Stopping criteria and mutation strategy based on the theoretical analysis for a class of evolutionary algorithms

In this paper the convergence properties for a class of evolutionary algorithms in continuous space are studied. Under the certain conditions, the convergence theorem is presented with the selection and mutation operators. Based on the convergence analysis, we discuss the stopping criteria. Furthermore, a new mutation strategy is constructed on the foundation of the principle of simulated annealing-like strategy. The numerical simulations are offered to illustrate the effectiveness of the strategy.