Analysis And Application Of A Hybrid Genetic Algorithm

Genetic Algorithm is developed by John. Holland who is a professor in Michigan University in United State, based on Darwin's theory of evolution. It is an efficient optimal technology. This paper introduces the elemental theory of the Genetic Algorithm at first, and gives a dedicate analysis to the problem which is found when we use the Genetic Algorithm to optimize the parameter of the PID controller. According to the achievement of predecessor and the application of the Genetic Algorithm optimizing the parameter of the PID controller, we notice that the standard Genetic Algorithm is not a globally optimal technology, and easily fallen in the "the early-maturing phenomenon". In this paper, we gives the strict testify based on the Markov Chain. For solving this problem, we develop some new algorithms of the Genetic Algorithm and use them to the optimal of PID parameter. The result shows that they are better than the standard Genetic Algorithm. But there is another problem on the application of the Genetic Algorithm which is the rate of convergence is too low. The Simplex Optimal method is another optimizing technology which can find a local best solution without the character of function. But there exits also a problem in the Simplex Optimal method which is the result is easier influenced by the initial values. Based on the characters of the Simplex Optimal method and the Genetic Algorithm, we develop some new Genetic Algorithms for the optimizing technology. These new Genetic algorithms combine the classic Genetic Algorithm and the Simplex Optimal method. At last, we use new Genetic Algorithm and the classic Genetic Algorithm to optimize the parameter of PID controller, respectively. Comparing their results, we find that the new Genetic Algorithm is distinctly better than the classic Genetic Algorithm.