Sustainable Energy Grids & Networks最新文献_第3页

Bridging detection and forecasting: Refined WaveNet for robust smart building energy monitoring 桥接检测和预测：用于稳健智能建筑能源监测的改进WaveNet

IF 5.6 2区工程技术

Sustainable Energy Grids & Networks Pub Date : 2026-03-01 Epub Date: 2026-02-02 DOI: 10.1016/j.segan.2026.102142

Cyrine Berrima, Viet Tra, Manar Amayri

{"title":"Bridging detection and forecasting: Refined WaveNet for robust smart building energy monitoring","authors":"Cyrine Berrima, Viet Tra, Manar Amayri","doi":"10.1016/j.segan.2026.102142","DOIUrl":"10.1016/j.segan.2026.102142","url":null,"abstract":"<div><div>Clean, high-resolution time series data is essential for precise predictions and effective control in smart building energy systems. Based on a refined WaveNet, this study proposes a novel architecture for unsupervised anomaly identification in univariate energy consumption time series. Without labeled data, the model is specifically intended to detect abnormal patterns and capture long-range temporal correlations. It uses gated activations, skip connections, and dilated causal convolutions to improve temporal fidelity, resilience, and sensitivity. To the best of our knowledge, this is the first anomaly detection strategy in this setting that integrates architectural improvements with a thorough assessment of the impact on downstream forecasting. High-frequency energy consumption time series are used to rigorously benchmark the Refined WaveNet against cutting-edge baselines, such as the original WaveNet, deep generative hierarchical learning (DGHL), and variational autoencoder (VAE). With a Precision-Recall Area Under the Curve of 99.23% and an F1 Score of 98.30%, the model outperforms the original WaveNet—which achieves only 41.96% in F1 Score—by more than 56 percentage points. We combine the detection module with a long short-term memory (LSTM) forecasting model in order to evaluate its practical usefulness. According to experimental findings, adding just 10% synthetic anomalies to the time series increases mean squared error (MSE) by more than thirty times, whereas using Refined WaveNet for preprocessing brings forecasting performance back to levels that are almost pristine. These results highlight anomaly detection’s significance as a fundamental element of time series analysis for intelligent energy systems and establish Refined WaveNet as a small, effective, and deployable solution for edge-based, real-time applications. The code is available at: <span><span>https://surl.li/oumfna</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102142"},"PeriodicalIF":5.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147395629","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Raw measurement supervised learning transformer for anomaly detection of power system digital twin updates 用于电力系统数字孪生更新异常检测的原始测量监督学习变压器

IF 5.6 2区工程技术

Sustainable Energy Grids & Networks Pub Date : 2026-03-01 Epub Date: 2025-11-28 DOI: 10.1016/j.segan.2025.102069

Zhiwei Shen, Felipe Arraño-Vargas, Georgios Konstantinou

{"title":"Raw measurement supervised learning transformer for anomaly detection of power system digital twin updates","authors":"Zhiwei Shen, Felipe Arraño-Vargas, Georgios Konstantinou","doi":"10.1016/j.segan.2025.102069","DOIUrl":"10.1016/j.segan.2025.102069","url":null,"abstract":"<div><div>Continuous updates are essential to ensure that a digital twin (DT) remains an accurate representation of its physical counterpart. The performance of DT applications heavily relies on how accurately the DT reflects its physical counterpart. DT updates, however, can be compromised by anomalous PT data stemming from physical twin (PT) measurements, communication malfunctions, and/or external attacks. Detecting such anomalies in PT data is crucial to ensuring the accuracy and reliability of DT, thereby generating only valid outcomes for associated applications. This paper proposes a detection method to identify anomalous PT data before its integration into the DT. The proposed raw measurement supervised learning Transformer (RM-SL-TF) facilitates a straightforward identification of PT data using raw measurements, eliminating the dependency on data preprocessing. The feasibility and effectiveness of the RM-SL-TF are demonstrated by using a power system digital twin (PSDT) that requires frequent updates. The resulting detection accuracy of anomalous PT data is comparable to, or even surpasses, that of other artificial intelligence (AI) algorithms that rely on input feature normalisation. By directly analysing raw measurements without normalising input features, the proposed approach is simpler, more flexible, and expandable, making it suitable for establishing and advancing the development and implementation of DTs for power systems and other industries.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102069"},"PeriodicalIF":5.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738369","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Coordinated scheduling mechanism of electric vehicle V2G and DR in integrated energy systems via deep reinforcement learning 基于深度强化学习的综合能源系统中电动汽车V2G和DR协调调度机制

IF 5.6 2区工程技术

Sustainable Energy Grids & Networks Pub Date : 2026-03-01 Epub Date: 2025-12-10 DOI: 10.1016/j.segan.2025.102086

Chao He , Junwen Peng , Wenhui Jiang , Jiacheng Wang , Sirui Zhang , Yi Zhang , Hong Na

{"title":"Coordinated scheduling mechanism of electric vehicle V2G and DR in integrated energy systems via deep reinforcement learning","authors":"Chao He , Junwen Peng , Wenhui Jiang , Jiacheng Wang , Sirui Zhang , Yi Zhang , Hong Na","doi":"10.1016/j.segan.2025.102086","DOIUrl":"10.1016/j.segan.2025.102086","url":null,"abstract":"<div><div>With the large-scale integration of electric vehicles (EVs) and the growing penetration of renewable energy, integrated energy systems (IES) are facing increased complexity in coordinated scheduling. This complexity arises from multi-source heterogeneity, heightened operational uncertainty, and the challenge of coordinating demand-side responses. To address these issues, we propose a coordinated optimization framework that integrates vehicle-to-grid (V2G) technology, demand response (DR) mechanisms, and carbon trading incentives. The framework facilitates dynamic coordination of flexible resources, such as EV charging/discharging, energy storage, grid electricity procurement, and heat pump loads. This improves operational flexibility, economic efficiency, and carbon reduction potential. To solve the multi-objective, non-convex optimization problem, we introduce a Deep Q-Network (DQN) algorithm from deep reinforcement learning. By utilizing policy learning, the algorithm dynamically optimizes operational decisions across various energy units, enabling adaptive scheduling in response to real-time system changes. Simulation results show that the proposed framework outperforms traditional rule-based and static strategies in terms of load regulation, carbon emission control, and operational cost. These findings highlight the broad applicability and scalability of the integrated scheduling mechanism with reinforcement learning for low-carbon dispatch in IES.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102086"},"PeriodicalIF":5.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738308","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Neural Lyapunov function with projection gradient descent for region of attraction estimation in AC/DC wind power systems 带投影梯度下降的神经Lyapunov函数用于交直流风力发电系统的吸引力区域估计

IF 5.6 2区工程技术

Sustainable Energy Grids & Networks Pub Date : 2026-03-01 Epub Date: 2026-02-02 DOI: 10.1016/j.segan.2026.102123

Zhaobin Du, Sicheng Shan, Yao Liu, Wenxian Zhao

{"title":"Neural Lyapunov function with projection gradient descent for region of attraction estimation in AC/DC wind power systems","authors":"Zhaobin Du, Sicheng Shan, Yao Liu, Wenxian Zhao","doi":"10.1016/j.segan.2026.102123","DOIUrl":"10.1016/j.segan.2026.102123","url":null,"abstract":"<div><div>Estimating the region of attraction (ROA) of an equilibrium point remains a fundamental yet challenging problem in power system transient stability analysis. This paper proposes a neural Lyapunov function-based framework for transient stability assessment and ROA estimation in nonlinear AC/DC power systems with wind power integration. A structurally positive-definite neural architecture is designed to simplify Lyapunov function learning and improve training robustness. To further enhance efficiency, a projection gradient descent (PGD)-based counterexample discovery strategy is introduced, which formulates the Lyapunov violation search as a constrained optimization problem within prescribed state boundaries. The proposed approach is validated on a standard 3-machine 9-bus system and a 9-bus AC/DC system with wind farms. Comparative results demonstrate that the method reduces training time by approximately 55% and enlarges the verified ROA by over 15% compared with existing neural Lyapunov benchmarks, while maintaining accurate post-fault stability assessment.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102123"},"PeriodicalIF":5.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147395734","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Delta-connected series-cascaded microgrids with extended power routing range control 具有扩展功率路由范围控制的三角洲连接串联级联微电网

IF 5.6 2区工程技术

Sustainable Energy Grids & Networks Pub Date : 2026-03-01 Epub Date: 2026-01-28 DOI: 10.1016/j.segan.2026.102128

Salman Ali , Santiago Bogarra Rodríguez , Muhammad Mansoor Khan , Felipe Córcoles

{"title":"Delta-connected series-cascaded microgrids with extended power routing range control","authors":"Salman Ali , Santiago Bogarra Rodríguez , Muhammad Mansoor Khan , Felipe Córcoles","doi":"10.1016/j.segan.2026.102128","DOIUrl":"10.1016/j.segan.2026.102128","url":null,"abstract":"<div><div>Series-cascaded microgrids (SCMGs) enable high-voltage synthesis and enhances control flexibility; however, power distribution among distributed generation (DG) sources in such systems remains insufficiently examined. The inherent intermittency of renewable energy sources introduces challenges in inter-module and inter-cluster power sharing, which can limit the injection of balanced three-phase currents into the grid. This study focuses on a delta-connected SCMG configuration incorporating both PV and battery-based DG units operating in grid-connected mode. Battery modules play a key role in voltage support, power balancing, and maintaining the reference capacitor voltage for PV-integrated units. A control methodology is developed, employing third harmonic current injection for inter-module power routing and zero-sequence current injection for inter-cluster power routing under adverse power flow conditions caused by PV variability. The optimal injection coefficients for both power sharing carriers are analytically derived, demonstrating an increase in the power routing range with a power routing factor of 100 % for the developed third harmonic current injection. The proposed method is validated through detailed MATLAB/Simulink simulations and Typhoon HIL experiments across balanced, partially imbalanced, PV-inactive, and reverse power-flow conditions. The results confirm stable seven-level output voltage, balanced grid currents, and robust inter-module and inter-cluster routing even under extreme conditions (χ<sub><em>bc</em></sub> ≈ 0), highlighting the superior balancing capability of the delta-connected topology combined with the proposed control strategy.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102128"},"PeriodicalIF":5.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147395730","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

P2P modeling formation coalitions and prosumers participation based on dynamic pricing algorithm and line congestion consideration 基于动态定价算法和考虑线路拥塞的P2P建模形成联盟和产消参与

IF 5.6 2区工程技术

Sustainable Energy Grids & Networks Pub Date : 2026-03-01 Epub Date: 2025-12-08 DOI: 10.1016/j.segan.2025.102099

Zhen Ji, Wei Sun, Bo Yan, BoHao Sun

{"title":"P2P modeling formation coalitions and prosumers participation based on dynamic pricing algorithm and line congestion consideration","authors":"Zhen Ji, Wei Sun, Bo Yan, BoHao Sun","doi":"10.1016/j.segan.2025.102099","DOIUrl":"10.1016/j.segan.2025.102099","url":null,"abstract":"<div><div>The rapid proliferation of distributed energy resources such as photovoltaic systems, wind turbines, battery energy storage systems, and electric vehicles has transformed residential microgrids into active, transactive energy communities. However, realizing fair, efficient, and scalable peer-to-peer energy sharing under stochastic household demand, dynamic pricing, and network constraints remains a major challenge. This study develops a hybrid centralized-decentralized peer-to-peer energy-sharing framework that models heterogeneous household prosumers five distinct types equipped with photovoltaic, wind turbine, battery energy storage, and electric vehicles within a demand-supply environment. The model integrates a home energy management system with dynamic pricing derived from the balance between Feed-in Tariff and Real-Time Pricing, augmented by congestion and degradation costs to ensure market fairness. A heuristic battery control algorithm and a two-level robust optimization based on the MILP and column-and-constraint generation method are implemented to coordinate energy exchanges between prosumers and the grid. Electric vehicles are treated as active market agents capable of bidirectional energy trading to enhance grid flexibility. Case studies involving 30, 120, and 240 households simulated using MATLAB to compare three operational scenarios without P2P trading, hybrid centralized-decentralized peer to peer trading, and large-scale community participation. The findings indicate that the proposed framework increases household self-consumption rates by 64.22 %, decreases grid energy imports by 52.5 %, and elevates prosumer revenue by 41.6 %, while preserving network stability and fairness. Hybrid market structure efficiently reduces peak energy costs, ensures strong local balance, and offers scalable basis for resilient, consumer-driven energy communities.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102099"},"PeriodicalIF":5.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145926051","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A hybrid model for efficient reliability assessment of power systems 电力系统高效可靠性评估的混合模型

IF 5.6 2区工程技术

Sustainable Energy Grids & Networks Pub Date : 2026-03-01 Epub Date: 2025-12-08 DOI: 10.1016/j.segan.2025.102091

Adil Waheed, Jueyou Li

{"title":"A hybrid model for efficient reliability assessment of power systems","authors":"Adil Waheed, Jueyou Li","doi":"10.1016/j.segan.2025.102091","DOIUrl":"10.1016/j.segan.2025.102091","url":null,"abstract":"<div><div>The reliability assessment of power systems ensures uninterrupted service and system stability. This paper proposes a hybrid approach consisting of Multi-Layer Perceptron (MLP) and Long Short-Term Memory (LSTM) networks to predict key reliability indices, such as Loss of Load Probability (LOLP), Expected Energy Not Supplied (EENS), and Loss of Load Frequency (LOLF). The proposed approach eliminates the need to solve multiple Optimal Power Flow (OPF) problems for each system state, thereby reducing computational time and complexity. In the training phase, the model learns from historical data and a limited set of pre-calculated OPF results. This process enables the model to capture the complex relationships between system states, load curtailment, and reliability indices. Once the training phase is complete, the model directly predicts reliability indices without the need to repeatedly solve OPF for every system state. Comparative analysis demonstrates that the proposed method achieves a high level of accuracy while significantly outperforming conventional techniques, such as Monte Carlo Simulation (MCS). The proposed model is also applied to well-known power systems, including the IEEE Reliability Test Systems (IEEE RTS, IEEE RTS-96) and the Saskatchewan Power Corporation (SPC) system in Canada. The results show that the MLP-LSTM model performs better and can solve OPF-based reliability assessments. Furthermore, the model reduces dependence on OPF and provides faster and more reliable analysis in real-time. This improvement facilitates better decision-making in power system planning and operations.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102091"},"PeriodicalIF":5.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738359","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Robust optimization of electric bus charging-operation scheduling considering charging discrepancy 考虑充电差异的电动客车充电调度鲁棒优化

IF 5.6 2区工程技术

Sustainable Energy Grids & Networks Pub Date : 2026-03-01 Epub Date: 2025-12-06 DOI: 10.1016/j.segan.2025.102084

Zhouzuo Wang , Xinghua Hu , Jiahao Zhao , Fang Liu , Lanping Si

{"title":"Robust optimization of electric bus charging-operation scheduling considering charging discrepancy","authors":"Zhouzuo Wang , Xinghua Hu , Jiahao Zhao , Fang Liu , Lanping Si","doi":"10.1016/j.segan.2025.102084","DOIUrl":"10.1016/j.segan.2025.102084","url":null,"abstract":"<div><div>Optimizing electric bus (EB) scheduling is crucial for advancing urban bus systems and reducing carbon emissions. In this study, we establish an EB scheduling model using a robust optimization paradigm to address the challenges associated with charging demand uncertainty during the operation period. To model the charging process of electric buses (EBs), we adopted a piecewise linear function to handle the nonlinear charging function. This approach improves the practicality of the model while ensuring basic realism. This study introduced a mixed-integer programming model to maximize the profit of the EB system, including the weighted delay time. The main constraints include the departure time window and the charging process. To account for the impact of multiple vehicle types on the scheduling of EBs, a distributed robust optimization model is established for the uncertainty of the EB operation. An instantiated analysis is conducted to schedule an EB line in a Chinese city. The results demonstrate that the distributed robust optimization model enhances the expected profit by approximately 27.27 %-54.24 % compared with the deterministic model. Additionally, the robust optimization model exhibits a steeper increase in expected profit as the uncertainty level increases. Furthermore, the mixed scheduling strategies with multiple vehicle types in the robust optimization model enhance the profit compared to the model relying solely on a single vehicle type. The results demonstrate the applicability and effectiveness of the proposed model for EB scheduling.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102084"},"PeriodicalIF":5.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738360","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Integrated optimization and game theory framework for fair cost allocation in community microgrids 社区微电网成本公平分配的集成优化与博弈论框架

IF 5.6 2区工程技术

Sustainable Energy Grids & Networks Pub Date : 2026-03-01 Epub Date: 2025-12-02 DOI: 10.1016/j.segan.2025.102076

K. Victor Sam Moses Babu , Pratyush Chakraborty , Mayukha Pal

{"title":"Integrated optimization and game theory framework for fair cost allocation in community microgrids","authors":"K. Victor Sam Moses Babu , Pratyush Chakraborty , Mayukha Pal","doi":"10.1016/j.segan.2025.102076","DOIUrl":"10.1016/j.segan.2025.102076","url":null,"abstract":"<div><div>Fair cost allocation in community microgrids remains a significant challenge due to the complex interactions between multiple participants with varying load profiles, distributed energy resources, and storage systems. Traditional cost allocation methods often fail to adequately address the dynamic nature of participant contributions and benefits, leading to inequitable distribution of costs and reduced participant satisfaction. This paper presents a novel framework integrating multi-objective optimization with cooperative game theory for fair and efficient microgrid operation and cost allocation. The proposed approach combines mixed-integer linear programming (MILP) for optimal resource dispatch with Shapley value analysis for equitable benefit distribution, ensuring both system efficiency and participant satisfaction. The framework was validated using real-world data across six distinct operational scenarios, demonstrating significant improvements in both technical and economic performance. Results show peak demand reductions ranging from 7.8 % to 62.6 %, solar utilization rates reaching 114.8 % through effective storage integration, and cooperative gains of up to $1,801.01 per day. The Shapley value-based allocation achieved balanced benefit-cost distributions, with net positions ranging from −16.0 % to +14.2 % across different load categories, ensuring sustainable participant cooperation.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102076"},"PeriodicalIF":5.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738367","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Reconstructing hourly power profiles from monthly billing data: A neural network framework with two-phase validation 从月度账单数据重构每小时电力概况：一种两阶段验证的神经网络框架

IF 5.6 2区工程技术

Sustainable Energy Grids & Networks Pub Date : 2026-03-01 Epub Date: 2026-01-23 DOI: 10.1016/j.segan.2026.102122

Morteza Aghahadi , Alessandro Bosisio , Edoardo Dacco , Davide Falabretti , Andrea Ruffini , Alessandro Cirocco

{"title":"Reconstructing hourly power profiles from monthly billing data: A neural network framework with two-phase validation","authors":"Morteza Aghahadi , Alessandro Bosisio , Edoardo Dacco , Davide Falabretti , Andrea Ruffini , Alessandro Cirocco","doi":"10.1016/j.segan.2026.102122","DOIUrl":"10.1016/j.segan.2026.102122","url":null,"abstract":"<div><div>Electrical grid planning requires accurate hourly power consumption profiles, yet utilities typically possess only monthly billing data. This study presents a neural network framework for reconstructing detailed hourly power profiles from aggregated monthly consumption features. Feature engineering transforms hourly consumption into 46 monthly aggregated features, including tariff-based totals and distribution ratios. Principal Component Analysis and K-means clustering identify 14 distinct user behavioral patterns. Three neural network architectures are systematically compared: Multi-Layer Perceptron, Long Short-Term Memory, and Gated Recurrent Unit networks. The methodology employs temporally separated validation, using 2022 data for training and 2023 data for validation, thereby assessing robustness to inter-annual variations in weather, economic conditions, and consumer behavior. Among the evaluated models, the Gated Recurrent Unit achieved the best overall performance with an <span><math><msup><mi>R</mi><mn>2</mn></msup></math></span> of 0.87 and a 40% reduction in mean squared error compared to XGBoost. For peak load estimation, which is critical for grid capacity planning, the proposed approach achieves a peak error of 18.3% for high-consumption users. Clustering stability analysis and evaluation across extreme user segments (high-consumption, high-volatility, and low-consumption) further confirm the robustness of the proposed methodology.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102122"},"PeriodicalIF":5.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146077424","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0