The Improvement Of Genetic Algorithms And Its Convergence Analysis And The Applicationins Network Security

Genetic Algorithm is a kind of searching method which simulates the nature evolution. It is simple, easy to implement and has strong robustness. Especially it doesn't need special field knowledge, but only need to use a self-adapting function as evaluation to instruct the searching process. As a result, its application area is very broad .It has been practically applied in numerous fields and has acquired notable results. So great number of scholars and people working in engineering field have played much attention on it Genetic Algorithm is a new developing technology .Despite its success on application domains, its theoretical fundament is relatively weak, there are still lots of problems to study and improve.This paper puts forward a kind of improved genetic algorithm on the basis of researching and analyzing present genetic algorithm development conditions and leads the thought of the greedy algorithm into the genetic algorithm, in the meantime, carries on a comparison to the improved genetic algorithm, the classic genetic algorithm and the greedy algorithm. The result is the improved genetic algorithm having excellent optimization characteristic, in the meantime the thesis elaborates on the convergence problem of classic genetic algorithm in detail and carries on a comparison to the convergence characteristic of improved genetic algorithm and that of classic algorithm proving that the searching process of this improved genetic algorithm is a limited hour, identical order and throughout Markov chain, in the meantime making use of the Markov chain theories to carry on a research on two kinds of algorithms, giving out the integrated proving method of convergence characteristic of classic genetic algorithm and that of improved genetic algorithm ,obtaining improved genetic algorithm converging on the overall optimized solution with 1 of probability, resolving the convergence problem of the classic genetic algorithm.In the mean time, the improved genetic algorithm is applied to the optimization of invading detected system in network safety. According to the present development conditions of invading detected system, a kind of acting more for distributed invading detected system is put forward. The idea of coordination detection is put forward in the process of system design, leading that the detection system improves detecting property obviously and applying improved genetic algorithm to optimize the process of coordination detection, obtaining excellent optimization results.