A multimodal adaptive detection algorithm for optical fiber mems Fabry-Perot sensor

Considering about the problem of distorted spectrum of optical fiber MEMS Fabry-Perot sensors which cannot be solved by the traditional peak finding algorithm, we have proposed an adaptive algorithm for peak finding to resolve the problem of spectral distortion. The multimodal signal is divided into a single-peak signal by the Hilbert transform, and a number of independent interference peak spectra with only one peak are obtained, and the number of data points on the left and right sides of the peak point is used as the basis to determine whether the distortion is determined. For the distortion spectrum, a distortion spectrum correction algorithm based on Gaussian fitting was proposed by constructing a correction function to achieve accurate peak finding of the interference spectrum.The simulation results based on MATLAB show that compared to traditional peak finding methods such as centroid algorithm, polynomial fitting method, Gaussian algorithm, etc., the algorithm is more suitable for different degrees of distorted spectrum, and the accuracy of which is higher with an average error of ±3pm. We have carried out experiments and the result shows that the algorithm error is less than 0.0051MPa within the pressure range of 0-3MPa.