Abstract

In industries, all the machines have different types of mechanism for its stable operation. Among that the most\nimportant mechanism for transmitting power and to regulate speed is the gears. Continuous monitoring of such gears is\nneeded because of its high precision and high speed of operation conditions. The teeth of the gear get distorted because of\nageing and varying load conditions which leads to non-stationary operation conditions. This paper is attempt to analyze the\neffectiveness of modern time-frequency distributions called the Short-Term Fourier Transform (STFT). Also, the new nonlinear\nfeatures called HU descriptors are extracted from the Short-Term Fourier Transform (STFT) to increase the\neffectiveness of non-stationary fault diagnosis in spur gears.