Gear Fault Diagnosis Based on Time Domain Indexes on Harmonic Wavelet Reconstruction Signal

The vibration character is analyzed with the methods of harmonic wavelet and time domain indexes of reconstruction with simulation signal and test data. Using harmonic wavelet compactly supported characteristic and box spectrum, the data is rebuild with some specific frequency wavelet coefficients. This not only realized the frequency domain filtering, the time domain information is also retained. The analysis of time domain indexes shows the method can identify the vibration amplitude growth derived from the noise or from the useful signals. The simulation and experimental results confirm that, with the harmonic wavelet reconstruction signal time domain indexes, the gear fault detection and diagnosis can be achieved under the condition of strong noise.