Study about Improvement of Reliability on Maintenance TechnologyBased on NPP Condition and Medical Application Using Time Series Analysis

将奇异谱变换（SST）应用于现役设备的实时诊断和预测，开发了一种简单、早期的异常检测系统。应用主成分分析（PCA）和奇异谱分析（SSA），提出了一种对医学X射线电影进行近期预测的方法。模拟肿瘤随XYZ台移动，CCD摄像机捕捉到的运动图像用SSA方法成功预测，预测精度为1mm，可信度为95%。

The simpler and earlier abnormality detection system was developed by applying the SST(Singular Spectrum Transform) for the Real-time diagnosis and prediction of active equipment. The movie-prediction method was developed to predict the near future of the medical X-ray movie by applying the PCA(Principle Component Analysis) and SSA(Singular Spectrum Analysis). The imitation tumor was moved with the XYZ-stage and its motion captured with the CCD cameras was successfully predicted by SSA method with the 1mm accuracy of 95% reliability.