Ill-conditioning during damped least squares optimization of optical systems

Christopher C. Hull

doi:10.1117/12.246704

23 August 1996 Ill-conditioning during damped least squares optimization of optical systems

Christopher C. Hull

Proceedings Volume 2774, Design and Engineering of Optical Systems; (1996) https://doi.org/10.1117/12.246704
Event: Optical Instrumentation and Systems Design, 1996, Glasgow, United Kingdom

Abstract

Three methods have been investigated for detecting linear dependence between variables during damped least squares optimization. They are the angle (delta) between column vectors calculated from the scalar product, the normalized standard deviation of the ratios of corresponding elements and the singular values computed from a singular value decomposition. Optimization runs were made on a test lens where attempts were made to eliminate the causes of ill- conditioning by removing variables which were close to linear dependence. Results showed an improvement in the final merit function for all techniques with the exception of singular values. There was also a difference in the final solution forms which indicated that attempting to improve the conditioning of the matrix equations at source was advantageous for the test lens examined with forms closer to the optimum for the two successful techniques used for detecting linear dependence.

Citation Download Citation

Christopher C. Hull "Ill-conditioning during damped least squares optimization of optical systems", Proc. SPIE 2774, Design and Engineering of Optical Systems, (23 August 1996); https://doi.org/10.1117/12.246704

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available