Fine alignment of the James Webb Space Telescope with a genetic algorithm

The quality of the image recorded by the future James Webb Space Telescope or JWST will depend on the alignment of the multiple optical elements constituting the telescope. Several steps are necessary to align the optical system. In this paper, we describe a genetic algorithm (GA) that optimizes the last alignment stage or fine-phasing stage. We split the population of candidate solutions into several sub-populations in order to take into account the different parameter dependencies. Each optical element has several (seven) degrees of freedom. The worse ranked individuals undergo the GA variation operations on all their genes independently guarantying that solutions with independent genes will be searched. Conversely, since the symmetry of the problem imposes that most rotations, translations and curvatures must be related, we apply GA variation operations on only three (pseudo)-genes for better ranked solutions. Our method improves the current state of the art.