Gear Fault Diagnosis and Classification Based on Fisher Discriminant Analysis

Abstract

Gears are the most essential parts in rotating machinery. So gear fault modes diagnosis and levels classification are very important in engineering practice. This paper present a novel method in gear fault recognition and identification using Fisher discriminant analysis (FDA) due to FDA can reduct dimension when analyse signal. The real data collected from a gearbox test rig is used to validate the method this paper proposed. And the effectiveness of the methodology was demonstrated by the results obtained from the analysis. Three kinds of fault modes and levels were identified. And energy was selected as feature parameter. The fault modes (normal, breaktooth and crack) were diagnosed at first, then the fault levels of breaktooth and crack are classified. The accurate rate of the method is approximate 89%.