Use of Wavelet Transform for Signal Denoising of Nuclear Power Plant Data

Abstract

A tool that reliably removed or reduced the noise from signals in nuclear processes would help considerably in monitoring the condition of safety related or process signals. In this paper, we describe wavelet-based filtering technologies, and how they have been developed and integrated into an existing Windows software package, the Plant Analysis Workbench (PAW). The added feature using the wavelet technology in PAW has been used to denoise realtime data collected in PLGS. Compared with the traditional filtering technology, the wavelet filtering technology in PAW can produce much more accurate and satisfactory results. This paper gives a complete description of the use of the wavelet shrinkage technique underlying the denoising algorithm used in PAW.