Real-time frequency estimation for sinusoidal signals with application to robust fault detection

Abstract

This paper has investigated the problem of estimating unknown frequencies of a given sinusoidal signal with disturbances and noises, as well as its usages for fault detection. On the basis of a parametric linear model of the signal, a gradient estimator with leakage is adopted to identify the frequencies. The estimation mechanism is applied to fault detection, as it achieves not only the tolerance to disturbances but also the sensitivity to faults. Simulations verify the feasibility and capability of the frequency identifier for estimating frequencies and detecting faults. Copyright © 2012 John Wiley & Sons, Ltd.