Transform domain methods for performance enhancement of EFPI sensor

Abstract

The paper presents advances in the processing technique based on transform domain methods for performance enhancement of extrinsic Fabry–Perot interferometric (EFPI) sensors. The paper introduces a new technique to estimate the cavity length from the instantaneous phase. Pre-processing of the EFPI spectrum signal is proposed using wavelet transform to filter out the low frequency background and broadband measurement noise. The filtering technique efficiently extracts the required interference spectrum in a low cavity length EFPI. The computation of the analytic representation of the spectrum signal has been carried out using Hilbert transform. The variation computed using the proposed algorithm in a continuous test with an EFPI of cavity length 66μm is found to be as low as ±1.13nm. The range of wavelength used for processing the EFPI signal for accurate estimation of cavity length also needs to be optimized. In this regard, a definite relation between the wavelength range and cavity length measurement stability have been found. Optimization improves sensor response time and multiplexing ability.