Fault Diagnosis of Gearbox Based on Interpolated DFT with the Maximum Sidelobe Decay Windows

As a flexible maintenance strategy, condition based maintenance (CBM) has been accepted by industry due to its efficiency in many engineering practices. Fault character extraction and diagnosis is a very important step in the process of CBM implementation. In order to enhance the efficiency and accuracy of diagnosis, it is a crucial task to improve the estimation accuracy of amplitude, frequency, and phase of signal. In this paper, zoom IpDFT is proposed to decompose signal and extract dominant signatures, which can reflect fault condition variation of machinery. Moreover, a novel index called Fault Index is established to describe condition development trend. In order to detect occurrence of early fault, a dynamic threshold is defined. Then, one set of vibration data collected from gearbox is used to validate the proposed method. Through comparison with other indices using gearbox vibration data, the proposed index with its dynamic threshold can not only describe the gear condition development trend with fast computation speed, but also detect the occurrence of early fault both automatically and artificially.