Hybrid framework combining grey system model with Gaussian process and STL for CO2 emissions forecasting in developed countries

Applied Energy, Год журнала: 2024, Номер 360, С. 122824 - 122824

Опубликована: Фев. 20, 2024

Язык: Английский

---

Leveraging LSTM-SMI and ARIMA architecture for robust wind power plant forecasting

Applied Soft Computing, Год журнала: 2025, Номер 170, С. 112765 - 112765

Опубликована: Янв. 15, 2025

Язык: Английский

---

Machine learning boosts wind turbine efficiency with smart failure detection and strategic placement

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Янв. 9, 2025

The research study objective seeks to improve the efficiency of wind turbines using state-of-the-art techniques in domain ML, making energy key player fashioning a favorable future. Wind Turbine Health Monitoring (WTHM) is typically achieved through either vibration analysis or by Supervisory Control and Data Acquisition (SCADA) data turbines, wherein conventional fault pattern identification time-consuming, guesswork process. This work proposed an intelligent automated approach early detection implementation HARO (Huber Adam Regression Optimizer) model, which combines Transformer networks with Lasso optimizer. model traditional models, including Huber Automatic Relevance Determination (ARD) Regressors, since Transformers are capable learning patterns sensors. overall results obtained illustrated that Housing Asset Repair Optimization tool had reduced downtime while enhancing, at same time, accuracy prediction future faults enable right-time planning for maintenance activities. aspect enhances process, minimizing human input hence slightly lessens HARO. further calls collective practice amongst researchers, practitioners policymakers address emerging issues power avoid duplication continuously keep improving on strategies employed towards energy. paper also shows ML has possibility turbine dependability establishing as essential component renewable systems.

Язык: Английский

---

Optimization of Sample Size, Data Points, and Data Augmentation Stride in Vibration Signal Analysis for Deep Learning‐Based Fault Diagnosis of Rotating Machines

Shock and Vibration, Год журнала: 2025, Номер 2025(1)

Опубликована: Янв. 1, 2025

In recent years, deep learning models have increasingly been employed for fault diagnosis in rotating machines, with remarkable results. However, the accuracy and reliability of these tasks can be significantly influenced by critical input parameters, such as sample size, number data points within each sample, augmentation stride vibration signal analysis. To address this challenge, paper proposes a new adaptive method based on Bayesian optimization to determine optimal combination parameters from raw signals enhance diagnostic performance models. This study utilizes one‐dimensional convolutional neural network (1‐D CNN) model classification. The proposed 1‐D CNN‐based is validated via collected motor rolling bearings achieves 100%. Compared existing methods, approach not only highest testing set but also demonstrates stable during training, even under varying operating conditions. These results indicate importance optimizing tasks.

Язык: Английский

---

Advanced techniques for wind energy production forecasting: Leveraging multi-layer Perceptron + Bayesian optimization, ensemble learning, and CNN-LSTM models

Case Studies in Chemical and Environmental Engineering, Год журнала: 2024, Номер 10, С. 100881 - 100881

Опубликована: Авг. 3, 2024

Due to the shortage of fossil fuels in many countries, power plants that rely on will be phased out favor wind turbines as primary source energy generation. These fuel wreak havoc natural world, making humans and other life forms susceptible illness. The production potential was investigated. Consequently, methods such XGBOOST, Multi-Layer Perceptron with Bayesian Optimization (MLP + BO), Gradient Boosting Regression Tree (GBDT), Ensemble (gradient boosting xgboost), CNN Long Short-Term Memory (CNN-LSTM) have been utilized. A mean square error (MSE) 7.2 45 seconds achieved using technique, an MSE 6.8 450 obtained CNN-LSTM method. Wind is readily available straightforward acquire globally, indicating its a reliable sustainable source.

Язык: Английский

---

Optimization of 2024-T3 Aluminum Alloy Friction Stir Welding Using Random Forest, XGBoost, and MLP Machine Learning Techniques

Materials, Год журнала: 2024, Номер 17(7), С. 1452 - 1452

Опубликована: Март 22, 2024

This study optimized friction stir welding (FSW) parameters for 1.6 mm thick 2024T3 aluminum alloy sheets. A 3 × factorial design was employed to explore tool rotation speeds (1100 1300 rpm) and (140 180 mm/min). Static tensile tests revealed the joints' maximum strength at 87% relative base material. Hyperparameter optimization conducted machine learning (ML) models, including random forest XGBoost, multilayer perceptron artificial neural network (MLP-ANN) using grid search. Welding parameter extrapolation were then carried out, with final predictions analyzed response surface methodology (RSM). The ML models achieved over 98% accuracy in regression, demonstrating significant effectiveness FSW process enhancement. Experimentally validated, resulted an joint efficiency of 93% outcome highlights critical role advanced analytical techniques improving quality efficiency.

Язык: Английский

---

ML: Early Breast Cancer Diagnosis

Current Problems in Cancer Case Reports, Год журнала: 2024, Номер 13, С. 100278 - 100278

Опубликована: Янв. 8, 2024

Breast cancer is the most common malignancy among women worldwide, often characterized by uncontrolled proliferation of breast cells, leading to formation lumps or tumors that can be detected through medical imaging such as X-rays. Distinguishing between benign and malignant presents a significant challenge in diagnosis cancer. In this study, machine learning methods, including Logistic Regression, Gradient Boosting, Ada Boost, Random Forest, Gaussian NB with Grid Search, were employed differentiate healthy individuals those malignancies. The results revealed Forest algorithm exhibited highest performance predicting cancer, accurately identifying 99% both affected individuals. Additionally, Boosting Boost demonstrated similar level accuracy, correctly distinguishing 98% Conversely, performed least effectively, an accuracy 91% differentiating individuals, highlighting its comparatively lower predictive capability for

Язык: Английский

---

DFTQuake: Tripartite Fourier attention and dendrite network for real-time early prediction of earthquake magnitude and peak ground acceleration

Engineering Applications of Artificial Intelligence, Год журнала: 2025, Номер 144, С. 110077 - 110077

Опубликована: Янв. 27, 2025

Язык: Английский

---

Non-Destructive Sweetness Classification of Khao Tang Kwa Pomelos Using Machine Learning with Acoustic and Image Processing

Journal of Food Composition and Analysis, Год журнала: 2025, Номер 142, С. 107385 - 107385

Опубликована: Фев. 22, 2025

Язык: Английский

---

PM2.5 concentration prediction using machine learning algorithms: an approach to virtual monitoring stations

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Март 8, 2025

One of the most important pollutants is PM2.5, which particularly to monitor pollutant levels keep concentration under control. In this research, an attempt has been made predict concentrations PM2.5 using four Machine Learning (ML) models. The ML methods include Light Gradient Boosting (LGBM), Extreme Regressor (XGBR), Random Forest (RF) and (GBR). mean maximum were recorded 32.84 µg/m3 160.25 µg/m2, respectively, indicating occurrence occasional episodes high pollution from 2016 2022. dropped below 30 µg/m2 in 2018 due reduced human activities during COVID-19 lockdowns but significantly increased because ongoing operation heavy industries post-COVID-19 2021. models performed very well predicting with around 95% their predictions falling within factor observed concentration. results presented that among algorithms, GBR confirmed good model performance compared other models, lowest MSE (5.33) RMSE (2.31), as accuracy measures. This suggests best for reducing large errors, making it more robust capturing variations levels. conclusion, study proposed a method obtain high-accuracy prediction are useful air quality monitoring on global scale improving acute exposure assessment epidemiological research.

Язык: Английский

---