油纸绝缘典型缺陷局放特性及缺陷类型识别

针对油浸式变压器纸质绝缘件内部缺陷导致局部放电和绝缘劣化的问题,结合三种典型缺陷纸板的局部放电图谱信号和深层神经网络的手段提出一种缺陷识别方法。根据不同类型油浸式变压器纸质绝缘件局部放电图谱信号的特点确定图谱的统计参数作为特征量。分析对比深层神经网络不同参数对识别效果的影响,寻找最优的深层神经网络结构。通过各类局部放电信号的特征量和深层神经网络来进行局部放电模式识别。研究结果表明:统计参数能够表征局部放电图谱信号的分布特征,优化深层神经网络可以提高模型的收敛速度和准确度。局部放电图谱信号的统计参数和深层神经网络相结合能够识别不同类型下油浸式变压器纸质绝缘件内部缺陷的局部放电信号,识别结果高于K-邻近法、支持向量机与反向传播神经网络。

运行环境对换流变压器油纸绝缘电场分布的影响

为研究多环境因素对换流变压器阀侧典型油纸绝缘结构电场分布的影响,在-20℃~80℃以温度和含水率为变量对变压器油和浸油纸板进行电导率和相对介电常数的测量。建立换流变压器阀侧主绝缘模型,利用实测数据建立动态材料属性,通过仿真得到不同温度、含水率在外施电压幅值为800kV的纯交流、83%直流含量、87.5%直流含量以及纯直流条件下的油纸绝缘结构电场分布。研究结果表明:当温度为-20℃或0时浸油纸板电导率随电场强度升高呈先降后升的趋势,而当温度为20℃或80℃时浸油纸板电导率随电场强度呈上升趋势。温度、含水率和外施电压直流含量均会对油纸绝缘电场分布产生影响,最大场强的位置会随外施电压直流分量与浸油纸含水率的变化而改变;运行温度下换流变主绝缘承受的场强高于常温下;水分会使绝缘纸板电导率明显上升甚至高过绝缘油,导致油中场强高于纸板中。

Reconstruction of Community Fiscal System: Background, Principles and Proposals

After the collapse of the unit-neighborhood system,the communities in China have got a nature of duality,from which a mixed dual duty structure is derived.This structure calls for a re-establishment of the principals for community financial system from such aspects as fund raising mode,budget supervision,and allocation program,etc.There exist some defects in the present community financial system,such as one-sided design,insufficient fund resource for community outlay,lower efficiency of fund operation,lacking of coordination in community construction investment,etc.which seriously restrict the functions of modern community.In order to construct a community oriented public fiscal system,we need to clarify the division of responsibilities between the government and the community,give full play to community’s potential in self-governance and construct a comprehensive fund allocation mechanism.