| 多神经网络方法在变压器油色谱故障诊断中的应用 |
| |
| 引用本文： | 徐志钮,律方成.多神经网络方法在变压器油色谱故障诊断中的应用[J].高压电器,2005,41(3):206-208. |
| |
| 作者姓名： | 徐志钮 律方成 |
| |
| 作者单位： | 华北电力大学电力系统保护与动态安全监控教育部重点实验室,河北,保定,071003 |
| |
| 摘 要： | 电力变压器的故障诊断对于变压器的维护起着至关重要的作用,诊断的可信度能给变压器维护提供更好的依据。为了克服单个神经网络不能给出诊断结果可信度的缺点,将多神经网络方法引入到变压器油色谱故障诊断中,利用多个神经网络对变压器诊断结果的方差给出了诊断结果的可信度;同时将多个网络输出的平均作为网络的诊断结果,减少了网络诊断的误差,提高了诊断的准确率。故障变压器实例验证了多神经网络方法的有效性。
|
| 关 键 词： | 变压器 故障诊断 油色谱分析 多神经网络 可信度 |
| 文章编号： | 1001-1609(2005)03-0206-03 |
Application of Multiple Neural Networks Method in DGA Fault Diagnosis of Transformer Oil |
| |
| XU Zhi-niu,LU Fang-cheng.Application of Multiple Neural Networks Method in DGA Fault Diagnosis of Transformer Oil[J].High Voltage Apparatus,2005,41(3):206-208. |
| |
| Authors: | XU Zhi-niu LU Fang-cheng |
| |
| Abstract: | Fault diagnosis is important to the maintenance work of power transformer. But the diagnosis should be reliable. Single ANN diagnosis method can not give an information of reliability. So, this paper uses multiple ANNs method to diagnose transformer fault based on dissolved gas-in-oil analysis(DGA). The results as well as the reliability is provided. The effect of multiple ANNs method is verified by practical applications of cases. |
| |
| Keywords: | transformer fault diagnosis DGA multiple ANNs reliability |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
