Load-independent condition assessment of gears using kolmogorovsmirnov goodness-of-fit test and autoregressive modeling

Abstract

A novel technique for detection of gearbox deterioration is proposed in this study. The proposed technique makes use of a time-varying autoregressive model and establishes a compromised autoregressive model based on healthy-state gear motion residual signals under varying load conditions and employs the Kolmogorov-Smirnov goodness-of-fit test statistic as a measure of gear condition. The order of the time-varying autoregressive model is selected by using a proposed fully automatic and highly effective model order selection technique with the aid of hypothesis tests. Validation of the proposed technique is carried out by using both simulated and real entire lifetime gear vibration signals and demonstrates that the proposed technique possesses appealing effectiveness in identifying the optimum autoregressive model order for robust gear condition detection under varying load conditions.