Improvement Of Genetic Algorithm And Its Application

The genetic algorithm(GA) is a kind of searching method using probability which simulates the natural evolution. It is simple and easy to implement, especially it does not need the special field knowledge and the search only dispend on the fitness, so it has been used in many broad fields. Now the genetic algorithm has got a lot of fruits and more scholars began to pay attention to it.The genetic algorithm which was presented by Holland in 1960s and generalized to standard GA in 1980s is becoming a popular domain increasingly in the middle of 1990s. GA has the ability of global search and is easy to implement. Initially, it is used for nonnumeric computing and then in optimization domain recently.The genetic algorithm is still a new developing technology. Despite its success in many domains, its theoretical fundament is relatively weak. There are still lots of problems to be studied and improved. The main works of the paper are as follows:Firstly, briefly introduce the standard genetic algorithm in the general situation of the development, basic concept, basic principle, rationale, convergence, character and its application. The paper summarize the realization technology about the standard genetic algorithm. Introduce the problem of curve fitting and its basic methods, the mathematical model of the traveling salesman problem (TSP) and its traditional methods, the basic theory of Partheno Genetic Algorithm and solution of TSP based on Partheno Genetic Algorithm.Secondly, according to the shortage of simple genetic algorithm in solving Logistic curve fitting: easy to produce premature convergence, easy to fall into local optimal equilibrium states, poor efficiency at evolutionary late stage. This paper presents an improved genetic algorithm. Dynamic adaptive strategy is introduced to adjust the crossing probability and mutating probability. Numerical results illustrate that the algorithm is feasible and effective.Finally, according to the mathematical model of the traveling salesman problem (TSP) and its solution based on Partheno Genetic Algorithm, This paper presents an improved Partheno Genetic Algorithm (ImpPGA). It puts forward a new genetic operators: untwist operator. This algorithm simulates the recurrence of nature evolution process, while providing fewer control parameters. Experiments prove that it can reach the satisfying optimization at a faster speed.