Two membered evolution strategy ES EVOL implementation

Abstract

It is well known that the genetic algorithms are designed as a general search method which can be, also used in the optimization of various technical systems. There is a large number of different genetic algorithms but for the optical system optimization only the adaptive steady-state genetic algorithm (ASSGA) is used because it is the only genetic algorithm which is well adapted to the optimization of complex technical systems. Contrary to the genetic algorithms, the evolution strategies are designed to be the optimization method for complex technical systems. All evolution strategies are very well suited for the optical system optimization. The following evolution strategies are implemented in the optical system optimization:

• the two membered evolution strategy ES EVOL which represents the most simple model for the evolution simulation with only two members: a parent and an offspring and with mutation as only one genetic operator;

• the multimembered evolution strategies ES GRUP, ES REKO and ES KORR which are further developments of evolution simulation from the first evolution strategy ES EVOL. The ES GRUP is developed from the ES EVOL and it introduces the population that is larger than two members but keeps mutation as an only genetic operator. The ES REKO is developed from the ES GRUP and it introduces the recombination as an genetic operator. This means that the ES REKO has two genetic operators: the mutation and the recombination. The ES KORR is a final and the most developed simulation of evolution, which has several genetic operators.