An Improved Genetic Algorithm of Suppressing Premature Convergence

Genetic algorithm has been successfully applied to a wide variety of problems.However,there are still some shortages.Considering premature convergence of the genetic algorithm,we proposed an improved genetic algorithm based on analyzing simple genetic algorithm′s genetic operator and control parameters.The improved genetic algorithm has used real number coding,arithmetic crossover operator,no-nuniform mutation operator,and selected the control parameters reasonably.In the early stage the improved genetic algorithm can search the solution space evenly,in the latter part can be more subtle local search,so individuals can enter,groups gradually focus the most advantage of the domain of attraction under certain conditions,and thus to prevent the premature convergence of the genetic algorithm.Theory analysis and example results show that the improved genetic algorithm performance is superior in a number of simple genetic algorithm,avoided genetic algorithm′s premature convergence well,raised the evolution efficiency of genetic algorithm,and has good validity and feasibility.