Fault Detection Algorithm for Power Distribution Network Based on Sparse Self-Encoding Neural Network

Xi, Peng; Pan, Feilai; Yongchao, Liang; Luo, Zhengxiu; Li, Long

doi:10.1109/icsgea.2017.19

Cited by 21 publications

(4 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…IF level 6 interface device value = fiber optic, subsequent fields include (fiber optic device installation location, number of fiber optic device ports, reverse order check box).IF level 6 interface device type value = network cable, subsequent fields are (fiber optic device installation location, number of fiber optic device ports, reverse order check box).IF level 6 interface Value of device type = terminal block, subsequent fields are (fiber optic device installation location, number of fiber optic device ports, inverse order check box, terminal block installation location, terminal block specification, terminal block parallel identification). The general hierarchical structure is shown in figure 2 (3) Feature analysis and AI deep learning image recognition model training are performed based on the dataset [8] .…”

Section: Data Modeling and Automatic Image Recognitionmentioning

confidence: 99%

Design of Secondary Circuit Visualization System for UHVDC Converter Stations

Liu¹,

He²,

Zhang³

et al. 2023

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

For the problems of complicated work content and error-prone workload of secondary circuits, a three-dimensional visualization method is proposed for the visualization and management of the physical circuits of the control and protection system of the converter station based on digital twin technology. The system can be used to improve the work efficiency of operation and maintenance personnel by constructing a secondary circuit information data model, generating corresponding data through artificial intelligence image recognition and manual review of secondary circuit drawings and schematics, and performing accurate corresponding circuit data analysis and 3D visualization management of the secondary physical circuits of control and security equipment, which can visualize the circuit information of control and security equipment in the station from outside to inside at multiple levels and multiple dimensions.

show abstract

Section: Data Modeling and Automatic Image Recognitionmentioning

confidence: 99%

Design of Secondary Circuit Visualization System for UHVDC Converter Stations

Liu¹,

He²,

Zhang³

et al. 2023

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

show abstract

“…The process of hyperparameter optimization can be timeconsuming and computationally expensive. In addition, selecting the appropriate model architecture for a specific power system application may require extensive experimentation and expertise [23].…”

Section:  Model Complexity and Parameter Tuningmentioning

confidence: 99%

Leveraging Deep Learning Algorithms for Predicting Power Outages and Detecting Faults: A Review

Rizvi¹

2023

AIR

View full text Add to dashboard Cite

Power outage prediction and fault detection play crucial roles in ensuring the reliability and stability of electrical power systems. Traditional methods for predicting power outages and detecting faults rely on rule-based approaches and statistical analysis, which often fall short of accurately capturing the complex patterns and dynamics of power systems. Deep learning algorithms, with their ability to learn automatically representations from large amounts of data, have emerged as promising solutions for addressing these challenges. In this literature review, we present an overview of deep learning algorithms applied to power outage prediction and fault detection. The purpose of this literature review is to explore the uses, effectiveness and advantages and disadvantages of utilizing deep learning algorithms in this domain. Various deep-learning models were explored in the context of power outage prediction and fault detection. Convolutional Neural Networks (CNNs) are effective in analyzing spatial dependencies and patterns in power system data, such as voltage levels and load distributions. Recurrent Neural Networks (RNNs), particularly Long Short Term Memory (LSTM) networks, excel in capturing temporal dependencies and patterns in time series data such as power demand and line currents. Generative Adversarial Networks (GANs) offer a unique approach by generating synthetic power system data for training purposes. This literature review involve collecting historical power system data from various sites. Deep-learning algorithms demonstrated promising results in power outage prediction and fault detection. It achieved a high accuracy rate of 95% in predicting power outages and accurately classified various fault types with an average precision of 92%. The findings highlight the advantages of deep learning algorithms in power outage prediction and fault detection. This literature review provides a comprehensive review of deep learning algorithms in the context of power outage prediction and fault detection.

show abstract

“…The wavelet transform convolves the signal with a translational stretching wavelet basis function with localization properties in both time and frequency domains, and decomposes the signal into components located in different frequency bands-time periods [24]- [26]. If the basic wavelet function (t) ∈ L 2 (R), satisfied:…”

Section: Wavelet Transform and Signal Singularitymentioning

confidence: 99%

Research on Distributed Power Distribution Fault Detection Based on Edge Computing

et al. 2020

View full text Add to dashboard Cite

With the advent of the era of Internet of Everything, the amount of data generated by edge devices in the distribution network has increased rapidly, bringing higher data transmission bandwidth requirements. At the same time, new applications have placed higher demands on the real-time nature of data processing, and traditional computing models have been unable to cope effectively. This paper proposes distributed power distribution fault detection based on edge computing, which can realize timely sensing and real-time response to distribution network faults, speed up distribution fault processing speed, shorten power outage time, improve power supply reliability and user satisfaction. Secondly, the basic principle of wavelet transform application in signal singularity detection is introduced, and a power signal fault signal analysis method based on wavelet transform is proposed. It not only makes full use of the advantages of wavelet transform in fault signal analysis, but also overcomes the shortcomings of traditional Fourier transform method, and verifies it through examples. Finally, based on the critical requirements of edge computing, such as agile connection, business real-time, data optimization, application intelligence, security and privacy protection, an evaluation system based on edge computing CROSS index fault detection model is proposed.

show abstract

Fault Detection Algorithm for Power Distribution Network Based on Sparse Self-Encoding Neural Network

Cited by 21 publications

References 10 publications

Design of Secondary Circuit Visualization System for UHVDC Converter Stations

Design of Secondary Circuit Visualization System for UHVDC Converter Stations

Leveraging Deep Learning Algorithms for Predicting Power Outages and Detecting Faults: A Review

Research on Distributed Power Distribution Fault Detection Based on Edge Computing

Contact Info

Product

Resources

About