Identifiability Considerations for Rotating Machine Fault Diagnosis and Prognosis

Abstract

It is important to develop reliable fault diagnosis and prognosis methods for critical mechanical assets such as wind turbines. Reliable fault diagnosis and prognosis methods ensure that the damage is detected early, the damage modes are accurately characterised, and the correct remaining life is inferred. This enables the appropriate maintenance decisions to be made and can decrease the risk of unexpected breakdowns. Identifiability is an important criterion for the development of new fault diagnosis and prognosis methods. Therefore, in this work, we present the identifiability problem for fault diagnosis and prognosis on academic examples and we place a specific emphasis on gearbox applications. This chapter provides an overview of the concepts and is intended for neophytes to experienced researchers and practitioners. Hence, the examples are purposefully simple. We specifically highlight the importance of sensor positioning and also discuss the influence of varying operating conditions on the diagnosis and prognosis steps. Thereafter, we present the fundamental steps in the fault diagnosis and prognosis process and highlight the associated challenges with identifiability. We also propose potential solutions for these challenges. Lastly, we propose requirements for the different phases of the fault diagnosis and prognosis steps, which could be beneficial when developing new methods.