Research on the Application of Double Tree Complex Wavelet Analysis in Mechanical Fault Recognition under Quantum Control

In view of the assumption that the wavelet coefficients in traditional wavelet analysis completely conform to the Gaussian probability distribution and ignore the difference of wavelet coefficients in different positions, a double tree complex wavelet analysis algorithm under quantum control is proposed. The algorithm takes into account the characteristics of "high peak" and "long tail" of wavelet coefficients, regards the vibration signal as the state superposition of fault signal and noise signal, introduces the mixture Gaussian distribution model, uses quantum theory to adjust and control the parameters in Bayesian theory, and realizes the nonlinear shrinkage of wavelet coefficients. Experimental results show that compared with the traditional wavelet soft threshold denoising, this algorithm has a stronger ability of denoising.