Denoising SPND signal by discrete wavelet analysis for efficient power feedback in regulating system of PHWRs under noisy environment

This paper aims to denoise a neutron power sensor data which is the output of Self Powered Neutron Detector (SPND) in each of the Liquid Zone Compartments (LZC) in a nuclear reactor core. Noisy data measured over long period is analyzed with different types of wavelets by one dimensional Discrete Wavelet Transform (DWT) technique to investigate the local feature of the measured signal by selecting proper basis wavelet to feedback the processed signal for global reactor power control and safety purposes. Denoising has also been carried out, considering the SPND signal, corrupted with fractional order (FO) noise. A relative Maximum Absolute Error (MAE) based analysis, adopted in this paper is done to precisely compare the wavelet performances in integer and fractional order type noisy SPND signal.