learning networks optimization algorithms were developed. The designed topologies are multilayer perceptron neural network, long short-term memory convolutional bidirectional network coupled with Bayesian optimization. models' performance was evaluated evaluation indicators mainly, root mean squared error, absolute percentage. Based simulation results, all them show considerable prediction results. Moreover, combination algorithm more robust in forecasting a error equal 0.23 m/s. estimated investigated for optimal Wind/Photovoltaic/Battery/Diesel management. handling approach lies continuity load supply sources as priority, batteries second order, finally diesel. proposed management strategy respects criteria satisfactory contribution percentage 71%.

Language: Английский

---