Robust Optimization of Nanoslit Array Sensor Based on Extraordinary Optical Transmission

Abstract

A robust optimization approach for plasmonic periodic array sensor based on extraordinary optical transmission is proposed using Kriging surrogate models to reduce the effects of uncertainty in various manufacturing processes while maintaining sensor performance. For systematic design with reasonable computation cost, the author adopt the universal Kriging models whose regression function is a polynomial. The gradient index and the multiobjective genetic algorithm are chosen as a robustness measure and a global optimization tool, respectively. The figure of merit and the gradient index are set as two objective functions, and the design variables are the slit width and height, respectively. The optical properties of interest are investigated using the finite-element method. The numerical optimization results show the proposed scheme to be powerful and efficient in designing nanoslity array sensors based on extraordinary optical transmission with fabrication uncertainty.