Genetic algorithm search to reach an optimal layout for DLS optimization

Before computers, conventional optical design methodologies employed paraxial ray tracing based on the thin lens model, typically executed in a two-step process: the first step involved third-order chromatic correction through judicious material selection and power distribution, while the second step focused on refining lens shapes to minimize residual monochromatic Seidel aberrations. Although modern computing enables full geometric ray-trace optimization, selecting optimal glass combinations remains a significant challenge due to the discrete nature of the glass parameter space. This paper proposes a hybrid optimization framework that combines two-ray paraxial analysis of a thick-lens model with a Genetic Algorithm (GA) search to optimize glass selection. Each candidate design in the GA population encodes a specific set of glass materials and undergoes local re-optimization via a simplex method for fitness evaluation. The merit function is constructed as a weighted sum of Seidel aberrations. The best-performing solution is then used as the initial design for full geometric ray-trace optimization, ensuring both global material selection and local performance refinement.