Helical Gearbox Defect Detection with Machine Learning Using Regular Mesh Components and Sidebands

Iulian Lupea,Mihaiela Lupea,Adrian Coroian

DOI: https://doi.org/10.3390/s24113337

IF: 3.9

2024-05-24

Sensors

Abstract:The current paper presents helical gearbox defect detection models built from raw vibration signals measured using a triaxial accelerometer. Gear faults, such as localized pitting, localized wear on helical pinion tooth flanks, and low lubricant level, are under observation for three rotating velocities of the actuator and three load levels at the speed reducer output. The emphasis is on the strong connection between the gear faults and the fundamental meshing frequency GMF, its harmonics, and the sidebands found in the vibration spectrum as an effect of the amplitude modulation (AM) and phase modulation (PM). Several sets of features representing powers on selected frequency bands or/and associated peak amplitudes from the vibration spectrum, and also, for comparison, time-domain and frequency-domain statistical feature sets, are proposed as predictors in the defect detection task. The best performing detection model, with a testing accuracy of 99.73%, is based on SVM (Support Vector Machine) with a cubic kernel, and the features used are the band powers associated with six GMF harmonics and two sideband pairs for all three accelerometer axes, regardless of the rotation velocities and the load levels.

engineering, electrical & electronic,chemistry, analytical,instruments & instrumentation

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to detect gear faults in spiral gearboxes by machine - learning methods. Specifically, the researchers focus on how to use the general meshing frequency (GMF) in vibration signals, its harmonics, and side - band frequencies to identify different health states of gears, including poor lubrication, normal state, and two structural defects - local pitting and local wear on the tooth surface. The main purpose of the research is to establish an effective gear - fault - detection model by analyzing the characteristics of vibration signals at different rotation speeds and load levels, so as to improve the accuracy of fault diagnosis.

The research questions proposed in the paper include:
1. **RQ1**: Among the vibration signals measured by a three - axis accelerometer at three rotation speeds and three load levels, which time - domain and frequency - domain statistical parameters are the most relevant for detecting the health state of the spiral gearbox?
2. **RQ2**: How much does adding the peak amplitudes of the GMF frequency band on all three axes to the selected statistical features improve the performance of the defect - detection task?
3. **RQ3**: Based on the amplitude and phase modulation of vibration signals, how effective is it to use the band power or/and peak amplitude related to GMF, harmonics, and side - band pairs as classification features?

To answer these questions, the researchers designed an experimental platform, collected vibration data in different health states, and extracted multiple types of features through signal - processing techniques. These features were then used to build multi - class classification models to achieve automatic detection of gear faults. The focus of the research is to explore how to effectively distinguish different gear - fault types by using specific frequency components in vibration signals.