Disturbance rejection in optical fiber specklegram sensors

Optical fiber specklegram sensors explore characteristics of modal interference fields to quantify physical and chemical measurands through a straightforward and relatively low-cost interrogation scheme. Despite their promising characteristics, speckle patterns are susceptible to extraneous mechanical and thermal effects, which impair their reliability in unconstrained environments. Therefore, this paper presents a method for rejecting disturbances in acquired speckle images, wherein a reference fiber tracks signal distortions while a proportional-integral-derivative controller-inspired routine processes error signals, producing noise-free virtual specklegrams that preserve the correlation coefficients within acceptable levels. Empirical tuning of the algorithm parameters produced a steady-state error of 0.02 and a percentage undershoot of 5.7% in response to a step-like stimulus. Experiments also evaluated the performance of a microbending displacement gauge under static perturbations, achieving absolute sensitivity (1.26×10^-2µm^-1) and resolution (4.61 µm) comparable to the undisturbed case. Moreover, dynamic tests highlighted the ability to suppress both gradual and abrupt distortions affecting the optical system, discriminating modulation signals from extraneous effects without intricate phase modulation or cumbersome machine-learning-based schemes. Such promising results inspire future developments toward robust fiber specklegram sensors for practical applications.