The Research Of Action Selection And Its Evolution Of Artificial Life

Artificial Life Science has become a focus of Information Science, Life Science and System Science in the world now, which now is a rapidly developed science. To research intelligent artificial life and intelligent agents, it is very essential to solve the problem of action selection of artificial life. Only if the agent of artificial life can choose an appropriate action by itself in the environment, then it can be called as intelligent artificial life. Therefore, the research of action selection of artificial life is significant. On the basis of investigation on the action selection of artificial life and intelligent agent, we provide a model which is about the action selection and its evolution of artificial life. We also construct the simulation system. The main works and conclusions of this paper are shown as follows:(1) We provide a model of action selection based on priority degree. The model can keep the merits of original priority model, at the same time, we can simulate the animal's action selection more vivid.(2) We also provide a new action selection evolution model which is based on Genetic Algorithms. The results of simulation show that the model can get good results with high efficiency.(3) We set up a simple simulation system on action selection of artificial life. It is the basis for realizing the model of action selection and its evolution. The system includes:a) We establish the object of computer artificial life with the properties of real life such as growing(aging), responding the stimulation, exchanging of energy and generating. b) We design the program of artificial life's actions such as eating, escaping and wondering.