The Application Of Artificial Neural Network And Genetic Algorithm In The Hull Structural Dynamics

With the developing of the science and technology, the performance of structure is mor< important than before. Especial in naval architecture, it 's very necessary to analyze th< dynamic capability in design phase.Artificial neural network (ANN) simulates the human brain's function. It can self-learn self-organize and self-accommodate. It's a new powerful tool to forecast in many fields.Genetic Algorithms (GA) simulates the biology evolution. It can find the best resul without others control. It's a new method in optimizationThis thesis is divided into four sections. Section one in Chapter 1 gives an introduction tc the significance of this task and the main content of the article.Section two in Chapter 2 introduces the basic principle of Finite Element Analyzed Method, and the way of how to analyze the dynamic question.Section there in Chapter 3 is concern a new method (ANN) to forecast ship global vibration. By comparison to the traditional method, ANN is more useful and convenient. At last, giving an illustration to prove it.Section four includes chapter 4 and chapter 5. A brief introduction is given to basic theories of Genetic Algorithms in chapter 4. In chapter 5, writing a GA program to optimize a structure that consists of beams and shells. The coding language is ANSYS Parametric Design Language (APDL). The program testifies that the GA can be utilized in structural dynamic optimization.