Lithium Battery Enhancement Through Electrical Characterization and Optimization Using Deep Learning

World Electric Vehicle Journal, Journal Year: 2025, Volume and Issue: 16(3), P. 167 - 167

Published: March 13, 2025

Research on lithium-ion batteries has been driven by the growing demand for electric vehicles to mitigate greenhouse gas emissions. Despite advances, still face significant challenges in efficiency, lifetime, safety, and material optimization. In this context, objective of research is develop a predictive model based Deep deep-Learning learning techniques. Based Learning techniques that combine Transformer Physicsphysics-Informed informed approaches optimization design electrochemical parameters improve performance lithium batteries. Also, we present training database consisting three key components: numerical simulation using Doyle–Fuller–Newman (DFN) mathematical model, experimentation with half-cell configured zinc oxide anode, set commercial battery discharge curves electronic monitoring. The results show developed Transformer–Physics physics-Informed can effectively integrate deep deep-learning DNF make predictions behavior estimate battery-charge capacity an average error 2.5% concerning experimental data. addition, it was observed could explore new allow evaluation without requiring invasive analysis their internal structure. This suggests assess optimize various applications, which significantly impact industry its use Electric Vehicles (EVs).

Language: Английский

---

A Novel Hybrid Approach Using an Attention-Based Transformer + GRU Model for Predicting Cryptocurrency Prices

Mathematics, Journal Year: 2025, Volume and Issue: 13(9), P. 1484 - 1484

Published: April 30, 2025

In this article, we introduce a novel deep learning hybrid model that integrates attention Transformer and gated recurrent unit (GRU) architectures to improve the accuracy of cryptocurrency price predictions. By combining Transformer’s strength in capturing long-range patterns with GRU’s ability short-term sequential trends, provides well-rounded approach time series forecasting. We apply predict daily closing prices Bitcoin Ethereum based on historical data include past prices, trading volumes, Fear Greed Index. evaluate performance our proposed by comparing it four other machine models, two are non-sequential feedforward models: radial basis function network (RBFN) general regression neural (GRNN), bidirectional memory-based long memory (BiLSTM) (BiGRU). The model’s is assessed using several metrics, including mean squared error (MSE), root (RMSE), absolute (MAE), percentage (MAPE), along statistical validation through non-parametric Friedman test followed post hoc Wilcoxon signed-rank test. Results demonstrate consistently achieves superior accuracy, highlighting its effectiveness for financial prediction tasks. These findings provide valuable insights enhancing real-time decision making markets support growing use models analytics.

Language: Английский

---

Plant diseases classification with Spectral Signature Taxonomy & Analysis Software (SSTAS)

Software Impacts, Journal Year: 2025, Volume and Issue: unknown, P. 100744 - 100744

Published: March 1, 2025

Language: Английский

---

transformerForecasting: Transformer Deep Learning Model for Time Series Forecasting

Published: March 7, 2025

Language: Английский

---

The Future of Real-Time Analytics : AI-Driven Insights at Scale

International Journal of Scientific Research in Computer Science Engineering and Information Technology, Journal Year: 2024, Volume and Issue: 10(6), P. 703 - 712

Published: Nov. 20, 2024

Real-time analytics is experiencing a transformative evolution driven by artificial intelligence and cloud computing advancements. This comprehensive article explores cutting-edge developments in AI-powered systems, examining their impact across stream processing engines, query optimization, predictive analytics, cloud-native architectures. The investigates how modern systems leverage deep learning, reinforcement transformer models to enhance capabilities, optimize resource utilization, enable sophisticated insights. Through detailed examination of adaptive processing, state management advances, edge integration, this analysis demonstrates AI-driven approaches are revolutionizing data efficiency, scalability, performance optimization. highlights significant improvements areas such as automated scaling, workload prediction, management, pipeline showcasing these technologies organizations generate actionable insights from real-time streams while maintaining high cost efficiency.

Language: Английский

---