Design and Implementation of Perception System Based on Self-Learning for Artificial Fish

In "artificial fish",computer animation is created with artificial life.To study Self-Evolutionary of animation role.A BP neural network based on Time Difference (TD)-TDBP is presented,and artificial fish can automatically learn in environment.Perceptive area and visibility of artificial fish are studied,and perception system is designed using the technologies of pattern recognition.The experimental results have shown that perception system of artificial fish based on self-learning provides artificial life with more independence and intelligence.