Determination of the Inputting Data Length for the Diagnosis of the Operation Trend Based on Wavelet Analysis

The trend features, which are named trend symptoms later, of the operation conditions of critical equipment in a nuclear power plant can reflect its operation conditions, especially the potential failures. The trend features can be acquired based on the related physical quantities, such as temperatures, pressures, and flows, measured by sensors. When judging the trend symptoms of the operation and equipment, it is often necessary to select a suitable inputting data length. This paper proposes a method to determine the inputting data length based on wavelet analysis. The equipment history data are filtered by wavelet analysis to remove the noise and burr, then the smooth data curve after filtering will be segmented by a time series segmentation method based on local maximum and minimum, and the analysis results are obtained by linear fitting of the data within each segment. Then, the shortest segment of the segmentation results is conservatively selected as the reference value for the inputting data length of the corresponding equipment. The effectiveness of the proposed method is demonstrated by testing with several pieces of equipment history data. The results show that the proposed method can improve real-time performance while ensuring the accuracy of determining the symptom of equipment operation data.