Order Tracking in Vibro-acoustic Measurements: A Novel Approach Eliminating the Tacho Probe

摘要

When analyzing the behavior of rotating or reciprocating machines, it is often necessary to estimate and track the phases and magnitudes of harmonic orders of the fundamental frequency of the machine. The fundamental frequency is typically measured using dedicated sensors, e.g., proximity probes or photo-sensors that require direct access to the rotating parts of the machine, which can be difficult. In this paper we propose a new order tracking method where the fundamental frequency is extracted from the measured vibration and/or acoustic signals. Hereby the need for a dedicated tacho probe is avoided. Apart from the fundamental frequency, the method can also estimate the amplitude and phase of specified orders. The method is founded on Bayesian statistics, where the posterior distribution of the fundamental frequency is computed conditioned on the measured data and a harmonic model of the signal. Limited prior knowledge is required for initialization. The new method is tested on vibration signals from the drivetrain of a passenger car and is compared with two other order tracking methods: re-sampling based order tracking and the Vold-Kalman tracking filter. The results are promising, and the method is proven useful in situations where obtaining a robust tacho reference is difficult, e.g., to estimate the running speed of a small turbo-charger.