Various Mutation Strategies in Enhanced Competitive Differential Evolution for Constrained Optimization

Application of differential evolution ( DE) to the constrained optimization problems is addressed. A self-adaptive competitive variant of DE with enhanced search for feasibility region is used. The influence of different kinds of mutation (rand/1/, randrl/1/, current-to rand/1/, and a search strategy combing two types of mutation) on the performance is studied. Experimental comparison of four novel variants of competitive DE is carried out on the benchmark set developed for the special session of IEEE Congress of Evolutionary Computation (CEC) 2010. The experimental results showed that tested variants of competitive DE perform almost equally and there is no significant influence of the kind of mutation on the overall feasibility rate. Some tested competitive DE variants performed equally or even better when comparing with the best algorithms of CEC 2010 competition in several benchmark problems.