Optimized deep neural network architectures for energy consumption and PV production forecasting

Accurate time-series forecasting of energy consumption and photovoltaic (PV) production is essential for effective energy management and sustainability. Deep Neural Networks (DNNs) are effective tools for learning complex patterns in such data; however, optimizing their architecture remains a signif...

Full description

Saved in:
Bibliographic Details
Main Authors: Eghbal Hosseini, Barzan Saeedpour, Mohsen Banaei, Razgar Ebrahimy
Format: Article
Language:English
Published: Elsevier 2025-05-01
Series:Energy Strategy Reviews
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2211467X25000677
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850226339472211968
author Eghbal Hosseini
Barzan Saeedpour
Mohsen Banaei
Razgar Ebrahimy
author_facet Eghbal Hosseini
Barzan Saeedpour
Mohsen Banaei
Razgar Ebrahimy
author_sort Eghbal Hosseini
collection DOAJ
description Accurate time-series forecasting of energy consumption and photovoltaic (PV) production is essential for effective energy management and sustainability. Deep Neural Networks (DNNs) are effective tools for learning complex patterns in such data; however, optimizing their architecture remains a significant challenge. This paper introduces a novel hybrid optimization approach that integrates Genetic Algorithms (GA) and Particle Swarm Optimization (PSO) to enhance the DNN architecture for more accurate energy forecasting. The performance of GA-PSO is compared with leading hyperparameter optimization techniques, such as Bayesian Optimization and Evolutionary Strategy, across various optimization benchmarks and DNN hyperparameter tuning tasks. The study evaluates the GA-PSO-enhanced Optimized Deep Neural Network (ODNN) against traditional DNNs and state-of-the-art machine learning methods on multiple real-world energy forecasting tasks. The results demonstrate that ODNN outperforms the average performance of other methods, achieving a 27% improvement in forecasting accuracy and a 22% reduction in error across various metrics. These findings demonstrate the significant potential of GA-PSO as an effective tool to optimize DNN models in energy forecasting applications.
format Article
id doaj-art-019e191f4b5547eea63e7fa29a8d3d9d
institution OA Journals
issn 2211-467X
language English
publishDate 2025-05-01
publisher Elsevier
record_format Article
series Energy Strategy Reviews
spelling doaj-art-019e191f4b5547eea63e7fa29a8d3d9d2025-08-20T02:05:07ZengElsevierEnergy Strategy Reviews2211-467X2025-05-015910170410.1016/j.esr.2025.101704Optimized deep neural network architectures for energy consumption and PV production forecastingEghbal Hosseini0Barzan Saeedpour1Mohsen Banaei2Razgar Ebrahimy3Technical University of Denmark, Department of Applied Mathematics and Computer Science, Copenhagen, Denmark; Corresponding author.Department of Computer Engineering, University of Kurdistan, IranTechnical University of Denmark, Department of Applied Mathematics and Computer Science, Copenhagen, DenmarkTechnical University of Denmark, Department of Applied Mathematics and Computer Science, Copenhagen, DenmarkAccurate time-series forecasting of energy consumption and photovoltaic (PV) production is essential for effective energy management and sustainability. Deep Neural Networks (DNNs) are effective tools for learning complex patterns in such data; however, optimizing their architecture remains a significant challenge. This paper introduces a novel hybrid optimization approach that integrates Genetic Algorithms (GA) and Particle Swarm Optimization (PSO) to enhance the DNN architecture for more accurate energy forecasting. The performance of GA-PSO is compared with leading hyperparameter optimization techniques, such as Bayesian Optimization and Evolutionary Strategy, across various optimization benchmarks and DNN hyperparameter tuning tasks. The study evaluates the GA-PSO-enhanced Optimized Deep Neural Network (ODNN) against traditional DNNs and state-of-the-art machine learning methods on multiple real-world energy forecasting tasks. The results demonstrate that ODNN outperforms the average performance of other methods, achieving a 27% improvement in forecasting accuracy and a 22% reduction in error across various metrics. These findings demonstrate the significant potential of GA-PSO as an effective tool to optimize DNN models in energy forecasting applications.http://www.sciencedirect.com/science/article/pii/S2211467X25000677Photovoltaic productionDeep neural networksMeta-heuristic algorithmsTime series forecasting
spellingShingle Eghbal Hosseini
Barzan Saeedpour
Mohsen Banaei
Razgar Ebrahimy
Optimized deep neural network architectures for energy consumption and PV production forecasting
Energy Strategy Reviews
Photovoltaic production
Deep neural networks
Meta-heuristic algorithms
Time series forecasting
title Optimized deep neural network architectures for energy consumption and PV production forecasting
title_full Optimized deep neural network architectures for energy consumption and PV production forecasting
title_fullStr Optimized deep neural network architectures for energy consumption and PV production forecasting
title_full_unstemmed Optimized deep neural network architectures for energy consumption and PV production forecasting
title_short Optimized deep neural network architectures for energy consumption and PV production forecasting
title_sort optimized deep neural network architectures for energy consumption and pv production forecasting
topic Photovoltaic production
Deep neural networks
Meta-heuristic algorithms
Time series forecasting
url http://www.sciencedirect.com/science/article/pii/S2211467X25000677
work_keys_str_mv AT eghbalhosseini optimizeddeepneuralnetworkarchitecturesforenergyconsumptionandpvproductionforecasting
AT barzansaeedpour optimizeddeepneuralnetworkarchitecturesforenergyconsumptionandpvproductionforecasting
AT mohsenbanaei optimizeddeepneuralnetworkarchitecturesforenergyconsumptionandpvproductionforecasting
AT razgarebrahimy optimizeddeepneuralnetworkarchitecturesforenergyconsumptionandpvproductionforecasting