ЗАСТОСУВАННЯ АДАПТИВНОГО МЕТОДУ ДИФЕРЕНЦІЙНОЇ ЕВОЛЮЦІЇ КОШІ ДЛЯ РОЗРАХУНКУ ОБ’ЄКТИВІВ

Abstract

For automated design of optical lens systems, the appropriate software is to be used. To provide a design process, modern algorithms of global optimization can be incorporated into such specialized software. In this paper, a recently published adaptive Cauchy differential evolution method is numerically studied for the purpose of lens design. This method is characterized by an availability of an internal mechanism for adapting two specific parameters (coefficients) of the classical differential evolution method, as well as by applying the Cauchy distribution to generate new values of these parameters. The obtained results confirm that the adaptive Cauchy differential evolution method enables to design an arbitrary optical system with the required functional parameters and an acceptable level of correction of both chromatic and monochromatic aberrations. The performance of a lens design process greatly depends on a structure of the given merit function. The time interval, required to carry out automated design of an optical system with a number of variables about 20, does not exceed a few hours. The image quality of the obtained lenses enables to use them with existing matrix image sensors.