Violation Intelligent Recognition Algorithm Based on Multi-source Data Fusion

- The rapid evolution of the power industry, coupled with the increasing integration of intelligent technologies, has spurred a significant demand for advanced monitoring and identification systems for power equipment and their operational status. As power infrastructure becomes more complex and int...

Full description

Saved in:
Bibliographic Details
Published inJournal of Electrical Systems Vol. 20; no. 10s; pp. 1820 - 1834
Main Authors Zhao, Bishun, Zuo, Shaoqing, Liu, Ling, Li, Chen, Xu, Min
Format Journal Article
LanguageEnglish
Published Paris Engineering and Scientific Research Groups 01.01.2024
Subjects
Accuracy
Algorithms
Artificial neural networks
Circulatory system
Clean energy
Construction management
Data integration
Demand analysis
Discrete Wavelet Transform
Electric industries
Feature extraction
Homogeneity
Image filters
Machine learning
Monitoring
Multimedia
Multisensor fusion
Notch filters
Optimization
Parameter identification
Parameter sensitivity
Power flow
Sensitivity analysis
Substations
Wavelet transforms
Workflow
Online AccessGet full text

Cover

More Information
Summary:- The rapid evolution of the power industry, coupled with the increasing integration of intelligent technologies, has spurred a significant demand for advanced monitoring and identification systems for power equipment and their operational status. As power infrastructure becomes more complex and interconnected, the need for precise monitoring and real-time analysis has become paramount to ensure reliability, safety, and efficiency. In this manuscript, Violation Intelligent Recognition Algorithm Based on Multi-source Data Fusion (VIRA-BO-MSDF) is proposed. The multimedia data includes real-time sensor monitoring data, video and image data, and Power Safety Workflow includes the correct process description of power operation. Initially the multimedia data is preprocessed utilizing Master-Slave Adaptive Notch Filter (MSANF) is used to clean the data. Then, feature extraction is done by Dual Tree Complex Discrete Wavelet Transform (DTCDWT) to extract the gray-scale statistical features such as Homogeneity, Entropy, Energy and harmony. Then, the extracted features are given to Relational Bilevel Aggregation Graph Convolutional Network (RBAGCN) is used to identifying violations in the power industry. In general, RBAGCN does not express some adaption of optimization strategies for determining optimal parameters to assure accurate identification of Violation Intelligent Recognition in power industry. Therefore, Circulatory System-based Optimization (CSBO) is proposed to optimize weight parameter of RBAGCN, which accurately identify violations in the power industry. The proposed model is implemented, efficacy is assessed utilizing some performance metrics likes accuracy, precision, specificity, sensitivity, Fl-Score, computation time, ROC, Mean Squared Error. The VIRA-BO-MSDF method provides 28.46%, 21.34 and 33.81% higher accuracy, 22.88%, 26.52% and 34.63% higher Precision, 28.46%, 21.34 and 33.81% higher specificity, 22.37%, 27.89%, and 31.37% higher sensitivity is analyzed with existing method such as Deep learning-depend substation remote construction management and AI automatic violation detection system (DLB-RCM-AVDS), Applying Deep Leaming-depend concepts for the detection of device misconfigurations in power systems (DLB-DMC-PS-AVD) and Deep OPF: A Feasibility-Optimized Deep Neural Network Method for AC Optimal Power Flow Problems (FODNN-PS-AVD)respectively.
ISSN:1112-5209