In the rapidly evolving field of artificial intelligence, creation and utilization synthetic datasets have become increasingly significant. This report delves into multifaceted aspects data, particularly emphasizing challenges potential biases these may harbor. It explores methodologies behind data generation, spanning traditional statistical models to advanced deep learning techniques, examines their applications across diverse domains. The also critically addresses ethical considerations...

With the continuous development of demand electric power users, analysis users' consumption(EPS) is becoming more and important. The frequency data acquisition dimension are increasing, which makes EPS behavior precise. user tag portrait technology brings a intuitive concise expression to user's EPS. Based on massive archives, load data, considering consumption characteristics influencing factors, library built, k-means algorithm used clustering achieve different types behavior. results 2000...

As a typical form of integrated energy system, micro system (MES) is promoted and popularized at the terminal distribution grid. To achieve efficient reliable utilization MES, two main problems need to be solved: how consider dynamic characteristics heterogeneous resources diverse MES operating states under multiple time scales coordinate interests both operator grid operator. This paper aims establish models specific devices in propose coordinated control strategy for multi MESs within...

In this paper, we propose a novel descriptor for face recognition on grayscale images, depth images and 2D+depth images. It is compact effective computed from the magnitude direction difference. can be concatenated with conventional descriptors such as well-known Local Binary Pattern (LBP) Weber (WLBP), to enhance their discrimination capability. To evaluate performance of our descriptor, conducted extensive experiments three types using four different databases. The experimental results...

There are many problems caused by more and electric power big data application such as the increase of learning cost use threshold, professional specialists can unscramble result using analysis so on. According to these problems, this paper introduces a tag portrait technology which is based on includes general idea, structure key implementation The information system explained translated into language text in daily production concretized presentation research, help people pay attention...

The depth/thermal information is beneficial for detecting salient object with conventional RGB images. However, in dual-modal detection (SOD) model, the robustness against noisy inputs and modality missing crucial but rarely studied. To tackle this problem, we introduce \textbf{Co}nditional Dropout \textbf{LA}nguage-driven(\textbf{CoLA}) framework comprising two core components. 1) Language-driven Quality Assessment (LQA): Leveraging a pretrained vision-language model prompt learner, LQA...

Visual Salient Object Detection (SOD) and Camouflaged (COD) are two interrelated yet distinct tasks. Both tasks model the human visual system's ability to perceive presence of objects. The traditional SOD datasets methods designed for scenes where only salient objects present, similarly, COD camouflaged present. However, both coexist, or neither is not considered. This simplifies existing research on COD. In this paper, explore a more generalized approach COD, we introduce benchmark called...

With continuous improvement of the retention rate household electrical appliances, peak valley difference power distribution network, especially rural network is expanding. A reasonable management method demand side needed to realize load shifting and economic operation network. At present, cost energy storage(ES) equipment decreasing, which provides a new idea for optimization operation. This paper studied charge/discharge model strategy ES, constructed multi-period optimal with ES. Sparse...

The rapid update of energy information technology promotes the transformation traditional power system to integrated systems (IESs) with multi coupling and coordinated control. In this paper, a strategy based on IESs provide ancillary services in emergency scenarios is proposed. composed dynamic potential assessment control demand decision distribution operator. On basis determining optimal operating state IES, consider adjustment capacity speed each controllable device, evaluate adjustable...

When serious HVDC transmission fault happens, Emergency load shedding measures will be taken. In order to achieve the precise shedding, it proposes a real-time control decision technology based on user's interruptible load. Some key technologies involved in process of are studied just like quick response, control, stability security. Designed mechanism used such as user dynamic sequence table, preparation, cycle supplement, failure switch list and so on. Finally, application scene is...

In view of the high frequency data update, large volume amount and response requirements industrial park integrated energy system, this paper proposes an online visual analysis technology based on SVG Ajax. Based technology, a real-time monitoring system for in parks was designed developed. The developed drawing management tool vector graphics energy-using equipment, model library physical characteristics equipment. used to construct main electrical scheme, flow diagram, steam diagram so on....

Clone attack is considered as a severely destructive threat in Internet of Things (IoT), because: 1) the may be easily launched due to deficiency hardware architecture and limited resources against physical capture compromise 2) it trigger large variety insider outsider attacks. Different from traditional clone detection approaches that ground on amount data traversing network (e.g., locations identities), this article tackles problem by answering following fundamental questions: do we...

With global information technology development, the era of big data is in offing.Power industry closely related with people's livelihood and it necessary to introduce improve its economy reliability.The completion Jiangsu Electricity Information Acquisition System accumulation historical both laid a rich foundation for study electricity data.Jiangsu Power Company has an advantage take full carry out researches.Based on data, research works short-term load forecasting were carried this paper,...

This paper develops a day-ahead scheduling model for microgrids to reduce their operation cost. In view of the intermittent and uncertain photovoltaic power demand, i.e. source-demand uncertainty, novel constrained robust optimization is incorporated into model, which minimizes expected cost historical distribution conditions restricts in worst-case scenarios. To obtain empirical scenario set with temporal correlation represents distribution, Nataf transformation-based Latin hypercube...

Abstract Visual search in millions of samples high-dimensional feature space is computationally expensive and challenging. A natural solution to reduce the dimension image representation by mapping each sample into compact binary code. In this paper, we propose a Ranking Based Semantic Hashing (RBSH) method tackle with problem. Observing that semantic structures carry complementary information, paper takes advantage supervision for training high quality hashing, between reduced code, help...

Recently diverse and heterogeneous energy sources like electricity, heat gas are widely interconnected integrated to achieve full cross-border exchange of flows, information flows value which makes the research system (IES) a topic interest. However, dynamic characteristics devices optimal operation strategy IES still need be further studied, especially with integration storage system. Besides, influence mass IESs on distribution grid is also worth taking note of. This paper aims establish...