Intelligent chaos Fish-Catching based on Neural-Network-Differential-Equation

In this paper we tackle a Fish-Catching task under a visual feedback hand-eye system with catching net. As the time of tracking and catching process continues, and releasing the caught immediately into the same pool, then the fish can somewhat get accustomed to the net motion pattern and gradually find out new strategies on how to escape from the pursuing net. For the sake of such innate ability being widely existed in animal´s behavior, the catching operation becomes tough and some effective intelligent method is needed to be conceived to go beyond the fishes´ intelligence. The purpose of this paper is to construct intelligent system to exceed the fishes´ intelligence in order to track and catch the fish successfully. Then we embed chaotic motion into the net motion of robot for better performance, and we have shown the chaotic net motion can overcome the fishes´ escaping strategies.