Numerical solution of adaptation problems by means of an evolution strategy

Adaptation problems are characterized as general extremum problems. For the numerical solution of such problems, the paper considers evolution strategies, that means direct stochastic methods being organized according to principles of the biological evolution. These strategies are assessed with special emphasis being placed on their efficiency. The paper presents a new evolution strategy simulating the “genetic load” principle — the EGL method, describing it as being efficient and convergent with probability 1. A successful application for the solution of a practically relevant adaptation problem is referred to.