A Creep Model of Steel Slag–Asphalt Mixture Based on Neural Networks

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(13), P. 5820 - 5820

Published: July 3, 2024

To characterize the complex creep behavior of steel slag–asphalt mixture influenced by both stress and temperature, predictive models employing Back Propagation (BP) Long Short-Term Memory (LSTM) neural networks are described compared in this paper. Multiple repeated recovery tests on AC-13 grade mix samples were conducted at different temperatures. The experimental results processed into a group independent test results, then divided training testing datasets. K-fold cross-validation was applied to datasets fine-tune hyperparameters effectively. Compared with curves, effects BP LSTM investigated, broad applicability proven. performance trained model observed 95% confidence interval around fit errors, thereby strain intervals for dataset obtained. suggest that had enhanced prediction deformation trends various Due potent generalization strength artificial intelligence technology, can be further expanded forecasting road rutting deformations.

Language: Английский

---

Exploring LSTM-based Attention Mechanisms with PSO and Grid Search under Different Normalization Techniques for Energy demands Time Series Forecasting

Knowledge Engineering and Data Science, Journal Year: 2024, Volume and Issue: 7(1), P. 1 - 1

Published: April 16, 2024

Advanced analytical approaches are required to accurately forecast the energy sector's rising complexity and volume of time series data. This research aims demand utilizing sophisticated Long Short-Term Memory (LSTM) configurations with Attention mechanisms (Att), Grid search, Particle Swarm Optimization (PSO). In addition, study also examines influence Min-Max Z-Score normalization in preprocessing stage on accuracy performances baselines proposed models. PSO Search techniques used select best hyperparameters for LSTM models, while attention mechanism selects important input LSTM. The compares performance (LSTM, Grid-search-LSTM, PSO-LSTM) proposes models (Att-LSTM, Att-Grid-search-LSTM, Att-PSO-LSTM) based MAPE, RMSE, R2 metrics into two scenarios normalization: Min-Max, Z-Score. results show that all have better than those model is shown Att-PSO-LSTM MAPE 3.1135, RMSE 0.0551, 0.9233, followed by Att-LSTM, PSO-LSTM, These findings emphasize effectiveness improving predictions methods performance. study's novel approach provides valuable insights forecasting demands.

Language: Английский

---

Explainable AI and optimized solar power generation forecasting model based on environmental conditions

PLoS ONE, Journal Year: 2024, Volume and Issue: 19(10), P. e0308002 - e0308002

Published: Oct. 2, 2024

This paper proposes a model called X-LSTM-EO, which integrates explainable artificial intelligence (XAI), long short-term memory (LSTM), and equilibrium optimizer (EO) to reliably forecast solar power generation. The LSTM component forecasts generation rates based on environmental conditions, while the EO optimizes model’s hyper-parameters through training. XAI-based Local Interpretable Model-independent Explanation (LIME) is adapted identify critical factors that influence accuracy of in smart systems. effectiveness proposed X-LSTM-EO evaluated use five metrics; R-squared (R 2 ), root mean square error (RMSE), coefficient variation (COV), absolute (MAE), efficiency (EC). gains values 0.99, 0.46, 0.35, 0.229, 0.95, for R , RMSE, COV, MAE, EC respectively. results this improve performance original conventional LSTM, where improvement rate is; 148%, 21%, 27%, 20%, 134% compared with other machine learning algorithm such as Decision tree (DT), Linear regression (LR) Gradient Boosting. It was shown worked better than DT LR when were compared. Additionally, PSO employed instead validate outcomes, further demonstrated efficacy optimizer. experimental simulations demonstrate can accurately estimate PV response abrupt changes patterns. Moreover, might assist optimizing operations photovoltaic units. implemented utilizing TensorFlow Keras within Google Collab environment.

Language: Английский

---

Machine learning and feature engineering-based anode potential estimation method for lithium-ion batteries with application

Journal of Energy Storage, Journal Year: 2024, Volume and Issue: 103, P. 114387 - 114387

Published: Oct. 31, 2024

Language: Английский

---

Indonesian Stock Price Prediction Using Neural Basis Expansion Analysis for Interpretable Time Series Method

Journal of Information and Communication Technology, Journal Year: 2024, Volume and Issue: 23(3), P. 361 - 392

Published: July 28, 2024

The stock market is an attractive investment venue for many individuals and companies. However, unexpected share price fluctuations can cause significant financial losses. In investment, predicting movements the most frequently discussed topic because it allows investors to make right decisions big profits. Therefore, a model needed predict future prices, one strategy maximising New state-of-the-art deep learning architectures time series forecasting are being developed yearly, making them more accurate than ever. commonly used network such solution Long Short-Term Memory (LSTM) architecture, but has limitations as long training interpretability. This study aims evaluate another state-of-theart solution, Neural Basis Expansion Analysis Interpretable Time Series (N-BEATS), in comparison with LSTM by utilising historical data of PT Bank Central Asia Tbk (one banking companies Indonesia) from 25 March 2013 21 2023. N-BEATS relatively new variable method that produce predictions using neural networks. architecture advantages interpretability, seamless applicability across diverse target domains without requiring modifications, fast training. Based on tests carried out prediction errors measured Mean Average Percentage Error (MAPE), was found outperformed MAPE value 1.05 percent. conclusion, this research shows use algorithms which contributes facilitating buying selling investors.

Language: Английский

---

Research on Supply Chain Demand Prediction Model Based on LSTM

Procedia Computer Science, Journal Year: 2024, Volume and Issue: 243, P. 313 - 322

Published: Jan. 1, 2024

Language: Английский

---

Solar Radiation Prediction Based on TCN‐N‐BEATS Sequence Modeling

Advances in Meteorology, Journal Year: 2024, Volume and Issue: 2024(1)

Published: Jan. 1, 2024

Solar radiation prediction research is a key area of interest in the realm solar energy utilization and has garnered significant attention recent times. In order to realize accurate make better serve photovoltaic (PV) power generation, this study proposes method based on sequence model, which integrates two kinds neural networks, namely, temporal convolutional network (TCN) basis expansion analysis (N‐BEATS). First, dataset preprocessed using Pearson’s correlation coefficient, outlier detection, normalized obtain valid relevant data; second, features TCN feature extraction N‐BEATS flexible extension are integrated predict radiation; then, model’s hyperparameters fine‐tuned grid search algorithm ensure precise predictions; last, correctness verified by comparing error metrics running time. Empirical findings indicate that TCN‐N‐BEATS model high accuracy short time overhead, it certain application value prediction, could offer valuable insights for predicting radiation.

Language: Английский

---