Nonlinear numerical optimization with use of a hybrid Genetic Algorithm incorporating the Modified Powell method

摘要

We propose the hybrid numerical optimization technique which incorporates the Genetic Algorithm (GA) into the Modified Powell method. Escaping from trapping in the local minimum, this procedure can seek out the global minimum with considerably fast convergence. The effectiveness was shown especially for the minimum search problem of a variable-separable multi peak (bottom) function having over 20 000 local minima within the parameter searching region.