Authors: T. Binder, C. Heitzinger and S. Selberherr
Affilation: Technical University Wien, Austria
Pages: 466 - 469
Keywords: optimization techniques, inverse modeling, simulation, semiconductors, microelectronics
We compare the two well-known global optimization methods, simulate annealing and genetic optimization, to a local gradient-based optimization techniques. We rate the applicability of each method in terms of the minimal achievable target value for a given number of simulation runs in an inverse modeling application. The gradient-based optimzer used in the experiment is based on the Levenberg-Marquardt algorithm. The actual implementation (llmin) was taken from MINPACK. The genetic optimzer (genopt) is based on GALIB. FOr the simulated annealing optimzer (siman) and implementation by L. Ingber was taken. All optimzers are capable of evaluating several targets in parallel.