基于卷积神经网络的换流变压器智能检测算法设计

打开文本图片集

中图分类号：TM402 文献标志码：A 文章编号：2095-2945（2025）20-0142-04

Abstract：Aimingattheshortcomingsof traditional manual inspectionsandregularinspections，suchasloweficiencyand vulnerabilitytohumanfactors，thispaperproposesanoverallplanforinteligentmonitoringtechologyforconvertetrasfoers basedonintellgentalgorithms.First，thegenerationmechanismofconvertertransformersoundsignalsisanalyzedindetai includingnormaloperationsoundsignalsandfaultstatesoundsignals.Secondly，thevoiceprintcolectionofconvertertransfoers andthecompositionofvoiceprintuderfaultconditionsaredescribed.Ten，aninteligentconvertertransfomerdetectionalgorit basedonconvolutionalneuralnetworks（CNNs）isproposed.inally，thefeasiblityandfectivenessofthealgoritmareverified throughexperimentalresults.Theresultsshowthattheintellgentalgorithmcanefectivelyimprovemonitoringaccuracyandearly warning efficiency，reduce false alarm rate，and improve the security and stability of the power system.

Keywords：intelligentalgorithm;convertertransformer；intelligntmonitoring;faultstatesound;convolutionalneuralnetwork

换流变压器作为高压直流输电系统中的关键设备，其安全可靠运行直接关系到整个电力系统的稳定性和供电质量。（剩余5280字）