-
Seeks minimum of cost function (over all cases specified)
-
Parameters for each optimization experiment are selectable by name from any one input parameter set item via a free parameter table
-
Micro-genetic algorithm (Krishnakumar, K., 1989, Proc. SPIE: Intelligent Control and Adaptive Systems 1196, Philadelphia, PA, 289-296), based on an implementation by Carroll (Carroll, D. L., 1996, in Developments in Theoretical and Applied Mechanics, Vol. XVIII, eds. Wilson, H. B, Batra, R. C., Bert, C. W. Davis, A. M. J., Schapery, R. A., Stewart, D. S. & Swinson, F. F., School of Engineering, The University of Alabama, pp.411-424)
-
User-definable bounds for each parameter or log-transformed parameter
-
Initialization from input table of parameter vectors or random
-
Local non-gradient direction set algorithm (Powell, M.J.D., 1964, Computer J. 7, 155-162) using Brent's line minimization algorithm (Brent R.P., 1973, in: Algorithms for Minimization without Derivatives, Prentice-Hall, Englewood Cliffs, N.J., 61-80)
-
User-definable bounds or no bound option for each parameter or log-transformed parameter
-
Repeat searches with different initial parameter vectors from input table or genetic algorithm output
-
Table of final parameter vectors and costs; optional output of all cost function values and corresponding parameter vectors to evaluations table(s)
-
Detailed optimizer report
-
Simulation output tables produced for final optimal parameter vector
