Estimation of Spectrum Envelope for Gear Motor Monitoring Using A Laser Doppler Velocimeter

Abstract

A new spectrum envelope estimation method for use in rotary machine monitoring is proposed. Aspects including the Internet of things, monitoring, sensing and signal processing have been considered for diagnostics of low-speed rotary machines. In this paper, signal processing in sensing when using a laser Doppler velocimeter for monitoring of low–rotational-speed machines such as gear motors is studied. It is possible to measure the shaft vibration of a rotary machine directly via laser Doppler velocimetry. Therefore, this type of velocimeter is suitable for replacement of the current monitoring methods used in vibration measurement of low-rotational-speed machines. However, the laser Doppler velocimeter shows low sensitivity to the harmonic spectrum structure, which is the most important information required when monitoring a gear structure. The balance of each harmonic spectrum structure can be confirmed using a spectral envelope. Therefore, a signal processing technique that addresses the problem of low sensitivity to harmonic components during spectral envelope estimation is proposed. By focusing on the peak frequency of the spectrum, which is a major technique in the signal processing field, a method for construction of an autocorrelation function and estimation of the spectrum envelope is proposed. The proposed spectral envelope estimation method is verified by an illustration using the gear meshing frequency. The effectiveness of the proposed estimation method is confirmed using a visualization graph.