CODEV设计资料.ppt

Fall 2001 CODE V Introductory Seminar: Day 3Optimizing Lens SystemsCustomizing CODE VReflective SystemsTilted/Decentered Systems Section 6Optimizing Lens SystemsThe AUTO Option Goal of Automatic Design (Optimization) Best performance that can be achieved, subject to problem constraints This requires A defined error function (sometimes called a “merit function”) A process to improve the error function A method of controlling boundary conditions (constraints) What is an Error Function? A single positive number that represents the quality of the optical system Structured such that smaller values are better than larger values 0 is the best error function Usually related to aberrations or image errors Can include contributions representing non-image error constraints Possible but usually not needed in CODE?V CODE?V Error Function Five types of error functions in CODE?V: Weighted transverse ray aberrations (default) Wavefront variance (OPD-based) MTF Fiber coupling efficiency User-defined Contains only image errors in the default case Constraints are handled separately Constraints can be included if desired Seldom necessary in practice Default CODE?V Error Function Default error function is basically a weighted RMS spot size Transverse ray aberrations are used (DX and DY from chief ray) All skew rays (no rays on meridional or sagittal planes) Traced for each wavelength, field, and zoom position Weighting is used Wavelength weighting (default is same as LDM) Field weighting (default is same as LDM, users may wish to override this) Aperture weighting (default tries for tight core with some allowed flare) Reducing the Error Function CODE?V uses a damped least squares (DLS) algorithm for optimization Very effective at finding the local minimum Two types of optimization Local optimization Finds the nearest local minimum in error function space Most commonly used method Global optimization Done with Global Synthesis? (you can try it very easily by choosing Optimizatio