Power supply quality prediction method based on LSTM and self-attention mechanism DOI
Yan Yang,

Yu Chang

Journal of Computational Methods in Sciences and Engineering, Journal Year: 2025, Volume and Issue: unknown

Published: April 26, 2025

Existing LSTM-based power quality (PQ) prediction models primarily rely on historical information, which limits their ability to fully capture contextual dependencies. Furthermore, these process inputs sequentially without accounting for the varying importance of different time steps, leading significant inaccuracies. To address limitations, this study proposes an enhanced PQ model that integrates Bidirectional Long Short-Term Memory (BiLSTM) with a Self-Attention (SA) mechanism. The BiLSTM module is introduced both forward and backward temporal dependencies, enabling more comprehensive long-term patterns in series data. SA mechanism dynamically adjusts steps through weighted summation, enhancing model’s focus critical features improving its capacity nonlinear relationships. from layer are then mapped connected generate final outputs. Experiments were conducted using data Nanchang as primary dataset, additional datasets Nanjing, Wuhan, Changsha, Beijing used generalization testing. results demonstrate BiLSTM-SA outperforms traditional LSTM across all metrics, achieving mean absolute error (MAE) 0.09 voltage deviation, 0.05 improvement over single-layer LSTM. Notably, maintains robust performance complex supply scenarios, generalized MAE only 0.2 Beijing. These findings highlight effectiveness combining reducing errors ensuring stability quality, offering advancement methodologies.

Language: Английский

Machine Learning Applications in Building Energy Systems: Review and Prospects DOI Creative Commons

D. Li,

Zhenzhen Qi,

Yiming Zhou

et al.

Buildings, Journal Year: 2025, Volume and Issue: 15(4), P. 648 - 648

Published: Feb. 19, 2025

Building energy systems (BESs) are essential for modern infrastructure but face significant challenges in equipment diagnosis, consumption prediction, and operational control. The complexity of BESs, coupled with the increasing integration renewable sources, presents difficulties fault detection, accurate forecasting, dynamic system optimisation. Traditional control strategies struggle low efficiency, slow response times, limited adaptability, making it difficult to ensure reliable operation optimal management. To address these issues, researchers have increasingly turned machine learning (ML) techniques, which offer promising solutions improving scheduling, real-time BESs. This review provides a comprehensive analysis ML techniques applied According results literature review, supervised methods, such as support vector machines random forest, demonstrate high classification accuracy detection require extensive labelled datasets. Unsupervised approaches, including principal component clustering algorithms, robust identification capabilities without data may complex nonlinear patterns. Deep particularly convolutional neural networks long short-term memory models, exhibit superior forecasting Reinforcement further enhances management by dynamically adjusting parameters maximise efficiency cost savings. Despite advancements, remain terms availability, computational costs, model interpretability. Future research should focus on hybrid integrating explainable AI enhancing adaptability evolving demands. also highlights transformative potential BESs outlines future directions sustainable intelligent building

Language: Английский

Citations

2
Optimal planning for integrated electricity and heat systems using CNN-BiLSTM-attention network forecasts DOI
Feng Li,

Shiheng Liu,

Tian-Hu Wang

et al.

Energy, Journal Year: 2024, Volume and Issue: 309, P. 133042 - 133042

Published: Aug. 31, 2024

Language: Английский

Citations

11
Using Crafted Features and Polar Bear Optimization Algorithm for Short-Term Electric Load Forecast System DOI Creative Commons

Mansi Bhatnagar,

Gregor Rozinaj, Radoslav Vargic

et al.

Energy and AI, Journal Year: 2025, Volume and Issue: unknown, P. 100470 - 100470

Published: Jan. 1, 2025

Language: Английский

Citations

1
Capturing complex electricity load patterns: A hybrid deep learning approach with proposed external-convolution attention DOI Creative Commons
Mohammad Sadegh Zare, Mohammad Reza Nikoo, Mingjie Chen

et al.

Energy Strategy Reviews, Journal Year: 2025, Volume and Issue: 57, P. 101638 - 101638

Published: Jan. 1, 2025

Language: Английский

Citations

1
Deep-Learning-Based Scheduling Optimization of Wind-Hydrogen-Energy Storage System on Energy Islands DOI
Qingxia Wu, Peng Long, Guoqing Han

et al.

Energy, Journal Year: 2025, Volume and Issue: unknown, P. 135107 - 135107

Published: Feb. 1, 2025

Language: Английский

Citations

1
Short- and medium-term power load forecasting model based on a hybrid attention mechanism in the time and frequency domains DOI

Z. J. Peng,

Xiaoyang Yang

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 127329 - 127329

Published: March 1, 2025

Language: Английский

Citations

1
Applications of Long Short-Term Memory (LSTM) Networks in Polymeric Sciences: A Review DOI Open Access
Ivan Malashin, В С Тынченко, Andrei Gantimurov

et al.

Polymers, Journal Year: 2024, Volume and Issue: 16(18), P. 2607 - 2607

Published: Sept. 14, 2024

This review explores the application of Long Short-Term Memory (LSTM) networks, a specialized type recurrent neural network (RNN), in field polymeric sciences. LSTM networks have shown notable effectiveness modeling sequential data and predicting time-series outcomes, which are essential for understanding complex molecular structures dynamic processes polymers. delves into use models polymer properties, monitoring polymerization processes, evaluating degradation mechanical performance Additionally, it addresses challenges related to availability interpretability. Through various case studies comparative analyses, demonstrates different science applications. Future directions also discussed, with an emphasis on real-time applications need interdisciplinary collaboration. The goal this is connect advanced machine learning (ML) techniques science, thereby promoting innovation improving predictive capabilities field.

Language: Английский

Citations

7
Multimodal Fusion of Optimized GRU–LSTM with Self-Attention Layer for Hydrological Time Series Forecasting DOI
Hüseyin Çağan Kılınç,

Sina Apak,

Furkan Ozkan

et al.

Water Resources Management, Journal Year: 2024, Volume and Issue: 38(15), P. 6045 - 6062

Published: Aug. 17, 2024

Language: Английский

Citations

6
Remote work might unlock solar PV's potential of cracking the ‘Duck Curve’ DOI Creative Commons
Kumar Biswajit Debnath, David P. Jenkins, Sandhya Patidar

et al.

Applied Energy, Journal Year: 2024, Volume and Issue: 367, P. 123378 - 123378

Published: May 16, 2024

Integrating renewable energy technologies into a decentralised smart grid presents the 'Duck Curve' challenge — disparity between peak demand and solar photovoltaic (PV) yield. Smart operators still lack an effective solution to this problem, resulting in need maintain standby fossil fuel-fired plants. The COVID-19 pandemic-induced lockdowns necessitated shift remote work (work-from-home) home-based education. primary objective of study was explore mitigating strategies for duck curve by investigating notable behaviour examining effect education on PV electricity use 100 households with battery storage southwest UK. This examined 1-min granular consumption data April–August 2019 2020. findings revealed statistically significant disparities demand. Notably, there 1.4—10% decrease average from April August 2020 (during following lockdown) compared corresponding months 2019. Furthermore, household reduced 24—25%, while self-consumption systems increased 7—8% during lockdown May increase particularly prominent morning afternoon, possibly attributed growing prevalence work-from-home dynamic shifts patterns emphasised role meeting evolving needs unprecedented societal changes. Additionally, might unlock PV's potential resolving Curve', urging further investigation implications infrastructure policy development.

Language: Английский

Citations

4
Identification and Correction of Abnormal, Incomplete Power Load Data in Electricity Spot Market Databases DOI Creative Commons
Jingjiao Li, Yifan Lv,

Zhou Zhou

et al.

Energies, Journal Year: 2025, Volume and Issue: 18(1), P. 176 - 176

Published: Jan. 3, 2025

The development of electricity spot markets necessitates more refined and accurate load forecasting capabilities to enable precise dispatch control the creation new trading products. Accurate relies on high-quality historical data, with complete data serving as cornerstone for both transactions in markets. However, at distribution network or user level often suffers from anomalies missing values. Data-driven methods have been widely adopted anomaly detection due their independence prior expert knowledge physical models. Nevertheless, single architectures struggle adapt diverse characteristics networks users, hindering effective capture patterns. This paper proposes a PLS-VAE-BiLSTM-based method identification correction by combining strengths Variational Autoencoders (VAE) Bidirectional Long Short-Term Memory Networks (BiLSTM). begins preprocessing, including normalization preliminary value imputation based Partial Least Squares (PLS). Subsequently, hybrid VAE-BiLSTM model is constructed trained loaded dataset incorporating influencing factors learn relationships between different features. Anomalies are identified corrected calculating deviation model’s reconstructed values actual Finally, validation public private datasets demonstrates that PLS-VAE-BiLSTM achieves average performance metrics 98.44% precision, 94% recall rate, 96.05% F1 score. Compared VAE-LSTM, PSO-PFCM, WTRR models, proposed exhibits superior overall performance.

Language: Английский

Citations

0