Technique of Smoothing Spectrum Data Based on Adaptive Filter

By using the traditional spectral line smoothing methods, improper filter parameter settings or a high frequency of smoothing will cause the spectral line distortion. To address this problem, the fast Kalman filter is used to implement the recursive least squares method according to the adaptive filtering principle. The forward and backward predictors are used to update the filter parameters to denoise and smoothen the energy spectrum data. Experiments are carried out to quantitatively compare and set up to analyze the proposed method with the least squares polynomial approximation method and the least mean-squares algorithm. The experimental results show that the proposed method can effectively reduce the noise in the energy spectrum while retaining the features of energy spectrum.