| In present study, Differential evolution, an evolutionary type of stochastic global optimum design search method was implemented using Matlab programming.; Parameters affecting this method and the effect of each parameter on the result and number of function evaluations were found out by numerical experiments. What should be the convergence criteria and how efficient different convergence criteria are, was checked with numerical experimental results. Efficiency of Differential evolution in solving nonlinear global optimization problems was checked using difficult global optimization problems with many variables.; For practical application of using Differential evolution to solve design optimization problems different methods of constraint handling was combined with Differential evolution and checked using nonlinear objective function with nonlinear constraints. Constraint handling method, which does not use penalty parameters and handling constraints using principle of Differential evolution, was developed.; Original algorithm of Differential evolution was modified at some steps to handle discrete design variables. The capability of Differential evolution was demonstrated by solving design problems with continuous and discrete design variables. |
