This paper introduces the implementation of a fuzzy control based intelligent robot fish. The robot fish can swim autonomously or be controlled through remote control. Possible actions of the robot fish by remote control include swimming forward, turning left, turning right, sinking, and rising. It also includes automatic electrical power sensing functions. When electricity is low, the robot fish will surface to recharge electricity. Furthermore, the robot fish can perform some independent functions, such as obstacle avoidance and target tracking. A camera inside the fish can monitor underwater surroundings and also track targets in the water.
　　This work uses two control methods, intuitive control and fuzzy control, to avoid obstacle and track targets, and then compares both of them by computer simulations and practical experiment. The comparison results show that fuzzy control fish performs better than the intuitive control fish. In order to prove the adaptability of the robot fish to different environments, the robot fish was tested not only in the pool of the laboratory, but also in various outdoor environments. The results show that the robot fish can be operated in all of the tested environments.