基于频域反卷积的局部放电特高频传感器时域参数研究

发布时间：2018-02-26 05:33

本文关键词： 特高频传感器有效高度时域参数反卷积 Guillaume-Nahman技术脉冲响应包络　出处：《高电压技术》2015年05期 　论文类型：期刊论文

【摘要】：时域信号特征分析是直流局部放电类型判断和严重程度分析的重要手段,而局部放电特高频检测传感器普遍采用频域有效高度作为主要性能参数,无法直观描述传感器的时域响应特性。因此提出一种通过时域参数表征特高频传感器性能的方案:在有效高度标定体系的基础上,经由Hilbert变换获取时域解析脉冲响应函数,并提取出包络峰值、包络宽度和振荡时间3种时域参数。利用Guillaume-Nahman技术克服了频域反卷积求取时域参数过程中的病态问题,避免脉冲响应淹没于高频噪声之中。使用该方案仿真研究了椭圆偶极子传感器和对数周期传感器的时域特性,结果显示后者的时域响应波形具有更高的峰峰值、更短的上升时间和更严重的拖尾效应,并通过模拟局部放电实验验证了该结论。研究结果表明,利用本方案提取出的时域参数与有效高度参数相比,能够更准确地描述UHF传感器在接收脉冲信号时的行为特征,可为局部放电UHF检测信号时域波形的有效对比和分析提供技术支撑。
[Abstract]:The characteristic analysis of time domain signal is an important means to judge the type and severity of DC partial discharge, and the effective height in frequency domain is generally used as the main performance parameter in UHF PD detection sensor. The time domain response characteristics of the sensor can not be described intuitively. Therefore, a scheme to characterize the performance of UHF sensor by time domain parameters is proposed. Based on the effective height calibration system, the analytical impulse response function in time domain is obtained by Hilbert transform. The time domain parameters of envelope peak value, envelope width and oscillation time are extracted, and the ill-conditioned problem in frequency domain deconvolution is overcome by using Guillaume-Nahman technique. The time domain characteristics of elliptical dipole sensor and logarithmic period sensor are studied by simulation. The results show that the time-domain response waveform of the latter has higher peak value. This conclusion is verified by simulated partial discharge experiments. The results show that the time domain parameters extracted by this scheme are compared with the effective height parameters. It can more accurately describe the behavior characteristics of the UHF sensor when it receives the pulse signal, and can provide technical support for the effective contrast and analysis of the time-domain waveform of the PD UHF detection signal.
【作者单位】：上海交通大学电气工程系;国网天津市电力公司电力科学研究院;
【基金】：国家重点基础研究发展计划(973计划)(2014CB239506)~~
【分类号】：TM855

【参考文献】