【作者】 王晶；

【导师】 刘建新；

【作者基本信息】 华北电力大学（河北） ， 电工理论与新技术， 2007， 硕士

【摘要】 电力变压器是电力系统的枢纽设备之一。对电力变压器进行早期故障预测对指导维护工作具有十分重要的意义。随着变压器状态的变化,其特征量的监测频率将提高,等间隔预测模型将不再适用。电气设备绝缘故障信息与特征量之间是一个典型的灰色系统。本文提出了变压器非等间隔灰色预测方法,还从初始条件选择、背景值改造等方面对原有模型进行了改进,并应用模型群优选组合预测法,克服了单一模型仅适于预测某一数据变化规律的不足。本文还介绍了基于灰色预测的新方法的变压器故障预测系统。多台变压器色谱数据和电气实验数据证明,该预测法能有效地预测出变压器中各种特征量的变化趋势,并有较高的预测精度;还可以预测变压器内部是否存在绝缘故障以及故障的类型,也表明具有较高的实用价值。更多还原

【Abstract】 Power transformers are key element in power system. It’s important for the maintenance to predict the faults in transformer. While the transformer estate change, the frequency of supervise will elevation, then the equal-gap grey fault prediction model is not applicable. The system between the fault information and character value is a typical grey system. The transformer non-equal-gap grey fault prediction is proposed which make improvements in the selection of initial conditions and the modification of background value, etc. The grey models group prioritizing forecasting method is used. A transformer fault forecast system based on the new method of grey forecast is introduced. The test of the chromatogram data and the electric experimental data of many transformers proved that the prediction model is effective for forecasting the tendency of the transformer factors, and the precision is remarkable. And it can forecast whether the transformer is breakdown and the kind of the fault accurately by the transformer interior fault diagnosis technique, and it is proved to be practical.更多还原