Tracking gearbox degradation based on stable distribution parameters: A case study

Gearboxes are widely used in the drivetrain of rotating machines, such as helicopters and wind turbines. Tracking gearbox degradation is essential to remaining useful life prediction and maintenance decision. Though many degradation features such as root mean square, FM0, and FM4, have been developed, the effectiveness of these features depends on individual components or specific applications. To deploy autonomous fault diagnosis system, new degradation features are often required because a larger number of features facilitate data fusion and generate more accurate degradation paths. This paper proposes to use the parameters of stable distribution to track the degradation of gearbox systems. Two sets of gearbox run-to-failure data, one from laboratory test, and the other from field wind turbine, are used to examine whether the distribution parameters are capable of serving as degradation features. The results show that stable distribution parameters can potentially monitor the degradation trajectory based on their trending profile.