Research and application of new threshold de-noising algorithm for monitoring data analysis in nuclear power plant

Under the complex condition of nuclear power plant, all kinds of influence factors may cause distortion of on-line monitoring data. It is essential that on-line monitoring data should be de-noised in order to ensure the accuracy of diagnosis. Based on the research of wavelet analysis and threshold de-noising, a new threshold denoising method based on Mallat transform is proposed. This method adopts factor weighing method for threshold quantization. Through the specific case of nuclear power plant, it is verified that the algorithm is of validity and superiority.     

Two-Factor Fuzzy Time Series for Equipment Data Prediction

The data forecasting of plant equipment plays an important role in assurance of the safe and reliable operation of the plant equipment. Thus, it is necessary to improve the accuracy of data forecasting of the equipment. A new two-factor fuzzy time series algorithm is proposed to forecast the data of the plant equipment. This method not only overcomes the limitations of one factor fuzzy time series algorithm, but also overcomes the drawbacks of traditional two-factor fuzzy time series algorithm. The collected data is used in the power plant to conduct experiments, where the metrics is Mean Absolute Percentage Error (MAPE). The results show that this method is superior to the existing two-factor fuzzy time series algorithms, and yields good results in the equipment prediction.