Abstract

The problem of optical coating design was first formulated as an optimization problem by Philip Baumeister in 1958 [1]. This approach is now the most widely used design technique for optical thin films. When formulated as an optimization problem it is possible to apply all the powerful tools of modern computational mathematics to the design of optical coatings. However, in many cases numerical solution of the resulting optimization problems are extremely difficult due to the large number of local minima in the merit function. Virtually all optimization methods are subject to convergence at local minima and for this reason a search for the global minimum or even "a locally optimal minimum" is a formidable problem.

© 1995 Optical Society of America