Displacement Estimation with an Optical Feedback Interferometer using an Evolutionary Algorithm

Self-mixing effect occurs when a small fraction of a laser beam backscattered from diffusely-reflecting targets reenters the laser active cavity. This generates notably variations of the optical output power (OOP) which can be monitored for sensing achievements as displacement measurements. A phase unwrapping algorithm has been performed for self-mixing sensor based on a laser under moderate feedback enabling the displacement sensor to achieve a resolution of 40 nm. Unfortunately, this algorithm leads to strong measurement errors when the signal is noisy. In order to improve the robustness of this algorithm, we characterize the variations of the OOP by computed Holder exponent using Differential Evolution Algorithms. The Holder exponent enables us to differentiate the shape of the signal in order to recover the target direction of displacement. This algorithm tested on experimental data allows to improve the performances of our sensor in terms of robustness to hostile environment.