Vector-Based Pattern Recognition Algorithm for Interferometric Fiber Sensor-Signals Analysis

A vector-based pattern recognition algorithm (V-PRA) for interferometric fiber sensor (IFS) signals is introduced. The proposed V-PRA transforms sensed interferometric signals into vectors to identify them by calculating the angular similarity with reference database vectors. This process accurately recognizes sensing patterns and assigns corresponding physical parameter values. The algorithm’s performance was tested in various IFS systems, including Fabry-Perot interferometers for refractive index measurement, multimode IFS for bending and strain monitoring, and multicore IFS for temperature monitoring. Experimental results confirm the effectiveness of the V-PRA as a specialized tool for IFSs.