Abstract: Structural complexity is first to use for the velocity time series in the field of impinging stream mixer (ISM)from the whole to each scale, combining spectral entropy with wavelet entropy algorithm. Spectral entropy algorithm depicts the structure of spectrum in the time series signal, reflecting the sequence of concentration of the frequency spectrum, which reflects the distribution of the overall average power for signal in ISM. To reflect the complexity of the signal in structure clearly, full advantage of the multi-resolution, multi-scale features of the wavelet transform are taken. The results indicated that the distribution of a4-scale wavelet entropy and the distribution of system complexity have similar structure, which indicates that the main character of the signal concentrated in the low-frequency detail components of a4. The wavelet entropy under different scales represent the tendency of the signal in every frequency range, reflecting the information of the signal in time and frequency domain. The structural complexity of the axial flow has no specific law along shaft line, but the radial is symmetric and get the maximum value in the core.

Key words: Impinging stream, Spectral entropy, Structural complexity, Wavelet entropy, Gear rattle at idle, Modeling and experiment, Tuning of pre-damper at a clutch, Vehicle transmission system