A hybrid optimization algorithm to identify unknown parameters of photovoltaic models under varying operating conditions

Engineering Applications of Artificial Intelligence, Год журнала: 2024, Номер 133, С. 108544 - 108544

Опубликована: Май 8, 2024

Язык: Английский

---

Solar radiation forecasting with deep learning techniques integrating geostationary satellite images

Engineering Applications of Artificial Intelligence, Год журнала: 2022, Номер 116, С. 105493 - 105493

Опубликована: Окт. 10, 2022

Язык: Английский

---

A multistep short-term solar radiation forecasting model using fully convolutional neural networks and chaotic aquila optimization combining WRF-Solar model results

Energy, Год журнала: 2023, Номер 271, С. 126980 - 126980

Опубликована: Фев. 20, 2023

Язык: Английский

---

Hybrid solar radiation forecasting model with temporal convolutional network using data decomposition and improved artificial ecosystem-based optimization algorithm

Energy, Год журнала: 2023, Номер 280, С. 128171 - 128171

Опубликована: Июнь 18, 2023

Язык: Английский

---

Forecasting solar energy production in Spain: A comparison of univariate and multivariate models at the national level

Applied Energy, Год журнала: 2023, Номер 350, С. 121645 - 121645

Опубликована: Авг. 10, 2023

Renewable energies, such as solar power, offer a clean and cost-effective energy source. However, their integration into national electricity grids poses challenges due to dependence on climate geography. While numerous studies have focused time series, few specifically addressed the critical task of forecasting production at level. Accurate national-level is crucial for optimizing management, informing policy development, promoting environmental sustainability. This study aims address associated with significant variability in renewable its impact grid stability by improving accuracy existing approaches. To achieve this goal, we evaluate effectiveness univariate multivariate approaches series data from ESIOS (the Spanish System Operator). Our primary focus leveraging external variables, irradiance data. end, propose methodology integrate forecasts historical power plants Spain improve performance models. Subsequently, compare classical regression techniques state-of-the-art deep learning algorithms, presenting models three forecast horizons (1 h, 24 48 h). Finally, assess our best comparing them official ESIOS. findings indicate that best-performing are deep-learning approaches, which benefit incorporating forecasts, particularly longer (24 h h), avoid detrimental effects Hughes Phenomenon, seems hamper non-deep-learning forecasters. The top-performing models, based Convolutional Networks + Recurrent Neural Networks, outperform reducing mean absolute error 41% 47.58%, respectively.

Язык: Английский

---

A Study of Optimization in Deep Neural Networks for Regression

Electronics, Год журнала: 2023, Номер 12(14), С. 3071 - 3071

Опубликована: Июль 14, 2023

Due to rapid development in information technology both hardware and software, deep neural networks for regression have become widely used many fields. The optimization of (DNNR), including selections data preprocessing, network architectures, optimizers, hyperparameters, greatly influence the performance tasks. Thus, this study aimed collect analyze recent literature surrounding DNNR from aspect optimization. In addition, various platforms conducting models were investigated. This has a number contributions. First, it provides sections models. Then, elements each section are listed analyzed. Furthermore, delivers insights critical issues related Optimizing simultaneously instead individually or sequentially could improve Finally, possible potential directions future provided.

Язык: Английский

---

Half-hourly electricity price prediction with a hybrid convolution neural network-random vector functional link deep learning approach

Applied Energy, Год журнала: 2024, Номер 374, С. 123920 - 123920

Опубликована: Июль 31, 2024

Digital technologies with predictive modelling capabilities are revolutionizing electricity markets, especially in demand-side management. Accurate price prediction is essential deregulated markets; however, developing effective models challenging due to high-frequency fluctuations and volatility. This study introduces a hybrid system that addresses these challenges through comprehensive data processing framework for half-hourly predictions. The preprocessing stage employs the Maximum Overlap Discrete Wavelet Transform (MoDWT) enhance input quality by reducing overlap revealing underlying patterns. model integrates Convolutional Neural Networks Random Vector Functional Link (CRVFL) deep learning approach. Bayesian Optimization fine-tunes MoDWT-CRVFL optimal performance. Validation of conducted using prices from New South Wales. results highlight efficacy model, achieving high accuracy superior Global Performance Indicator (GPI) values approximately 1.61, 1.33, 1.85, 1.30, 0.78 Summer, Autumn, Winter, Spring, Annual (Year 2022), respectively, outperforming alternative models. Similarly, Kling–Gupta Efficiency (KGE) metrics proposed consistently surpassed those both decomposition-based standalone For instance, KGE value was 0.972, significantly higher than 0.958, 0.899, 0.963, 0.943, 0.930, 0.661, 0.708, 0.696, 0.739, 0.738 MoDWT-LSTM, MoDWT-DNN, MoDWT-XGB, MoDWT-RF, MoDWT-MLP, Bi-LSTM, LSTM, DNN, RF, XGB, MLP, respectively. methodologies this optimize energy resource allocation, market prices, network management, empowering operators make informed decisions resilient efficient market.

Язык: Английский

---

Point-based and probabilistic electricity demand prediction with a Neural Facebook Prophet and Kernel Density Estimation model

Engineering Applications of Artificial Intelligence, Год журнала: 2024, Номер 135, С. 108702 - 108702

Опубликована: Июнь 10, 2024

Язык: Английский

---

Novel deep learning hybrid models (CNN-GRU and DLDL-RF) for the susceptibility classification of dust sources in the Middle East: a global source

Scientific Reports, Год журнала: 2022, Номер 12(1)

Опубликована: Ноя. 11, 2022

Abstract Dust storms have many negative consequences, and affect all kinds of ecosystems, as well climate weather conditions. Therefore, classification dust storm sources into different susceptibility categories can help us mitigate its effects. This study aimed to classify the in Middle East (ME) by developing two novel deep learning (DL) hybrid models based on convolutional neural network–gated recurrent unit (CNN-GRU) model, dense layer learning–random forest (DLDL-RF) model. The Dragonfly algorithm (DA) was used identify critical features controlling sources. Game theory for interpretability DL model’s output. Predictive were constructed dividing datasets randomly train (70%) test (30%) groups, six statistical indicators being then applied assess model performance both (train test). Among 13 potential (or variables) sources, seven variables selected important non-important DA, respectively. Based DLDL-RF – a with higher accuracy comparison CNN-GRU–23.1, 22.8, 22.2% area classified very low, low moderate susceptibility, whereas 20.2 11.7% representing high classes, clay content, silt precipitation identified three most game through permutation values. Overall, found be efficient methods prediction purposes large spatial scales no or incomplete from ground-based measurements.

Язык: Английский

---

Integrated Multi-Head Self-Attention Transformer model for electricity demand prediction incorporating local climate variables

Energy and AI, Год журнала: 2023, Номер 14, С. 100302 - 100302

Опубликована: Сен. 23, 2023

This paper develops a trustworthy deep learning model that considers electricity demand (G) and local climate conditions. The utilises Multi-Head Self-Attention Transformer (TNET) to capture critical information from G, attain reliable predictions with (rainfall, radiation, humidity, evaporation, maximum minimum temperatures) data Energex substations in Queensland, Australia. TNET is then evaluated models (Long-Short Term Memory LSTM, Bidirectional LSTM BILSTM, Gated Recurrent Unit GRU, Convolutional Neural Networks CNN, Deep Network DNN) based on robust assessment metrics. Kernel Density Estimation method used generate the prediction interval (PI) of forecasts derive probability metrics results show developed accurate for all substations. study concludes proposed predictive tool has high accuracy low errors could be employed as stratagem by modellers energy policy-makers who wish incorporate climatic factors into patterns develop national market insights analysis systems.

Язык: Английский

---