Application of a near-field acoustic holography-based diagnosis technique in gearbox fault diagnosis

Abstract

With the combination of the near-field acoustic holography (NAH) and pattern recognition technique, a NAH-based diagnosis technique is proposed and applied to diagnose gearbox faults by analyzing the sound field distribution information first. After visualizing the sound fields under different working conditions by NAH, the spatial gray level co-occurrence matrices based textural features can be obtained from the NAH images. Then, the support vector machine is employed to identify different working conditions and diagnose the faults. Two gearbox experiments with different gear faults and fault severity are studied in a semi-anechoic chamber to verify the NAH-based diagnosis technique. The experimental results demonstrate that the NAH-based diagnosis method is feasible and effective, and can be anticipated as a choice for gearbox fault diagnosis.