Study on Fault Diagnosis Technology for Nuclear Power Plants Based on Time Series Data Mining
-
摘要: 将时序数据挖掘引入核电厂故障诊断,把核电厂的故障诊断当作序列监督学习问题来对待,并采用滑动窗算法将序列监督学习问题转化为经典的监督学习问题。针对反应堆失水事故(LOCA)进行的仿真实验结果表明,在采用滑动窗算法后,诊断精度有一定的提高,再进一步对滑动窗内的时序数据进行特征提取后,诊断精度有了更大的提高,可以解决经典算法无法解决的问题。Abstract: Time series data mining is applied to the fault diagnosis for nuclear power plants.dow method is used to convert the problem to a standard supervised learning problem.Simulation experiment is carried out by LOCA.The simulation results show that the diagnostic accuracy has certain improvement when the sliding-window method is applied.Furthermore,extracting the feature of the time series data in the sliding-window,the diagnostic accuracy is improved greatly.Some problems which can not be solved by classical algorithm can be diagnosed by time series data mining method.
-
Key words:
- Nuclear power plants /
- Fault diagnosis /
- Time series data mining /
- Sliding-window
-
计量
- 文章访问数: 16
- HTML全文浏览量: 5
- PDF下载量: 0
- 被引次数: 0