The Research Of Implement Techniques Of Genetic Algorithms

An overlap between life sciences and engineering is vivid characteristic of development of modern science and technology. It is also a research focus of corresponding fields. People devised genetic algorithm by simulating the genetic and evolution mechanism of biology. GA is mimic-life algorithm in grandness. As genetic algorithm has vivid mathematical character and is suitable for any function, it has broaden application and to study the algorithm has great significance.At present, the studying of algorithm is focus on its application. It has no steadfast theory fundament. This paper will continue investigating the implement strategy and techniques of algorithm. To improve implement techniques through analyzing mechanisms of algorithm.This paper will introduce every step of implement techniques of algorithm in detail, analyze binary coding and gray coding, improve the method of fitness scaling and combine chaos with genetic algorithm. The main works and conclusions are shown as follows:Studying on coding. We contrast binary coding with gary coding and try to find the reasonable explanation of superiority of gray coding. Our view is that gray coding not only has good global searching ability but also good local searching ability.Studying on fitness scaling. Fitness scaling was proposed to fight the premature convergence of GA. In fact, the existent fitness scaling does not fulfill the desire. The paper let it be more reasonable.Studying on algorithm improving. Chaotic sequence is a complete autonomic process after its iteration. It is more similar to the evolution process than the stochastic process. We obtained chaos-genetic algorithms by using chaotic variable in initializing group, crossover operator, and mutation operator of genetic algorithms. The computation results indicate that CGA has good computational efficiency in function optimization.