Automatic strain detection in a Brillouin Optical Time Domain sensor using Principal Component Analysis and Artificial Neural Networks

In this paper the performance of a distributed optical fiber sensor based on the Stimulated Brillouin Scattering (SBS) for dynamic strain detection is analyzed. The proposed scheme is based on the employment of Principal Component Analysis (PCA) to help in the detection and localization of the dynamic events employing the signal offered by a Slope-assisted Brillouin Optical Time Domain Analysis (BOTDA) sensor system. Results will demonstrate that the selection of the proposed processing scheme might prove useful, allowing identification of these events using the first PCA components. Additionally, an Artificial Neural Network (ANN) has been designed to be fed by the outputs of the PCA stage to perform the required classification.