Advances in Sparrow Search Algorithm: A Comprehensive Survey

Archives of Computational Methods in Engineering, Journal Year: 2022, Volume and Issue: 30(1), P. 427 - 455

Published: Aug. 22, 2022

Language: Английский

---

Battery SOH estimation method based on gradual decreasing current, double correlation analysis and GRU

Green Energy and Intelligent Transportation, Journal Year: 2023, Volume and Issue: 2(5), P. 100108 - 100108

Published: July 17, 2023

In intelligent lithium-ion battery management, the state of health (SOH) is essential for batteries' running in electric vehicles. Popularly, SOH estimated by using suitable features and data-driven methods. However, it difficult to extract appropriate characterizing from charging discharging data batteries owing various charges (SOCs) working conditions batteries. order effectively estimate SOH, an estimation method based on gradual decreasing current, double correlation analysis gated recurrent unit (GRU) proposed this paper. Firstly, current constant voltage phase measured as raw data. Then, select combined different categories features. Meanwhile, number input also ensured method. Finally, GRU algorithm employed set up a model whose learning rate improved sparrow search (SSA) purpose capturing hidden relationship between SOH. The adaptability validated experiments single pack. Additionally, contrast are performed show advanced performance

Language: Английский

---

A CNN-BiGRU-AM neural network for AI applications in shale oil production prediction

Applied Energy, Journal Year: 2023, Volume and Issue: 344, P. 121249 - 121249

Published: May 22, 2023

Language: Английский

---

Recent Versions and Applications of Sparrow Search Algorithm

Archives of Computational Methods in Engineering, Journal Year: 2023, Volume and Issue: unknown

Published: Feb. 7, 2023

Language: Английский

---

Hybrid neural network-based metaheuristics for prediction of financial markets: a case study on global gold market

Journal of Computational Design and Engineering, Journal Year: 2023, Volume and Issue: 10(3), P. 1110 - 1125

Published: April 29, 2023

Abstract Technical analysis indicators are popular tools in financial markets. These help investors to identify buy and sell signals with relatively large errors. The main goal of this study is develop new practical methods fake obtained from technical the precious metals market. In paper, we analyze these different ways based on recorded for 10 months. novelty research propose hybrid neural network-based metaheuristic algorithms analyzing them accurately while increasing performance indicators. We combine a convolutional network bidirectional gated recurrent unit whose hyperparameters optimized using firefly algorithm. To determine select most influential variables target variable, use another successful recently developed metaheuristic, namely, moth-flame optimization Finally, compare proposed models other state-of-the-art single deep learning machine literature. finding that metaheuristics can be useful as decision support tool address control enormous uncertainties

Language: Английский

---

Modeling and compensation of small-sample thermal error in precision machine tool spindles using spatial–temporal feature interaction fusion network

Advanced Engineering Informatics, Journal Year: 2024, Volume and Issue: 62, P. 102741 - 102741

Published: July 30, 2024

Language: Английский

---

Ensemble learning using multivariate variational mode decomposition based on the Transformer for multi-step-ahead streamflow forecasting

Journal of Hydrology, Journal Year: 2024, Volume and Issue: 636, P. 131275 - 131275

Published: May 7, 2024

Language: Английский

---

LSTM–Transformer-Based Robust Hybrid Deep Learning Model for Financial Time Series Forecasting

Sci, Journal Year: 2025, Volume and Issue: 7(1), P. 7 - 7

Published: Jan. 10, 2025

The inherent challenges of financial time series forecasting demand advanced modeling techniques for reliable predictions. Effective is crucial risk management and the formulation investment decisions. accurate prediction stock prices a subject study in domains investing national policy. This problem appears to be challenging due presence multi-noise, nonlinearity, volatility, chaotic nature stocks. paper proposes novel model based on deep learning ensemble LSTM-mTrans-MLP, which integrates long short-term memory (LSTM) network, modified Transformer multilayered perception (MLP). By integrating LSTM, Transformer, MLP, suggested demonstrates exceptional performance terms capabilities, robustness, enhanced sensitivity. Extensive experiments are conducted multiple datasets, such as Bitcoin, Shanghai Composite Index, China Unicom, CSI 300, Google, Amazon Stock Market. experimental results verify effectiveness robustness proposed LSTM-mTrans-MLP network compared with benchmark SOTA models, providing important inferences investors decision-makers.

Language: Английский

---

Nature-inspired optimization algorithms and their significance in multi-thresholding image segmentation: an inclusive review

Evolving Systems, Journal Year: 2022, Volume and Issue: 13(6), P. 889 - 945

Published: Feb. 21, 2022

Language: Английский

---

An improved neural network model for battery smarter state-of-charge estimation of energy-transportation system

Green Energy and Intelligent Transportation, Journal Year: 2023, Volume and Issue: 2(2), P. 100067 - 100067

Published: Jan. 20, 2023

The safety and reliability of battery storage systems are critical to the mass roll-out electrified transportation new energy generation. To achieve safe management optimal control batteries, state charge (SOC) is one important parameters. machine-learning based SOC estimation methods lithium-ion batteries have attracted substantial interests in recent years. However, a common problem with these models that their performances not always stable, which makes them difficult use practical applications. address this problem, an optimized radial basis function neural network (RBF-NN) combines concepts Golden Section Method (GSM) Sparrow Search Algorithm (SSA) proposed paper. Specifically, GSM used determine optimum number neurons hidden layer RBF-NN model, its parameters such as base center, connection weights so on by SSA, greatly improve performance estimation. In experiments, data collected from different working conditions demonstrate accuracy generalization ability results experiment indicate maximum error model less than 2%.

Language: Английский

---