Inverse Design of a PCF-Based SPR Sensor Using Machine Learning

A unique Photonic Crystal Fiber (PCF)-based Surface Plasmon Resonance (SPR) sensor design using a Machine Learning Inverse approach has been proposed in this work. The sensor is designed to predict the input parameters required to achieve a desired output. The inverse design method for the PCF-based SPR sensor streamlines the design process by predicting optimal parameters and reducing trial-and-error iterations. It also enables exploring a wider parameter space, leading to innovative sensor designs, with predictions falling within a ±10% error range around the zero line, indicating high accuracy. This approach minimizes computation time and resources, eliminating the need for extensive simulations, and makes the sensor a cost-effective, time-saving, and highly adaptable alternative for various sensing applications.