Population size for EvolutionaryAlgorithms is usually an empirical parameter. We study the population size from aspects of fitness landscapes’ ruggedness and Probably Approximately Correct (PAC) learning theory.

The evolutionary approach to multiple function optimization formulated in the rst part of the paper [1] is applied to the optimization of the low-pressure spool speed governor of a Pegasus gas turbine engine. This study illustrates how a technique such as the Multiobjective Genetic Algorithm can be applied, and exempli es how design requirements can be re ned as the algorithm runs. Several obje...

There is a lot of experimental evidence that crossover is, for some functions, an essential operator of evolutionary algorithms. Nevertheless , it was an open problem to prove for some function that an evolutionary algorithm using crossover is essentially more eecient than evolutionary algorithms without crossover. In this paper, such an example is presented and its properties are proved.