Designing lens systems taking account of glass selection by real-coded genetic algorithms

Presents a lens design method based on a real-coded genetic algorithm (GA) taking account of glass selection. In designing lenses for a single wavelength, it is only required to optimize curvature of surfaces and gaps between surfaces that are continuous variables. The authors have proposed a real-coded GA using the UNDX as a crossover operator for lens design, which has shown good performance in optimizing continuous variables, curvature of surfaces and gaps between surfaces. However, the GA cannot be applied to designing lenses for multiple wavelengths such as camera lenses, microscope ones and telescope ones. In designing lenses for multiple wavelengths, it is also required to optimize glasses of lens elements that are discrete variables. In the paper, we enable the real-coded GA using the UNDX to optimize glasses as well as the curvatures of surfaces and gaps between surfaces by devising an encoding/decoding scheme. We show the effectiveness of the proposed method through some experiments.