傅里叶算法测量介质损耗的误差分析与应用  

Error Analysis on Dielectric Loss Measurement by Fourier Algorithm and Its Application

作　　者：徐志钮[1] 赵丽娟[2] 律方成[1] 丁傲[1] 

机构地区：[1]河北省输变电设备安全防御重点实验室(华北电力大学),河北省保定市071003 [2]华北电力大学电子与通信工程系,河北省保定市071003

出　　处：《电网技术》2011年第12期124-129,共6页Power System Technology

基　　金：国家重点基础研究发展计划项目(973项目)(2009CB724508);中央高校基本科研业务费专项基金资助(11MG42)~~

摘　　要：为减少傅里叶算法用于电容型设备介质损耗(介损)计算的误差,基于信号仅含基波分量和信号频率变化的前提下对傅里叶算法用于介损计算的误差进行理论推导和分析。获得了介损角计算误差的表达式,该表达式由电压与电流信号的相位差、电流信号的初始相位、信号频率和信号长度组成,即介损角的计算误差受这些因素影响。分析结果表明:在初始相位变化范围内介损角误差的最大值、最小值大致互为相反数,介损角最大误差与频率偏离系数和电压与电流信号的相位差近似存在线性关系;介损角最大误差约为频率偏离系数乘以电压与电流信号的相位差。因此,欲减少介损角测量误差可采用以下方法:缩小采样得到电压与电流信号的相位差,比如电流信号通过移相?/2的传感器,比如末屏串入电容或Rogowski线圈获得,该方法有望使频率波动导致的误差降低为真实值的1/100左右。仿真和实验数据的计算验证了以上分析结论。To reduce the error in the dielectric loss calculation of capacitive equipments by Fourier algorithm, under the premises that the signal only contains fundamental component and signal frequency varies, the error caused by applying Fourier algorithm in dielectric loss calculation is theoretically derived and analyzed. The expression of calculation error of dielectric loss angle is obtained. The expression is composed of phase difference between voltage and current, initial phase of current signal, signal frequency and duration of sampling time, namely the calculation error is influenced by these factors. Analysis results show that inside the variation range of initial phase the maximum value and minimum value of the error of dielectric loss angle are mutually opposite number approximately, and the relation between the error of dielectric loss angle and frequency deviation coefficient a well as that between the error of dielectric loss angle and the phase difference of voltage signal and current signal are approximately linear; the maximum error of dielectric loss angle is approximately equal to the absolute value of frequency deviation coefficient multiplied by phase difference between voltage signal and current signal. Thus, to reduce the measurement error of dielectric loss angle, the phase angle difference between the voltage signal and leakage current signal should be decreased by following measures： the current signal should be passed through the sensor that can make the phase shaft of 90°, or Rogowski coil or capacitance is connected in series with the tap of the measured equipment. These measures can reduce the measurement error caused by frequency fluctuation to about one hundredth of its true value.Above-mentioned conclusions are verified by results from simulation and calculation of experimental data.

关 键 词：电容型设备 介损 傅里叶算法 相位差 频率波动 误差 

分 类 号：TM835.4[电气工程—高电压与绝缘技术]