Optimization of filter radiometer spectral response

The design of filters with specific spectral characteristics is a requirement not only for the design of filter radiometers, but also for many applications in optical measurements. The most general type of absorptive filters are composite subtractive-additive filters and the general problem of filter radiometer spectral response optimization using such filters is formulated. The algorithm and software realization of constrained optimization for various objective functions with arbitrary weight functions are described. Successive random search and Hooke-Jeeves methods are employed in the optimization and several goodness-of-fit criteria are used for evaluation of the results. Illustrative numerical examples are presented.