Abstract

In the optimization of lens design, the damped least squares method (DLS) allows us to find a solution easily, but it corresponds to a local minimum of the merit function which happens to lie near the starting design. There must be many other solutions in the wider area, some of which can be much better than the first solution. When the design falls into a local minimum, it is almost impossible with DLS to get out of that trap to find other solutions, because the control of damping factor forces the design to remain within that trap.

© 1998 Optical Society of America