A De-noising Algorithm of Turbulence Signal Based on Maximum Likelihood

Considering the problem that noise energy affects the quality of the wavenumber spectrum and the statistical characteristics of turbulence, this paper proposes a de-noising algorithm of turbulence signal based on maximum likelihood. First, the turbulence data obtained by the independently developed turbulence observation instrument (TOI) are taken as the original data, and based on the spectrum fluctuation feature, the differences between the observed spectrum and the Nasmyth empirical spectrum are taken as the processing object. Then, the cross-validation method is applied in data preprocessing to get the characteristic data, and next, through the maximum likelihood method, the discriminate function of the characteristic data is given to identify and eliminate the noise signal in the observed spectrum. Finally, the flume comparison experiment between TOI and acoustic Doppler velocimeter (ADV) is used to verify the effectiveness of the algorithm. The results show that the algorithm is feasible and effective and also improves the accuracy of turbulence data.