Multicriteria Gear Monitoring System Based on Deep Neural Networks

Gears are commonly used mechanical components in various fields of power transmission. They offer stability and high transmission efficiency. However, the issue of gear lifespan remains unavoidable. In this study, we have developed a gear monitoring system that employs deep neural networks (DNNs) and integrates data from the time domain, frequency domain, shorttime Fourier transform (STFT), and discrete wavelet transform (DWT). This system is designed to monitor the occurrence of wear in both spur and helical gears. In this research, we expand its practical applications, implement various gear fault detection methods based on deep neural networks, provide multicriteria for gear monitoring, and offer experimental results demonstrating its effectiveness.