摘要
对油纸绝缘等效电路参数进行准确辨识是当前研究的一个难点。为此摒弃了传统利用恢复电压特征量对油纸绝缘等效电路建模和参数辨识方法,并提出了一种电路建模和参数辨识的新方法。恢复电压曲线含有丰富的介电弛豫响应信息,利用微分解谱得到包含介电弛豫响应信息的子谱线,分析子谱线弛豫参数与等效电路参数的关系并求解等效电路参数。该方法可以准确地确定等效电路拓扑结构,能避开复杂的方程组运算而准确辨识电路中的参数量,并可求解出各极化支路的非典型自由弛豫的线型参数。同时,对多台变压器绝缘等效电路的部分参数与油纸绝缘老化状态的关系进行了研究,结果表明:变压器绝缘状态良好时,恢复电压曲线可分解5条子谱线,线型参数较小且接近0.5;子谱线数量越多,线型参数越大,则变压器油纸绝缘老化越严重。因此,利用子谱线数量和线型参数能准确评估油纸绝缘老化状态。
Analyzing the equivalent circuit of oil paper insulation is one of the effective ways to diagnose aging of transformer insulation. It is difficult to accurately identify the parameters in current research. We abandon the method of using the recovery voltage characteristics to identify the parameters of oil paper insulation, and present a new method to identify circuit model and parameter. Adopting the voltage curve which contains rich dielectric relaxation response information and the differential spectrum response voltage of dielectric relaxation response information which contains dielectric relaxation response information, we analyze the relationship between the relaxation parameters and the parame- ters of the equivalent circuit. The new method can determine the equivalent circuit topology, and avoid complex computation equations. Meanwhile, it can solve the linear parameters of atypical free relaxation, and the results show the method is feasible and accurate. Finally, the relationship between equivalent circuit parameters of transformer insulation and aging status is studied. The results indicate that the recovery voltage can be decomposed to five sub-lines, and the line shape factors close to 0.5 when the insulation performance is good, the number of sub-lines and line shape factors increases with ageing. Therefore, the number of sub-lines and linear parameters can accurately reflect the state of oilpaper insulation aging.
出处
《高电压技术》
EI
CAS
CSCD
北大核心
2017年第6期1988-1994,共7页
High Voltage Engineering
基金
国家自然科学基金(61174117)~~
关键词
油纸绝缘
参数辨识
时域微分解谱
恢复电压
绝缘诊断
oil-paper insulation
parameter identification
time-differential unfolding
recovery voltage
insulation diagnosis
作者简介
贺德华（通信作者）,1983-，男，博士生,主要从事电气设备故障诊断的研究,E-mail：153367542@qq．com
