Abstract
The paper presents a non-interference measurement and diagnostic method for the parametric identification of vibrations in rotating engine blades, based on blade tip-timing (BTT) measured by optical sensors. Because of the inherent under-sampling nature associated with BTT measurement, effective algorithms are needed to extract key vibration parameters such as frequency and amplitude from the measurement. In this paper, an enhanced estimation of signal parameters via rotational invariance technique (ESPRIT) is proposed. The main advantage of this technique is its ability to analyze both single and multi-mode blade vibrations spreading across a wide dynamic range, while accommodating the effect of varying rotational speeds and sensor installation errors. Analysis and numerical simulation have shown that the method can effectively improve the accuracy and robustness of vibration frequency and amplitude estimation compared to the traditional ESPRIT.
