International Symposium on Symbiotic Nuclear Power (ISSNP2024)

This paper presents an innovative measurement noise reduction algorithm strategy that combines techniques based on Empirical Mode Decomposition (EMD), Singular Value Decomposition (SVD), and Wavelet Transform (WT), with the aim of enhancing the quality of signals in nuclear power plant monitoring systems. This algorithm is capable of effectively separating noise components from the original signal, thereby extracting more clear and stable effective signals for use in areas such as state monitoring, fault diagnosis, and life prediction. The implementation process involves first decomposing complex signals into a series of Intrinsic Mode Functions (IMFs) using EMD technology, followed by further decomposition using SVD, and finally performing a multiscale analysis and reconstruction of the signal through wavelet transform. The results indicate that this algorithm outperforms traditional denoising methods by significantly reducing noise levels while preserving signal detail, resulting in improved signal-to-noise ratios and smoothness. Additionally, the algorithm showed excellent stability and adaptability in the application on a simulated operational platform of nuclear power plants, providing valuable references and support for further developments in state monitoring, fault diagnosis, and life prediction in nuclear power plants.

EMD,SVD,WT,Noise Reduction,nuclear power plant