A tri-objective Particle Swarm Optimizer for designing line defect Photonic Crystal Waveguides

This paper proposes a novel tri-objective approach for optimizing the structure of line defect Photonic Crystal Waveguides (PCW). A nature-inspired algorithm called Multi-Objective Particle Swarm Optimization (MOPSO) is employed as the optimizer. The three objectives considered are maximization of group index, maximization of bandwidth, and minimization of Group Velocity Dispersion (GVD). In addition, the optimization process is subject to five constraints in order to guarantee the feasibility of the structures obtained and prevent bad mixing in the final optimized structures. The results show that the tri-objective MOPSO is able to find 20 optimized structures for line defect PCW. The comparative study verifies the significant improvement of the optimized structures compared to current structures. Moreover, post analysis of the results reveals the importance of holes and their possible physical behaviours in line defect PCW. Finally, the implementation considerations and investigations show that the optimized structures are feasible for manufacturing with a resolution of 1 nm.