Autoregressive Model-Based Gear Fault Diagnosis

Abstract

Abstract This paper presents a model-based technique for the detection and diagnosis of gear faults. Based on the signal averaging technique, the proposed technique first establishes an autoregressive (AR) model on the vibration signal of the gear of interest in its healthy-state. The model is then used as a linear prediction error filter to process the future-state signal from the same gear. The health condition of the gear is diagnosed by characterizing the error signal between the filtered and unfiltered signals. The technique is validated using both numerical simulation and experimental data. The results show that the AR model technique is an effective tool in the detection and diagnosis of gear faults and it may lead to an effective solution for in-flight diagnosis of helicopter transmissions.