首页|期刊导航|计算技术与自动化|基于振动频谱相似度的GIS隔离开关接触状态自动检测

基于振动频谱相似度的GIS隔离开关接触状态自动检测OA

Automatic Detection of Contact State of GIS Isolation Switch Based on Vibration Spectrum Similarity

中文摘要英文摘要

GIS 隔离开关由于受到高电压及大电流的影响,其振动信号难以准确捕捉,进而影响了接触状态检测的准确性.为此,提出了一种基于振动频谱相似度的 GIS 隔离开关接触状态自动检测方法.通过求取隔离开关轮廓线上各个采样点的绝对坐标确定隔离开关的振动源,并利用特征函数获取隔离开关振动信号,引入振动频谱相似度方法将 GIS 隔离开关的振动信号从时域转换到频域,并求取特征点与时频点的余弦相似度,结合调制信号相位提取振动信号的特征频段,以此为依据,建立特征频段与隔离开关接触状态之间的对应关系,并通过经验公式将隔离开关的接触状态量化为健康指数,从而确定其接触状态.实验结果表明,利用所提方法对 GIS 隔离开关接触状态进行自动检测,得到的漏检率在 20%以下,检测精度较高.

Due to the influence of high voltage and high current,GIS isolation switches are difficult to accurately capture their vibration signals,which in turn affects the accuracy of contact state detection.Therefore,a GIS isolation switch con-tact state automatic detection based on vibration spectrum similarity is proposed.By obtaining the absolute coordinates of each sampling point on the contour line of the isolation switch to determine the vibration source of the isolation switch,and using the characteristic function to obtain the isolation switch vibration signal,the vibration spectrum similarity method is in-troduced to convert the vibration signal of the GIS isolation switch from the time domain to the frequency domain,and the cosine similarity between the characteristic points and the time-frequency points is calculated.Combined with the modulation signal phase,the characteristic frequency band of the vibration signal is extracted.Based on this,the correspondence be-tween the characteristic frequency band and the contact state of the isolation switch is established,and the contact state of the isolation switch is quantified as a health index through empirical formulas to determine its contact state.The experimen-tal results show that using the proposed method to automatically detect the contact state of GIS isolation switches results in a missed detection rate of less than 20%and high detection accuracy.

郑科森

国网西咸新区供电公司,陕西 咸阳 712000

信息技术与安全科学

振动频谱相似度GIS隔离开关接触状态特征频段状态检测

vibration spectrum similarityGIS isolation switchcontact statuscharacteristic frequency bandstatus de-tection

《计算技术与自动化》 2026 (3)

64-68,5

10.16339/j.cnki.jsjsyzdh.202603010

评论