With the construction and promotion of Ubiquitous Power Internet Things (UPIoT), it is an increasingly urgent challenge to comprehensively improve recognition accuracy gasinsulated switchgear (GIS) partial discharge (PD), incorporate model into UPIoT intelligent terminals supported by edge computing in embedded systems.Therefore, this paper proposes a novel MobileNets convolutional neural network (MCNN) identify GIS PD patterns.We first construct pattern classification datasets means...

Intelligent fault diagnosis methods, especially convolutional neural network (CNN), have made significant progress in gas-insulated switchgear (GIS) partial discharge (PD) diagnosis, which are attributable to two reasons: 1) the training and testing samples come from identical distribution; 2) there massive labeled data with PD information. However, owing specific operating conditions of GIS, collecting same distribution is difficult field conditions. With purpose resolving dilemma...

With the increase in applications of face verification, increasing attention has been paid to their accuracy and security. To ensure both safety these systems, this paper proposes an encrypted face-verification system. In paper, features are extracted using deep neural networks then with Paillier algorithm saved a data set. The framework whole system involves three parties: client, server, verification server. server saves user ID, performs client is responsible for collecting requester's...

The convolutional neural network (CNN) achieves excellent performance in pattern recognition owing to its powerful automatic feature extraction capability and outstanding classification performance. However, the actual samples obtained are unbalanced, accurate diagnoses difficult for existing methods. A method partial discharge (PD) gas-insulated switchgear (GIS) that uses a generative adversarial (GAN) CNN on unbalanced is proposed. First, novel Wasserstein dual discriminator GAN used...

Partial discharge (PD) is one of the major form expressions gas-insulated switchgear (GIS) insulation defects. Because PD will accelerate equipment aging, online monitoring and fault diagnosis plays a significant role in ensuring safe reliable operation power system. Owing to feature engineering or vanishing gradients, however, existing pattern recognition methods for GIS are complex inefficient. To improve accuracy, novel method based on light-scale convolutional neural network (LCNN)...

The construction of the ubiquitous power internet things (UPIoT) provides a new feasible solution for gas‐insulated switchgear (GIS) online monitoring and fault diagnosis, but it also puts forward greater requirements time accuracy. How to find an effective real‐time model that can be applied UPIoT mobile terminals has become urgent problem needing solved. To this end, study proposes lightweight convolutional neural network (LCNN) GIS partial discharge (PD) pattern recognition using three...

This paper uses a neural network approach transformer of taxi driver behavior to predict the next destination with geographical factors. The problem predicting is well-studied application human mobility, for reducing traffic congestion and optimizing electronic dispatching system’s performance. According Intelligent Transport System (ITS), this kind task usually modeled as multi-class problem. We propose novel model Deep Wide Spatial-Temporal-Based Transformer Networks (DWSTTNs). In our...

Convolutional neural networks (CNNs) have promoted the development of insulation defect diagnosis for gas-insulated switchgear (GIS) because their excellent feature extraction and classification capabilities. However, CNN ignores correlation between local areas space, resulting in insufficient utilization. Moreover, deploying applying methods explored massive laboratory data to complex few-shot conditions on-site is a difficult problem currently. Therefore, this study proposes novel domain...

Localization based on differences in the timing of an ultrahigh-frequency signal is difficult to apply actual settings because strict requirement for accurate calculation difference arrival time between discharge pulses. With development digital twinning, operational state equipment mapped onto a virtual model, and large-scale simulation data can be obtained assist localization partial (PD) gas-insulated switchgear (GIS). this aim, paper proposes novel deep domain-invariant long short-term...

Recently, convolutional neural networks (CNNs) have made certain achievements in gas-insulated switchgear (GIS) partial discharge (PD) pattern recognition. However, these methods rely on the availability of massive PD samples and how to apply CNN constructed laboratory field GIS recognition has become an urgent problem. To solve problems, we propose a small sample using one-dimensional (1DCNN) domain adversarial transfer learning (DATL). First, novel 1DCNN is achieve high-accuracy...

Abnormal discharge in gas-insulated switchgear (GIS) is a key cause of insulation failure, and it also an external manifestation defects. Sensitive source localization important goal GIS partial (PD) monitoring. However, most existing PD methods rely on time-delay estimation, which not only requires high-precision synchronous sampling device but suffers from serious interference. To this end, collaborative domain adaptation network (CDAN) proposed for localization. First, squeezing wavelet...

Destination prediction is an essential task in various mobile applications and up to now many methods have been proposed. However, existing usually suffer from the problems of heavy computational burden, data sparsity, low coverage. Therefore, a novel approach named DestPD proposed tackle aforementioned problems. Differing earlier that only considers starting current location partial trip, first determines most likely future then predicts destination. It comprises two phases, offline...

Photovoltaic (PV) output is greatly affected by meteorological factors. If it has no efficient factors, the prediction accuracy for PV a little low. Although Radial Basis Function (RBF) network already widely utilized in photovoltaic prediction, its error too large. An algorithm forecasting evaluation of short-term based on fuzzy clustering data and joint Genetic Algorithm Programming System (GAPS) proposed this paper to increase accuracy. Selecting three main types data, including...

Intelligent transportation system needs to solve the main problems in traffic safety. This paper focuses on safety caused by fatigue driving based image recognition of key technologies for research and analysis. proposes that location face facial feature points classification detection are links determine rate. In analysis localization algorithm skin color modeling, a corner-based optimization method is proposed optimize region. Based binary human eye algorithm, bi-directional integral...

The data-driven models have achieved remarkable advancements in diagnosing gas-insulated substation (GIS) insulation defects on specific massive datasets. However, restricted by field operating conditions and sample scarcity of GIS, the above methods are difficult to achieve high-precision robust diagnosis GIS on-site. To settle these problems, we propose an innovative domain adversarial graph convolutional network (DAGCN) for few-shot improve accuracy diagnosis, a (GCN) with multiple...

针对PM2.5浓度预测中传统机器学习算法无法对数据内部隐藏特征进行深层次挖掘,而深度学习算法在数据较少情况下效果不佳的问题,综合考虑深度学习与随机森林的特点,提出一种基于深度学习与随机森林的PM2.5浓度预测组合模型。模型以气溶胶光学厚度(AOD)遥感数据、气象再分析数据和PM2.5地面观测数据构建训练数据集,通过深度学习方法对训练数据内部深层次隐含特征进行提取,将提取得到的隐含特征用于随机森林模型训练,并使用随机森林回归算法得到PM2.5浓度的预测值。为验证方法的有效性,以河南省区域2018年—2019年的PM2.5浓度估算为例,将原始特征与利用CNN、LSTM和CNN_LSTM所提取特征共同构建的新特征分别通过随机森林回归、支持向量回归以及K近邻回归等3种传统机器学习方法进行训练和预测。实验结果表明,在较少数据情况下PMCOM模型无论是在整体预测还是在分季节预测场景下均具有较好的预测精度,其中以LSTM为特征选择器,RF为回归器的组合模型是本实验的最优模型,在即使只有35%的数据作为训练样本时,整体预测实验中R2仍可达0.89,各季节预测实验中R2均在0.75以上。

Abstract Due to the requirement for highly precise synchronous sampling and substantial reliance on time difference calculations, current partial discharge (PD) localization based of arrival is only applicable in certain situations. As digital twin technology has advanced, it possible employ virtual models support gas‐insulated switchgear (GIS) PD localization. To do this, we propose a meta‐learning (ML) network with aid actual GIS Firstly, model was established acquire an auxiliary...

Location prediction has attracted much attention due to its important role in many location-based services, including taxi route navigation, traffic planning, and advertisements. Traditional methods only use spatial-temporal trajectory data predict where a user will go next. The divorce of semantic knowledge from the one inhibits our better understanding users’ activities. Inspired by architecture Long Short Term Memory (LSTM), we design ST-LSTM, which draws on trajectories future locations....

In this study, the partial discharge (PD) characteristics of nonuniform electric fields in SF6 and SF6/N2 gases from PD inception to intermittent breakdown under negative dc voltage were systematically studied. The effects gas mixture ratio field distortion on analyzed terms (PDIV), amplitude, repetition frequency, time-resolved (TRPD) patterns. PDIV was higher than that within a certain range ratios, which is related degree distortion. greater electrical distortion, mixing SF6. difference...