On a Novel Evolutionary Optimization Strategy

摘要

A novel algorithm for global nonlinear unconstrained optimization, which makes use of the interesting properties of the KS function in the framework of genetic optimization, is presented. It is particularly well suited for the optimization of continuous, highly multi-modal functions. The proposed technique is intended to improve the balance between exploitation and exploration with respect to classic genetic schemes in order to increase global convergence succes rate and convergence speed.