Aiming at the problems of difficult multi-source heterogeneous data fusion, limited resources of edge terminals, and insufficient recognition accuracy of small-target faults in power inspection and equipment state perception, this paper proposes a multi-source data fusion edge pattern recognition method based on a lightweight deep network. The revised method constructs a cloud-edge-end collaborative recognition architecture, clarifies the original difference-guided local–global feature extraction design, adds a multi-dimensional coordinate collaborative attention mechanism for multimodal and temporal feature competition, and strengthens small-target fault detection through a high-resolution four-head multi-scale structure. The experimental section further reports dataset statistics, hyperparameter settings, state-of-the-art baselines, and robustness comparisons under strong illumination, rainy/foggy, occlusion, and cross-scenario conditions. Experimental results show that the proposed method achieves 99.2% mAP in power defect detection tasks. The compressed model maintains 98.7% mAP while improving inference speed and deployment feasibility on edge platforms such as Jetson Nano and Raspberry Pi 4B.

In the integrated energy system, there is a strong coupling relationship among the electricity, cooling, and heating loads. Traditional single-task prediction models are unable to fully exploit the common features and differential information among tasks. To address this issue, this paper proposes a multi-task load prediction model that integrates efficient channel attention and multi-head self-attention (MT-ECA-BIGRU-ATT). This model uses the maximum mutual information coefficient to select features, employs the efficient channel attention-optimized temporal convolutional network to extract common spatiotemporal features, captures the specific spatiotemporal dependencies of each load through independent bidirectional gated recurrent units, and introduces multi-head self-attention to achieve weighted differentiation of features among tasks. Experimental results show that compared with the single-task model, the weighted average absolute percentage error and weighted root mean square error of the proposed model are reduced by 45.3% and 46% respectively; compared with the multi-task model without the multi-head attention module, the errors are reduced by 12.28% and 12.43% respectively; the multi-head self-attention mechanism can adaptively allocate weights according to the physical driving laws of electricity, cooling, and heating loads, effectively avoiding the negative transfer phenomenon, and verifying the superiority of the model in the collaborative prediction of multiple loads.

The carbon emissions in the power industry account for over 40% of the total national emissions. Under the new power system, it faces dual rigid constraints of load supply-demand balance and total carbon emission control. In response to the problems of insufficient source-load coordination in the existing scheduling models, the absence of carbon emission constraints on the load side, and low solution efficiency, this paper proposes an intelligent optimization scheduling model under dual constraints. Firstly, a load priority evaluation system is constructed from three dimensions: environment, economy, and reliability. Secondly, the load electricity carbon efficiency and stepped carbon trading mechanism are introduced to build a multi-constraint optimization model with the goal of minimizing the comprehensive scheduling cost. Finally, the enhanced fogging optimization algorithm with a chaotic game search strategy is adopted to efficiently solve the problem. The example results show that the proposed model can dynamically allocate load adjustment quantities based on priority levels, with low-priority users undertaking more reduction tasks. The enhanced fogging optimization algorithm requires the fewest iterations, has a significant advantage in convergence speed and solution accuracy, providing an effective solution for intelligent scheduling in industrial parks under dual constraints.

Aiming at the challenges in power equipment inspection images, including tiny defective targets, complex backgrounds, limited deployment conditions of edge devices and the difficulty of existing models balancing detection accuracy and inference efficiency, this paper proposes a lightweight defect recognition model, LAFD-Net, integrated with an attention mechanism for edge power equipment. The backbone network adopts depthwise separable convolution, Ghost feature generation and lightweight residual connection to reduce model parameters and computational complexity. A multi-attention fusion module is designed to enhance the response of critical defect areas from channel, spatial and positional dimensions. A multi-scale and multi-level feature fusion structure is constructed to improve the detection performance of small targets and weak-texture defects. Meanwhile, an adaptive keypoint detection head is introduced to boost the localization accuracy of defects with ambiguous boundaries. Experimental results show that LAFD-Net achieves a parameter volume of 7.54 M and computational complexity of 18.2 G. Its mAP@0.5 and mAP@0.5:0.95 reach 94.3% and 63.2%, respectively, with an inference speed of 118 FPS on the server platform. Deployed on a Jetson Xavier NX edge device, the single-frame inference latency is 18.3 ms, the detection speed is 54.6 FPS and the power consumption is 12.2 W. The proposed model realizes an optimal trade-off among detection accuracy, model size and edge inference efficiency and can satisfy the real-time defect recognition requirements of power inspection.

Data in power grid digital operation exhibit multisource heterogeneous characteristics, resulting in low integration efficiency and slow anomaly detection response. To address this, this paper proposes a method for power grid digital operation data integration based on K-medoids clustering. The basic service layer utilizes an Field Programmable Gate Array parallel architecture. This enables millisecond-level synchronous acquisition and dynamic preprocessing of multisource data, such as mechanical vibration, partial discharge signals, and temperature. The implementation is based on the analysis of the power grid digital operation structure. The data are then fed back to the cloud service layer, which, through business integration services, data analysis, and data access services, performs data filtering and analysis. Subsequently, the data are input to the application layer via the database server. The application layer employs a K-medoids clustering method that introduces a density-weighted Euclidean distance metric and an adaptive centroid selection strategy, significantly enhancing the clustering performance of multisource data. In particular, the proposed architecture supports real-time data processing and can be extended to cross-modal scenarios, including integration with speech-to-text systems in power grid monitoring. By aligning with low-latency neural network principles, this method facilitates timely decision-making in intelligent operation environments. Experiments confirm the method's efficacy. It acquires and integrates multisource heterogeneous power grid digital operation data effectively. The data throughput of different power grid digital operation data sources all exceed 110 MB/s. The silhouette coefficient of the integrated data sets is greater than 0.91, indicating that the integration of power grid digital operation data using this method exhibits good separability and reliability, enabling rapid detection of data anomalies within the power grid, thus laying a solid foundation for the operation and maintenance management of power grid digital operation.

In response to the growing integration of renewable energy and electric vehicle loads in distribution networks, this paper presents an optimized access scheme leveraging deep learning. We propose a Multi-Scale Topology-Aware Graph Neural Network (MT-GNN) to capture the spatial and electrical characteristics of the network, coupled with a spatiotemporal feature fusion module utilizing a dual attention mechanism to handle dynamic load and generation uncertainties. An end-to-end multitask learning framework integrates access location, capacity, and timing decisions, enhanced by a Soft Actor-Critic reinforcement learning module for adaptive strategy optimization. Experimental results demonstrate superior reliability and economic performance under uncertain conditions.

An accurate recognition of human-led operation scenarios in power grids is pivotal for developing responsive human-robot collaborative systems, enhancing operational safety, and enabling intelligent robotic assistance. This study proposes a novel deep learning framework for interpreting such scenarios from multisensor data. An optimized Bidirectional Long Short-Term Memory network is employed to decode the bidirectional temporal dependencies inherent in human operation sequences, with an attention mechanism pinpointing critical action phases. To model the structural context of the operational environment, a pruned Graph Neural Network efficiently captures the topological relationships within the grid, providing essential spatial features for robotic path or task planning. Fusing these temporal and spatial models enables a digital twin capable of robust, multi-source data integration for understanding human actions. Experimental results demonstrate that the proposed algorithm achieves a peak accuracy of 98.2% and an F1 score of 0.96, with an AUC of 0.92, significantly outperforming baseline methods. In classifying and predicting human-executed grid operation scenarios, the model average accuracy rate reaches 98.6% — an improvement of 8.3% to 12.5% over existing models — while maintaining a sub-53[Formula: see text]s response time suitable for near-real-time robotic feedback. These findings validate the framework’s strong performance and adaptability, offering a foundation for advanced humanoid robotic applications in state monitoring, fault warning, and intelligent scheduling of power systems.

Smart contracts rely on blockchain oracles to access off-chain data, yet existing oracle designs often face challenges such as untrustworthy data sources, weak temporal guarantees, and limited verifiability. This work presents Ivy Oracle, a robust and time-trustworthy data feed framework that enhances the reliability and auditability of off-chain information for smart contracts. Ivy Oracle integrates trusted execution environments (TEEs) for secure data acquisition, an external time server for authenticated timestamps, and a PageRank-based trust model to evaluate source credibility. We implement and evaluate Ivy Oracle on the Ethereum Sepolia testnet, demonstrating that it achieves up to 63.6% lower on-chain gas consumption than Chainlink for signature verification while maintaining only a slight increase in communication overhead due to its dual-attestation mechanism. These results confirm that Ivy Oracle provides strong time trustworthiness and data reliability with minimal performance cost, making it suitable for latency-sensitive blockchain applications.

The widespread adoption of cyber–physical–social systems (CPSSs) in the power industry has necessitated power marketing inspection as a critical component for ensuring secure and reliable operations of power systems. However, this effort entails significant challenges resulting from the massive volume of inspection data and complexity of electricity usage scenarios. Traditional inspection methods often fail to identify potential risks and abnormal behaviors effectively; to address this, we propose an intelligent security identification model for CPSS-based power marketing inspection by integrating advanced artificial intelligence techniques to enhance security defense and risk management. The proposed model incorporates a work order correlation matching algorithm, a fault interval detection algorithm, an electricity consumption prediction algorithm, and a business anomaly identification algorithm. Here, the users are first categorized based on multisource data to detect abnormal electricity usage precisely. Then, the model employs a correlation algorithm to uncover the intrinsic links between fault handling and electricity refund work orders for the same user, thereby revealing potential security vulnerabilities. Subsequently, the fault interval detection algorithm is used to locate fault periods, and the electricity consumed within these intervals is dynamically estimated using a prediction algorithm. Finally, an intelligent classification model based on recurrent neural networks and long short-term memory networks is developed by leveraging key security features to identify abnormal business behaviors accurately. Experiments were then conducted on three publicly available power industry datasets, and the results demonstrate that the proposed model significantly outperforms traditional methods in terms of accuracy, recall, and F1-score for security event detection. The proposed approach effectively enhances the safety and reliability of power marketing inspection for CPSSs while offering a novel technical framework for power system protection and privacy preservation.

With the rapid development of the Internet, big data information fusion technology, as a business model that breaks the traditional software marketing model, has attracted more and more attention from Internet providers. In the big data environment, multi-tenancy emerges as the times require. Cloud computing providers charge based on multi-tenant usage. Therefore, the compatibility of resources and the interest relationship between multiple tenants and suppliers become the focus of this paper. This article provides a big data cloud computing program for the multi-tenant version. Cloud platform includes development support area, technology management area and service support area. After the user pays and approves, he becomes a multi-tenant of application development, and can test his own application and bring it to the cloud platform. This article provides a reference for resource management and service scheduling of containerized big data platforms. Considering the differences and characteristics of big data services, this paper designs a big data service discovery technology, which provides a technical support for the flexible scale of the data platform. Experimental research shows that the multi-tenant cloud data transmission method adopted in this paper reduces the data transmission processing time by 20

Power equipment failure prediction method has the problem of high cumulative deterioration, and a power equipment failure prediction method based on dynamic ontology modeling technology is designed to solve the above problem. It evaluates the health status of power equipment, clarifies the performance degradation range of equipment according to the characteristics reflected in different stages, constructs a residual life judgment model by combining the mechanism of reliability function, clarifies the performance degradation conditions and failure threshold of power equipment, and optimizes the fault prediction process by using dynamic ontology modeling technology. The test results showed that the mean values of cumulative degradation of the power equipment failure prediction method in the paper and three other power equipment failure prediction methods are 1.612, 3.263, 3.207, and 3.234, respectively, indicating that the power equipment failure prediction method designed after incorporating dynamic ontology modeling technique has higher use value.

In order to improve the anti-jamming ability of monitoring process of power distribution, a new monitoring system based on LoRa wireless technology for power distribution is proposed. With the support of Lo Ra data acquisition network, we set up multisource sensors to collect data of power distribution, design the high and low level switching module of WAKE, and control its working state. Under the interference of external factors, the system monitoring data integrity is good, RSSI value within-50d Bm, anti-interference performance of the system is significantly improved.

In the future, the development of energy will change from a single energy system to an integrated energy system. The recyclable comprehensive energy system has promoted the revolution of energy production and consumption, and built a clean, low-carbon, safe and efficient energy system. GIS provides network planning engineers with more intuitive, more image, more efficient and more accurate network planning scheme. The design and construction of recyclable comprehensive energy platform based on network GIS is a complex process. The main objective of the project is to create a system to effectively manage all data related to renewable energy and transport from one location to another. In addition, the platform should be able to provide information about the status and performance of these resources throughout their life cycle.

文章分析了异构图模校验的总体思路,细致阐述了图模校验工具的研究内容和实现要点.对图模校验规则引擎、异构图模接入适配器、图模校验方法和图模校验结果的可视化展示技术进行了研究.通过基于异构系统、多部署模式的图模校验差异性适配和校验方法的研究,开发对应的变电站图模校验工具,实现了多个厂家、多个型号调度主站系统图模的定制适配、统一接入、模型校验和问题可视化展示,有力地支撑了自动化主站变电图模与GIS平台输配电图模的交互和融合工作.

With the powerful computing power of cloud computing, cloud security terminals are widely used in many industries. The information construction of the power industry will generate a large amount of data. In order to improve the efficiency of information management in the power industry and reduce the success rate of information cracking, this paper applies the cloud security terminal to the power information management. After deploying the power industry information management cloud security terminal architecture, build a terminal access control model. Cloud node load is predicted through Bayesian model to ensure smooth power information management. The improved Wu-Manber algorithm is used to protect the power information layer and ensure the safety of power circuit information. In the example verification, after the cloud security terminal is applied, the success rate of information being cracked is less than 10%, the response speed of this method is significantly improved, and the level of power information management is improved.

本文利用I EC国际电力模型标准,通过"云大物移智"等数字化技术,针对"发—输—变—配—用"各电压等级、各业务域数据,通过唯一数据编码系统源,实现数据横向贯通,构建统一的电网资源设备模型维护入口,实现发输变配图模多源异构数据的高度融合.

In order to understand the application performance of text mining technology in power enterprise complaint work order, this paper analyzes. Firstly, the text mining technology is summarized, and the basic principle of this technology is described. Secondly, the construction method of the work order text mining model is analyzed. Finally, the validity of the model is verified by a case. The results show that text mining technology can analyze the complaint work orders of power enterprises and then classify all work orders according to the analysis results, which is conducive to manual work and plays a role in improving work efficiency. At the same time, the application of text mining technology model in this paper is effective and has a certain application value. Because of the large number of modern power users, the number of complaint work orders increases, so the traditional labor is difficult to deal with all complaint work orders efficiently, and text mining technology can break through the limitations of labor and improve the status of complaint work orders.

为了提高智能电网管理水平,提出基于熵权法的智能电网管理水平评价指标量化方法.采用多直流馈入方法构建智能电网管理水平评价指标体系,以输入电流、功率因素以及电压等参数为约束指标,建立智能电网管理水平评价指标参数模型;结合无功潮流耦合分析方法进行智能电网管理水平参数分析,通过恒功率、恒电流补偿方法进行智能电网管理的最优代价函数分析,建立智能电网管理的代价约束模型;通过电压电流变化约束分析的方法,进行智能电网管理水平评价的熵权指标参数分析,提取智能电网管理水平评价体系的熵权特征量;结合大数据挖掘和智能调度,实现智能电网管理水平评价指标的量化评价.仿真结果表明,采用该方法进行智能电网管理水平评价的量化分析能力较好,评价结果准确可靠,提高了智能电网管理水平和电力资源调度能力.

In the process of building data warehouse based on Data Vault (DV), the creation and use of metadata involve multiple schemas and views of data from different sources, which inevitably bring inconsistency problems such as metadata missing, duplication and disconnection. However, due to the lack of precise semantics in DV metamodel, it is difficult to use automatic logical reasoning to detect these problems. This paper presents a formal logic DL dv in the describing logic family. Formalization upon the metamodel and metadata of DV data warehouse are analyzed in detail by using the descriptive logic DL dv . It is proved that the feasibility of applying DL dv to DV metamodel for automatically consistency reasoning.

随着电力系统智能化水平的不断提高,电网中产生的数据体系也越来越庞大,而数据的质量会直接影响电力系统的运行分析和规划决策.文中基于数据挖掘技术提出一种电网时序数据质量维护体系,筛选不合格的数据,并确定数据所存在的问题,为分析出现问题的原因提供便利.对电力数据及传输过程进行了分析,并指出了可能存在的问题.不同地区的数据具有自身不同的特点,为了提高检测速度,基于决策树算法先对历史数据样本进行决策分析.以某地区的数据训练集为例,对该地区电力数据检测流程进行分析,得到适合该区的检测顺序.针对数据合理性难以检测的问题,利用基于聚类的离群检测法筛选出问题数据,并尝试分析问题数据产生原因.通过算例证明了所提时序数据质量维护流程的有效性和可靠性.