Parametric Bayesian model for rotating blade frequency tracking with single probe blade tip timing

Abstract

Blade tip timing (BTT) has become the most promising online monitoring method for turbine blade health. Most existing BTT data analysis methods have a narrow application scope for different types of BTT data and slow calculation speed, which adversely affects the practical industrial application of BTT. In this study, we propose a parametric Bayesian model for rotating blade frequency tracking with a single probe. Compared with existing methods, this method has a wider application range and fast computation speed; moreover, it minimizes the number of probes used. Numerical simulation and experimental results show that the method is highly feasible and robust. In addition, the advantages of this method in signal amplitude estimation are demonstrated. The results are significant for online blade health monitoring of turbines.