The Study On Improved Genetic Algorithm And Its Application In Optimization Questions

Genetic algorithm is one kind of organic and adaptive random searching algorithm thatsimulates the evolution process and mechanism in nature to solve the optimization problems.It express the process of solving a problem as the process of survival of the fittest of thechromosome using for reference the idea of "survive the superior and eliminate the inferior,survival of the fittest" in Darwin's nature selection and the genetic mutation theory ofMendel. It makes the population search the best acclimation individual by the incessantevolution to the chromosomes, including the selection operator,the crossover operator,themutation operator and so on, and then gain the optimal solution or satisfactory solution. Genetic algorithm is a highly all-purpose optimal algorithm. Its coding technique andgenetic operator is more simple. It has few requests to the condition of constrain in anoptimization problem. It is very good at parallel and global searching. It can solve a greatmany actual questions and it has applied well in many fields such an machine learning,pattern recognition,image disposal,optimization controller,combinatorial optimization ,manage decision-making and so on. Despite genetic algorithm has been applied well in many fields, after all it is an newsubject, its theory and method still have crudeness and itself has several shortages to beameliorated and consummated ulteriorly. At present, the study of genetic algorithm mostlyconcentrates on two aspects as follow: (1)The improvements on genetic algorithm design and executive strategy. Theseimprovements mainly concentrate upon two aspects: one is done to each factor ( includecoding method,fitness evaluation method,selection operator,crossover operator andmutation operator) of the algorithm in order to search the optimal solution or satisfactorysolution at a rapid pace; the other one is done to its executive strategy ,which combine thegenetic algorithm and other optimization methods such as simulated annealing algorithm,enlighten algorithm,parallel algorithm and gradient method to enhance its convergent rateand its ability of searching the optimal solution. (2)The study in genetic algorithm elementary theory. Although we have much theorystudy in genetic algorithm and we also can provide some theory to prove it could search theglobal optimal solution, genetic algorithm still has faultiness in theory. There are manyproblems still have not settled well, for instance the convergence rate estimator ofmultifarious improved algorithms have not been done well. In this paper, the improvement to genetic algorithm and the application of genetic -- 53algorithm in actual fields have been done . Firstly, the theory of genetic algorithm is analyzed and summarized. In this paper, thefactors of genetic algorithms are analyzed synthetically those including coding method,genetic operator,fitness evaluate method and the parameters of genetic algorithm; Someindexes for estimating the capability of genetic algorithm are given and some shortages ofgenetic algorithm are pointed out ; The convergence of genetic algorithm is provedcombining the model of Markov; Some familiar hybrid genetic algorithms combininggradient method,hill climbing method,simulated annealing algorithm which are morefeasible in local searching are given. Secondly, some improvements are done to genetic algorithm using for reference the ideaof niche and the adaptive idea, according to the shortage of getting in local minimal easilyand convergence early in genetic algorithm; the solution of improved algorithm and that ofsome improved algorithms former are compared, the genetic algorithm presented in thispaper are more easily to search the optimal solution or satisfactory solution than that informers'. Thirdly, the estimate model of surface water environmental quality is set up according tothe "surfa...