Due to its outstanding privacy-related characteristics, Federated Learning (FL) has recently become a popular solution for the IIoT's data privacy and scalability issues. However, more research is needed determine how unique streaming in IIoT Settings affects FL-enabled architectures, with affecting accuracy reducing convergence performance. To achieve this goal, paper explains learning problem an framework enabled by FL. Afterward, it outlines two issues relevant situation: catastrophic...

The false data injection attack (FDIA) can cause the unstable operation of power grid by injecting into grid, which brings serious challenges to modern new system. However, diverse redundancy and inconsistent temporal in frequent interactive sensors significantly improves difficulty detecting attacks. To solve this aforementioned problem, we propose a multi-sensing fusion model based on hybrid deep learning network. Firstly, alignment technique is employed preprocess original multi-source...

When the system fails, a large amount of alarm information floods into dispatching center in short period time, making it difficult to quickly diagnose and process information. Dispatchers rely on wealth expertise, historical real-time grid data make optimal decisions based changing conditions. Aiming at complex power high requirements for judgment, this paper proposes method analysis decision-making knowledge graph, which condenses topology, operating procedures, disposal plans other...

Renewable energy, such as wind power, has witnessed rapid development recently on account of increasing public awareness environmental protection along with sustainable power supply. In this background, speed prediction plays an important role in helping ensure stable and economic operation modern systems a high renewable penetration. However, the stricter data regulation rules competitive demands markets set limit to employing datasets from different farms train more accurate model via...

In case of failure in the distribution network area, considering safety and stability system operation, priority should be given to reconstruction based on switch on-off completely restore power loss load. this context, economy fault recovery, firstly, recovery model is established with minimizing as objective function. Then, differential evolution algorithm natural number coding used improve mutation strategy complex order better solve model. Finally, applicability method verified by ieee33...

With the advancement of industrialization and development social economy, electric power resources have become an essential guarantee for ensuring efficient operation society. As a critical implementation step in design systems, load forecasting not only ensures safe stable systems but also assists accomplishing reasonable distribution tasks. It has significant technical economic importance. However, existing research on primarily relies expert or general time series methods, which seldom...

The rapid development of the global economy has brought a lot fossil energy consumption and environmental pollution, such as greenhouse effect caused by car exhaust. In order to fundamentally replace use energy, electric vehicles have been vigorously promoted governments all over world in recent years. However, vehicle charging pile encountered new problem process promotion: load is often unbalanced time space, which requires an accurate power forecasting scheduling model. past, algorithms...

Cyberattack on power system brings new challenges the development of modern system. Hackers may implement false data injection attack (FDIA) to cause unstable operating conditions However, from different internet things usually contains a lot redundancy, making it difficult for current efficient discriminant model precisely identify FDIA. To address this problem, we propose deep learning network-based fusion handle features measurement in Proposed includes enrichment module and module. We...

Abstract With the rapid advancement of national strategies such as "Internet+" and "Made in China 2025", industrial control systems have been widely used various industries energy, municipalities, transportation, water conservancy aerospace, etc. The security networks is always affecting lifeline economy. Therefore, has not given enough attention, resulting frequent network incidents, which makes us realize importance creating an situational awareness system that integrates assessment...

Power consumption forecasting is an important part of the macro planning industry and energy sector, accurate power load very for grid management dispatching. At present, most takes region as object, but residents small medium-sized enterprise users are basic units electricity consumption, their regional forecasting. compared with load, enterprises more uncertain difficult to forecast. Therefore, this study combines adaptive spectral clustering (ASC) method support vector quantile regression...

Efficient utilization of massive recorded data from fault indicators in distribution grid is beneficial to improve diagnosis. In this paper, the state detection and diagnosis based on indicator were realized by combining high-dimension statistical analysis artificial intelligence methods. Spectral residual algorithm was used obtain saliency map rolling time windows, which can filter out useless background information. By training model map, a GRU network with high real-time accuracy...

In order to solve the problems of frequent changes in distribution network topology and difficult identification, a identification method for medium voltage based on Copula entropy (CE) is proposed. First, function containing multivariable dependencies established. Second, CE defined by estimated using nonparametric obtain correlation matrix between nodes. Finally, maximum spanning tree algorithm used complete network. The measures precision dependency global variables, measured not related...

In recent years, with the clean energy consumption, decentralized supply and market-oriented trading, composition of power system has become increasingly complex. As one most important consumer groups, mining residents' consumption behavior is great value to strengthen demand side management, improve efficiency promote development smart grid. Therefore, this paper studies pattern recognition associated factors based on unsupervised clustering Apriori. We use hourly load curve Shanghai...

Abstract It is critical for the smart grid to mining power users’ consumption behaviors. The accumulation of data provides possibility conducting experiments research focus on behavior analysis. This paper proposes a residential classification method based typical load patterns called UCM-LP, classify behavior. firstly performs fuzzy grouping characteristics multiple users through two-stage clustering, then divides cluster label obtain final categories user. Finally, this verified algorithm...

With the development of social economy and continuous refinement industrial sectors, especially increasing importance manufacturing sector in economic growth, research on changes internal structure industry has gradually become an important part structure. In this paper, model regional manufacturing(RM) based electric Power data(PD) is analyzed studied. This paper analyzes capitalization process design PD, discusses value analysis PD assets supervision mechanism PD; The current situation...

A knowledge graph is an intelligent database that integrates artificial intelligence technology and traditional database. It a base represents concepts, entities their relationships in the objective world form of graph. can well reflect relationship between entities. In current era big data, although accuracy abnormal value detection power data much higher than methods, lack explicability, so it difficult to trace origin values. Abnormal values caused by errors must also be checked manually...

The integrated energy system is a new that coordinates variety of subsystems such as electricity, cooling and heating system. While amount renewable devices storage are allowed to be connected Therefore, the has more types data flow larger scale topo. To solve problem large-scale real-time communication in this system, paper proposed multi-path routing method based on classification flows' tasks. Besides, geographic information considered heuristic operator accelerate computation routing....