配电网密闭式关柜局部放电特高频信号时延估计算法

Time Delay Estimation Algorithm for Partial Discharge Ultra High Frequency Signals in Closed Cabinets of Distribution Networks

针对配电网局部放电特高频信号累积量大幅上升,导致时延估计误差增大的问题,提出了配电网密闭式关柜局部放电特高频信号时延估计算法。通过高速信号采集系统采集局部放电特高频信号,依据卡尔曼滤波平滑处理采集信号,计算每段信号中的高阶累积量,傅里叶变换得到的信号高阶累积量。利用过零检测方法确定信号零点位置,获取任意信号角度的相移,依据位置信息最优遍历变换信号累积量,以实现高阶次搜索,完成局部放电特高频信号时延估计。试验结果表明,时延估计误差减小,时延估计值达到稳定的迭代次数较少,接收用时降低为20 ms,能够满足局部放电特高频信号时延估计需求。

In view of sharply increasing cumulant of distribution network partial discharge(PD)ultra high frequency(UHF)signals leading to rise of time delay estimation error,a time delay estimation algorithm was proposed for PD UHF signals in closed cabinets of distribution networks.A high-speed signal acquisition system was used to collect PD UHF signals,and the collected signals were processed in the Kalman filtering-smoothing method.High-order cumulant of each signal section,and high-order cumulant of signals was obtained through Fourier transform.Zero position of the signal was determined in the zero-crossing detection method,phase shift of any signal angle was obtained,and ergodic transformation signal cumulant was optimized according to position information,so as to realize high-order search and conduct time delay estimation of PD UHF signals.Results showed that time delay estimation error was reduced,the number of iterations of steady time delay estimation value stayed low,and receiving time was reduced to 20 ms,thus satisfying the requirement of time delay estimation of PD UHF signals.