Automatic Correction of Residual Aberrations

Abstract

A method for the automatic design of optical systems, programed on an electronic digital computer, is described. The method is based on the reduction of an error function which is determined from the spot diagrams of the system. The minimization is achieved by an iterative advance along the steepest slope of this function. With the aid of Lagrangian multipliers secondary demands, such as avoidance of large curvatures, can also be fulfilled. An example of the final correction of a Cooke triplet is outlined.