Hyperparameter Tuning of Load-Forecasting Models Using Metaheuristic Optimization Algorithms—A Systematic Review DOI Creative Commons
Umme Mumtahina, Sanath Alahakoon, Peter Wolfs

et al.

Mathematics, Journal Year: 2024, Volume and Issue: 12(21), P. 3353 - 3353

Published: Oct. 25, 2024

Load forecasting is an integral part of the power industries. Load-forecasting techniques should minimize percentage error while prediction future demand. This will inherently help utilities have uninterrupted supply. In addition to that, accurate load can result in saving large amounts money. article provides a systematic review based on Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) framework. presents complete framework short-term using metaheuristic algorithms. consists three sub-layers: data-decomposition layer, optimization layer. The layer decomposes input data series extract important features. used predict result, which involves different statistical machine-learning models. optimizes parameters methods improve accuracy stability model Single models from results. However, they come with their limitations, such as low accuracy, high computational burden, stuck local minima, etc. To hyperparameters these need be tuned properly. Metaheuristic algorithms cab tune considering interdependencies. Hybrid combining three-layer perform better by overcoming issues premature convergence trapping into minimum solution. A quantitative analysis deep-learning presented. Some most common evaluation indices that are evaluate performance discussed. Furthermore, taxonomy state-of-the-art articles provided, discussing advantages, contributions, indices. direction provided researchers deal hyperparameter tuning.

Language: Английский

Short-term electric load forecasting based on enhanced GA-BP multi-parameter algorithm DOI
Jixuan Wang,

Kegui Wu,

Yujing Wen

et al.

Measurement, Journal Year: 2025, Volume and Issue: unknown, P. 117855 - 117855

Published: May 1, 2025

Language: Английский

Citations

0
Probabilistic load forecasting based on quantile regression parallel CNN and BiGRU networks DOI
Yuting Lu, Gaocai Wang,

Xianfei Huang

et al.

Applied Intelligence, Journal Year: 2024, Volume and Issue: 54(15-16), P. 7439 - 7460

Published: June 7, 2024

Language: Английский

Citations

3
ELFNet: An Effective Electricity Load Forecasting Model Based on a Deep Convolutional Neural Network with a Double-Attention Mechanism DOI Creative Commons
Zhao Pei, Guang Ling, Xiang‐Xiang Song

et al.

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(14), P. 6270 - 6270

Published: July 18, 2024

Forecasting energy demand is critical to ensure the steady operation of power system. However, present approaches estimating load are still unsatisfactory in terms accuracy, precision, and efficiency. In this paper, we propose a novel method, named ELFNet, for short-term electricity consumption, based on deep convolutional neural network model with double-attention mechanism. The Gramian Angular Field method utilized convert electrical time series into 2D image data input proposed model. prediction accuracy greatly improved through use extract intrinsic characteristics from data, along channel attention spatial modules, enhance crucial features suppress irrelevant ones. ELFNet compared several classic learning networks across different horizons using publicly available real demands Belgian grid firm Elia. results show that suggested approach competitive effective forecasting.

Language: Английский

Citations

3
An IDBO-optimized CNN-BiLSTM model for load forecasting in regional integrated energy systems DOI

Zhonge Su,

Guoqiang Zheng, Guodong Wang

et al.

Computers & Electrical Engineering, Journal Year: 2024, Volume and Issue: 123, P. 110013 - 110013

Published: Dec. 21, 2024

Language: Английский

Citations

3
Identification and prediction of the degree of multidimensional returning to poverty risk for the household in China through the novel hybrid model: Based on the survey data of China Family Panel Studies (CFPS) DOI Creative Commons

Jinsong Zhang,

Tonggen Ding,

Linmao Ma

et al.

Heliyon, Journal Year: 2024, Volume and Issue: 10(21), P. e38783 - e38783

Published: Oct. 1, 2024

Language: Английский

Citations

2
A new method of network traffic prediction based on combination model DOI
Guohao Li, Zhongda Tian

Peer-to-Peer Networking and Applications, Journal Year: 2024, Volume and Issue: 17(3), P. 1075 - 1090

Published: Feb. 8, 2024

Language: Английский

Citations

1
The forecasting of surface displacement for tunnel slopes utilizing the WD-IPSO-GRU model DOI Creative Commons

Guoqing Ma,

Xiaopeng Zang,

Shitong Chen

et al.

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Sept. 5, 2024

Language: Английский

Citations

1
A novel gated dual convolutional neural network model with autoregressive method and attention mechanism for probabilistic load forecasting DOI
Yilei Qiu,

Shunzhen Wang,

Shuai Zhang

et al.

Applied Intelligence, Journal Year: 2023, Volume and Issue: 53(17), P. 20256 - 20271

Published: April 4, 2023

Language: Английский

Citations

3
A hybrid-driven remaining useful life prediction method combining asymmetric dual-channel autoencoder and nonlinear Wiener process DOI
Yuhang Duan, Zhen Liu, Honghui Li

et al.

Applied Intelligence, Journal Year: 2023, Volume and Issue: 53(21), P. 25490 - 25510

Published: Aug. 9, 2023

Language: Английский

Citations

2
The Forecasting of Surface Displacement for Tunnel Slopes Utilizing the WD-IPSO-GRU Model DOI

Guoqing MA,

Xiaopeng Zang,

Shitong Chen

et al.

Research Square (Research Square), Journal Year: 2024, Volume and Issue: unknown

Published: April 26, 2024

Abstract Continuous displacement prediction of tunnel slope deformation can serve as a basis for evaluating stability. For this purpose, fusion optimized model based on wavelet decomposition (WD), particle swarm optimization with genetic algorithm enhancement (IPSO), and gated recurrent unit (GRU) termed WD-IPSO-GRU is proposed. Initially, WD preprocesses noise features in field monitoring data; subsequently, IPSO dynamically sets learning factors weights, optimizing the number neurons iteration times GRU hidden layers L1 L2, introduces Dropout technique to prevent overfitting, enhancing performance long-term sequence tasks. Finally, leveraging optimal solution enables GNSS surfaces. Results indicate that compared GRU, neural network (RNN), long short-term memory (LSTM) models, demonstrates higher accuracy. The root mean square error (RMSE), absolute percentage (MAPE), coefficient determination (R²) site 02 are 0.16, 0.18%, 0.95 respectively, providing new approach prediction.

Language: Английский

Citations

0