Machine learning techniques have been extensively developed in the field of electricity theft detection. However, almost all typical models primarily rely on consumption data to identify fraudulent users, often neglecting other pertinent household information such as gas data. This article aims explore untapped potential data, a critical yet overlooked factor In particular, we perform theoretical, qualitative, and quantitative correlation analyses between Then, propose two model-agnostic...

ABSTRACT Neural networks have been widely used for electricity theft detection recently. However, their decision‐making process is often not transparent, which limits the understanding of basis decisions. To address this limitation, letter proposes an explainable method with gradient‐weighted class activation mapping (Grad‐CAM). Specifically, Grad‐CAM extended to generate fraud scores by computing gradient‐based importance input features, highlighting suspicious activities. Simulation...

The integrations of advanced metering infrastructure and smart meters make it possible to detect electricity thieves by analyzing consumption readings. However, the detection accuracies traditional models are limited due their difficulty in capturing periodicity latent features from To solve this problem, a graph attention network (GAT)-based model is proposed improve accuracy fresh viewpoint on domains. First, new strategy presented transform raw one-dimensional readings into dynamic...

Offshore wind farms (OWFs) with modular multilevel converter high-voltage dc (MMC-HVdc) have become an important form of renewable energy utilization. However, if a fault occurs at the tie line between MMC and OWF, steady-state current point will be equal to zero when negative-sequence is suppressed, so traditional differential protection may fail operate. To cope this issue, article proposes new pilot method based on Minkowski distance. For internal faults, OWF different transient currents,...

Support Vector Machines (SVMs) have achieved significant success in the field of power transformer fault diagnosis. However, challenges such as determining SVM hyperparameters and their suitability for binary classification still exist. This paper proposes a novel method diagnosis, called ECOC-WSO-SVM, which utilizes White Shark Optimizer (WSO) error correcting output codes to optimize SVMs. First, t-distributed Stochastic Neighbor Embedding (t-SNE) is employed reduce dimensionality...

Recently, distributed generators (DGs) have been widely integrated into distribution network, so that the network is gradually transforming an active (ADN). Due to influence of meteorological conditions, output DGs has high uncertainty. At same time, considering increasing variety loads in ADNs, uncertainty load demand user side also increasing. In order fully consider measurement and quantitatively evaluate operational status, this paper proposes a steady-state analysis method for ADNs...

After a fault occurs in the distribution network, real-time balance between power supply and load demand must be met. Load can restored through upper sources nearby of distributed generators. Therefore, it is necessary to construct transmission path each node, namely matrix. Furthermore, access new controllable devices such as soft open point (SOP) makes possible for quickly by providing loss loads. In this paper, recovery model network considering SOP established. The dynamic node firstly...

The 10kV distribution network is an essential component of the power system, and its stable operation crucial for ensuring reliable supply. However, various factors can lead to faults in network. In order enhance safety reliability distribution, this paper focuses on analysis caused by natural factors, operational human equipment factors. It elucidates hazards resulting from proposes corresponding preventive measures different types aim mitigate or reduce impact faults, safe system.

Abstract Recently, distributed generators (DGs) have been widely integrated into distribution network, so that the network is gradually transforming an active (ADN). Due to influence of meteorological conditions, output DGs has high uncertainty. At same time, considering increasing variety loads in ADNs, uncertainty load demand user side also increasing. In order fully consider measurement and quantitatively evaluate operational status, this paper proposes a steady‐state analysis method for...

With the widespread use of electric vehicles, rapid development power electronics and smart grids making more nonlinear devices connected to grid, quality problem is becoming prominent, active filters (APFs) are widely used reduce harmonic voltage current in grid. In this paper, optimal configuration dimensions APFs for distribution networks based on improved Beluga whale optimization (IBWO) proposed. The IBWO algorithm proposed paper characterized by fast convergence high accuracy....