Numerical Simulation Techniques Research and Application in Genetic Algorithm Design

ABSTRACT

Numerical simulation techniques are also called computer simulation, which take the computer as a means to study all kinds of engineering and physical problems even natural objective through numerical calculation method and image display. This paper studied the numerical simulation techniques and try to solve two-dimensional convectiondiffusion equation parameter identification inverse problem by the genetic algorithm. Firstly, the finite element method was illustrated to solve the steady problem of two-dimensional convection-diffusion equation before it compute parameter identification inverse problem each time. Subsequently, it can search the best approximate solution from many initial points and obtained the global optimum solution by means of crossover operator and mutation operator. Finally, the paper discussed the computer simulation of GA for solving the inverse problem, and puts forward a new method for solving inverse problem: Genetic algorithm based on the best disturbed iteration. The results of numerical simulation show that the genetic algorithm has the higher accuracy and the quicker convergent speed. And it is easy to program and calculate and is of great application.

Keywords: