Energy and AI最新文献_第4页

Opening the AI black-box: Symbolic regression with Kolmogorov–Arnold Networks for advanced energy applications 打开人工智能黑盒子：先进能源应用的Kolmogorov-Arnold网络符号回归

IF 9.6

Energy and AI Pub Date : 2025-08-23 DOI: 10.1016/j.egyai.2025.100595

Nataly R. Panczyk, Omer F. Erdem, Majdi I. Radaideh

{"title":"Opening the AI black-box: Symbolic regression with Kolmogorov–Arnold Networks for advanced energy applications","authors":"Nataly R. Panczyk, Omer F. Erdem, Majdi I. Radaideh","doi":"10.1016/j.egyai.2025.100595","DOIUrl":"10.1016/j.egyai.2025.100595","url":null,"abstract":"<div><div>While most modern machine learning methods offer speed and accuracy, few promise interpretability or explainability—two key features necessary for highly sensitive industries, like medicine, finance, and engineering. Using eight datasets representative of one especially sensitive industry, nuclear power, this work compares a traditional feedforward neural network (FNN) to a Kolmogorov–Arnold Network (KAN). We consider not only model performance and accuracy, but also interpretability through model architecture and explainability through a post-hoc SHapley Additive exPlanations (SHAP) analysis, a game-theory-based feature importance method. In terms of accuracy, we find KANs and FNNs comparable across all datasets when output dimensionality is limited. KANs, which transform into symbolic equations after training, yield perfectly interpretable models, while FNNs remain black-boxes. Finally, using the post-hoc explainability results from Kernel SHAP, we find that KANs learn real, physical relations from experimental data, while FNNs simply produce statistically accurate results. Overall, this analysis finds KANs a promising alternative to traditional machine learning methods, particularly in applications requiring both accuracy <em>and</em> comprehensibility.</div></div>","PeriodicalId":34138,"journal":{"name":"Energy and AI","volume":"22 ","pages":"Article 100595"},"PeriodicalIF":9.6,"publicationDate":"2025-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145045870","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Semi-supervised battery state of health estimation for field applications 现场应用的半监督电池健康状态估计

IF 9.6

Energy and AI Pub Date : 2025-08-21 DOI: 10.1016/j.egyai.2025.100575

Nejira Hadzalic , Jacob Hamar , Marco Fischer , Simon Erhard , Jan Philipp Schmidt

{"title":"Semi-supervised battery state of health estimation for field applications","authors":"Nejira Hadzalic , Jacob Hamar , Marco Fischer , Simon Erhard , Jan Philipp Schmidt","doi":"10.1016/j.egyai.2025.100575","DOIUrl":"10.1016/j.egyai.2025.100575","url":null,"abstract":"<div><div>Battery electric vehicles are exposed to highly diverse operating conditions and driving behaviors that strongly influence degradation pathways, yet these real-world complexities are only partially captured in laboratory aging tests. This study investigates a semi-supervised learning approach for robust estimation of battery state of health, defined as the ratio of remaining to nominal capacity. The method integrates a multi-view co-training algorithm with a rule-based pseudo labeling mechanism and is developed and validated using field data from 3000 BMW i3 vehicles with battery capacity of 60Ah, collected since 2013 across 34 countries. The available data comprises standardized full charge capacity measurements, which serve as ground truth labels. The proposed training and validation pipeline is designed to address challenges inherent in real-world data generation and is particularly advantageous during early deployment of new battery technologies, when labeled data is scarce. By incrementally incorporating newly available labeled data into both evaluation and retraining, the model adapts to heterogeneous aging patterns observed in the field. Comparative analysis demonstrates that, relative to a supervised benchmark, the proposed method reduces estimation error by 28% under limited-label conditions and by 6% under optimally labeled scenarios, highlighting its robustness for field applications.</div></div>","PeriodicalId":34138,"journal":{"name":"Energy and AI","volume":"22 ","pages":"Article 100575"},"PeriodicalIF":9.6,"publicationDate":"2025-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144908147","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Physics-embedded graph learning unlocks integrated energy system modeling 物理嵌入式图学习解锁集成能源系统建模

IF 9.6

Energy and AI Pub Date : 2025-08-20 DOI: 10.1016/j.egyai.2025.100597

Chongshuo Yuan , Xiaojie Lin , Wei Zhong

引用次数: 0

Meta-heuristic federated learning aggregation methods for load forecasting 负荷预测的元启发式联合学习聚合方法

IF 9.6

Energy and AI Pub Date : 2025-08-20 DOI: 10.1016/j.egyai.2025.100594

Efstathios Sarantinopoulos , Vasilis Michalakopoulos , Elissaios Sarmas , Vangelis Marinakis , Liana Toderean , Tudor Cioara

{"title":"Meta-heuristic federated learning aggregation methods for load forecasting","authors":"Efstathios Sarantinopoulos , Vasilis Michalakopoulos , Elissaios Sarmas , Vangelis Marinakis , Liana Toderean , Tudor Cioara","doi":"10.1016/j.egyai.2025.100594","DOIUrl":"10.1016/j.egyai.2025.100594","url":null,"abstract":"<div><div>Federated learning (FL) is essential to energy transition as it leverages decentralized energy data and machine learning to collaborative train global energy predictive models across distributed energy resources while preserving data privacy. This paper introduces one of the first FL frameworks that efficiently integrates swarm intelligence-based aggregation methods to large-scale energy consumption forecasting, by extending the TensorFlow Federated Core framework with specialized functional enhancements. The primary objective is to enhance forecasting accuracy in decentralized learning settings. We investigated the effectiveness of various nature-inspired metaheuristics for optimizing the aggregation of local model updates from distributed energy resource nodes into a global model for load forecasting tasks, including Grey Wolf Optimization (GWO), Particle Swarm Optimization (PSO), and Firefly Algorithm (FFA) against the standard Federated Averaging (FedAvg) algorithm. Using a real-world dataset comprising of 4,438 distinct energy consumers, we demonstrate that metaheuristic aggregators consistently outperform the most well-known method, Federated Averaging in predictive accuracy. Among these approaches, GWO emerges as the superior performer achieving up to 23.6% error reduction. Our findings underscore the significant potential of metaheuristic-based aggregation mechanisms in improving FL outcomes, particularly in energy forecasting applications involving large-scale distributed data scenarios.</div></div>","PeriodicalId":34138,"journal":{"name":"Energy and AI","volume":"22 ","pages":"Article 100594"},"PeriodicalIF":9.6,"publicationDate":"2025-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145007633","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Abnormal state detection of diesel engine with flexible monitoring thresholds 灵活监测阈值的柴油机异常状态检测

IF 9.6

Energy and AI Pub Date : 2025-08-13 DOI: 10.1016/j.egyai.2025.100589

Xinwei Wang , Xiaolong Zhu , Guobin Pei , Kunyu Cai , Jiewei Lin

{"title":"Abnormal state detection of diesel engine with flexible monitoring thresholds","authors":"Xinwei Wang , Xiaolong Zhu , Guobin Pei , Kunyu Cai , Jiewei Lin","doi":"10.1016/j.egyai.2025.100589","DOIUrl":"10.1016/j.egyai.2025.100589","url":null,"abstract":"<div><div>In this paper, a condition monitoring method based on enhanced hierarchical fuzzy entropy and support vector data description (SVDD) is proposed to detect the abnormal state of diesel engine. Aiming at the lack of characterization of traditional statistical parameters, entropy algorithm is introduced to extract the fault mode information rich in vibration signal. Combined with the enhanced hierarchical analysis process, the enhanced hierarchical fuzzy entropy algorithm is proposed to enhance the extraction of high and low frequency fault related information in vibration signals. SVDD is used to construct anomaly detection model and set the threshold automatically. And the parrot optimization algorithm is used to optimize the parameters of the support vector data model to improve the adaptability and discrimination of the model. In order to verify the effectiveness of the proposed method, an experiment of a six-cylinder diesel engine under multiple working conditions was carried out to obtain the diesel cylinder head vibration data set under multiple conditions. Through feature analysis and discriminant analysis, the results show that compared with the existing entropy algorithm and intelligent search algorithm, the proposed method shows better representation and discriminant ability, and the average precision and recall rate under multiple loads are 99.32 % and 92 %, respectively.</div></div>","PeriodicalId":34138,"journal":{"name":"Energy and AI","volume":"21 ","pages":"Article 100589"},"PeriodicalIF":9.6,"publicationDate":"2025-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144865412","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A flow-based model for conditional and probabilistic electricity consumption profile generation 条件和概率用电量分布生成的基于流的模型

IF 9.6

Energy and AI Pub Date : 2025-08-11 DOI: 10.1016/j.egyai.2025.100586

Weijie Xia , Chenguang Wang , Peter Palensky , Pedro P. Vergara

引用次数: 0

A quantum neural network model for short term wind speed forecasting using weather data 利用天气资料预测短期风速的量子神经网络模型

IF 9.6

Energy and AI Pub Date : 2025-08-09 DOI: 10.1016/j.egyai.2025.100588

Otto Menegasso Pires , Erick Giovani Sperandio Nascimento , Marcelo A. Moret

{"title":"A quantum neural network model for short term wind speed forecasting using weather data","authors":"Otto Menegasso Pires , Erick Giovani Sperandio Nascimento , Marcelo A. Moret","doi":"10.1016/j.egyai.2025.100588","DOIUrl":"10.1016/j.egyai.2025.100588","url":null,"abstract":"<div><div>The use of computational intelligence has become commonplace for accurate wind speed and energy forecasting, however the energy-intensive processes involved in training and tuning stands as a critical issue for the sustainability of AI models. Quantum computing emerges as a key player in addressing this concern, offering a quantum advantage that could potentially accelerate computations or, more significantly, reduce energy consumption. It is a matter of debate if purely quantum machine learning models, as they currently stand, are capable of competing with the classical state of the art on relevant problems. We investigate the efficacy of quantum neural networks (QNNs) for wind speed nowcasting, comparing them to a baseline Multilayer Perceptron (MLP). Utilizing meteorological data from Bahia, Brazil, we develop a QNN tailored for up to six hours ahead wind speed prediction. Our analysis reveals that the QNN demonstrates competitive performance compared to MLP. We evaluate models using RMSE, Pearson’s R, and Factor of 2 metrics, emphasizing QNNs’ promising generalization capabilities and robustness across various wind prediction scenarios. This study is a seminal work on the potential of QNNs in advancing renewable energy forecasting, advocating for further exploration of quantum machine learning in sustainable energy research.</div></div>","PeriodicalId":34138,"journal":{"name":"Energy and AI","volume":"21 ","pages":"Article 100588"},"PeriodicalIF":9.6,"publicationDate":"2025-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144841868","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Hybrid deep learning architecture for temperature gradient control of a solid oxide electrolysis cell under fluctuating wind power 波动风力下固体氧化物电解池温度梯度控制的混合深度学习架构

IF 9.6

Energy and AI Pub Date : 2025-08-09 DOI: 10.1016/j.egyai.2025.100592

Jiayu Zhu , Yun Zheng , Wenlai Zhao , Wei Yan , Jiujun Zhang

{"title":"Hybrid deep learning architecture for temperature gradient control of a solid oxide electrolysis cell under fluctuating wind power","authors":"Jiayu Zhu , Yun Zheng , Wenlai Zhao , Wei Yan , Jiujun Zhang","doi":"10.1016/j.egyai.2025.100592","DOIUrl":"10.1016/j.egyai.2025.100592","url":null,"abstract":"<div><div>The co-electrolysis of CO₂ and H₂O through solid oxide electrolysis cells (SOECs), powered by renewable energy sources, offers a promising pathway to achieving carbon neutrality in the chemical industry. However, the inherent intermittency of renewable energy generation, such as wind power, leads to unstable power input for electrolysis. This variability induces significant thermal stress in SOECs, potentially causing cracks or even system failure. To address this challenge, a hybrid deep learning architecture (HDLA) was developed to control the temperature gradient of SOECs. The architecture combines a convolutional neural network (CNN) and a long short-term memory (LSTM) model for wind power prediction, a multi-physics model for temperature gradient simulation, and a linear neural network regression model to simulate the temperature distribution in SOECs. Training and verification are conducted using 16 datasets from an industrial wind farm. The results demonstrate that the application of HDLA successfully reduce the temperature gradient of SOECs from ±20°C to ±5°C. Additionally, the potential wind power utilization achieved near-complete wind power utilization, increasing from 18 % to 99 %. This real-time control strategy, which optimizes flow regulation, effectively mitigates thermal stress, thereby extending the lifespan of SOECs and ensuring continuous carbon reduction, efficient conversion, and utilization.</div></div>","PeriodicalId":34138,"journal":{"name":"Energy and AI","volume":"21 ","pages":"Article 100592"},"PeriodicalIF":9.6,"publicationDate":"2025-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144858078","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Identifying latent workforce capacities for extreme heat resilience: An artificial intelligence assisted approach 识别潜在的劳动力能力极端热弹性：人工智能辅助方法

IF 9.6

Energy and AI Pub Date : 2025-08-05 DOI: 10.1016/j.egyai.2025.100580

Jieshu Wang , Patricia Solís

{"title":"Identifying latent workforce capacities for extreme heat resilience: An artificial intelligence assisted approach","authors":"Jieshu Wang , Patricia Solís","doi":"10.1016/j.egyai.2025.100580","DOIUrl":"10.1016/j.egyai.2025.100580","url":null,"abstract":"<div><div>Extreme heat events, intensified by climate change, pose critical challenges to public health, infrastructure, and workforce resilience. Despite the urgency of these challenges, there is no systematic framework to identify workforce adaptive capacities that can help build regional heat resilience. This study introduces a novel large language model assisted approach, using task-level data from the O*NET dataset, to identify workforce capacities that enhance heat resilience. By defining heat-solution tasks as activities mitigating heat impacts, protecting public health, or improving infrastructure, we classify heat-solution occupations and dual-impact occupations, which are both vulnerable to heat and critical to heat resilience. A case study of the state of Arizona in the United States analyzed 16,398 tasks across 663 occupations, identifying 110 heat-solution occupations (about 14 % of Arizona’ workforce) and 31 dual-impact occupations. The study reveals how energy-relevant occupations, such as HVAC technicians, solar installers, and retrofit specialists, contribute to climate adaptation, linking occupational roles to the clean energy transition and resilient infrastructure. By leveraging large language models, our method provides a scalable, AI-powered tool to analyze workforce data and identify capacities necessary for energy efficiency and hazard resilience. The findings not only demonstrate the potential of large language models in workforce analysis but also contributed to shaping Arizona’s first Extreme Heat Preparedness Plan. This study offers a scalable method to uncover latent capacities and informs policies on workforce development, safety regulations, and climate-resilient infrastructure, serving as a model for other regions facing similar challenges.</div></div>","PeriodicalId":34138,"journal":{"name":"Energy and AI","volume":"21 ","pages":"Article 100580"},"PeriodicalIF":9.6,"publicationDate":"2025-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144889504","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An FPGA-accelerated multi-level AI-integrated simulation framework for multi-time domain power systems with high penetration of power converters 基于fpga加速的高功率变换器多时域系统多级ai集成仿真框架

IF 9.6

Energy and AI Pub Date : 2025-08-05 DOI: 10.1016/j.egyai.2025.100574

Chen Liu , Peng Su , Hao Bai , Xizheng Guo , Alber Filbà Martínez , Jose Luis Dominguez Garcia

{"title":"An FPGA-accelerated multi-level AI-integrated simulation framework for multi-time domain power systems with high penetration of power converters","authors":"Chen Liu , Peng Su , Hao Bai , Xizheng Guo , Alber Filbà Martínez , Jose Luis Dominguez Garcia","doi":"10.1016/j.egyai.2025.100574","DOIUrl":"10.1016/j.egyai.2025.100574","url":null,"abstract":"<div><div>The increasing integration of renewable energy sources and power electronic devices has significantly increased the complexity of modern power systems, making modeling and simulation challenging due to multi-time scale dynamics and multi-physics coupling. To address these challenges, this paper proposes a multi-level simulation framework based on unified energy flow theory. The framework structures systems hierarchically using energy transmission functions and unified energy information flow-based surrogate models with defined ports, ensuring compatibility with artificial intelligence algorithms. By integrating AI techniques, such as back propagation neural networks, the framework predicts variables with high computational complexity, improving accuracy and simulation efficiency. A multi-level simulation architecture leveraging Field Programmable Gate Arrays (FPGAs) enables faster-than-real-time system-level simulation and real-time component-level modeling with time resolution as small as 5 nanoseconds. A DC microgrid case study with photovoltaic generation, battery storage, and power electronic converters demonstrates the proposed method, achieving up to a 500× speedup over traditional Simulink models while maintaining high accuracy. The results confirm the framework’s ability to capture multiphysics interactions, optimize energy distribution, and ensure system stability under dynamic conditions, providing an efficient and scalable solution for advanced DC microgrid simulations.</div></div>","PeriodicalId":34138,"journal":{"name":"Energy and AI","volume":"21 ","pages":"Article 100574"},"PeriodicalIF":9.6,"publicationDate":"2025-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144829482","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0