Development of Underwater Robot Hand and Its Finger Tracking Control

Manipulator is the most important part for underwater flexible robot manipulation, and the grasping and manipulating capability is an important indication for intelligent robot. But current generation underwater robots generally employ the claw-like end effectors, which is simple and easy to control, but not able to perform complex operations. In this paper, a non-anthropomorphic underwater robot hand is presented. The characteristics of the hand, transmission mechanism, corrosion prevention and seal, sensor system, and control system are described and discussed. A fuzzy CMAC was adopted to establish the tracking control system for improving the fine performance capacity. The neural network structure of the fuzzy CMAC was designed, and the input and output formula of each layer was deduce. Based on BP arithmetic, the associative strength, core and width of Gaussian membership function were obtained through training. In final, the simulation of the finger tracking control system was carried out, the results indicate that the finger can follow the scheduled track, the stability is good, the position precision is high, and the control system can be used in the underwater robotic hand. In brief, this hand is capable of grasping a wide variety of objects of different shape, size and weight, and the work area of the underwater robot will expand when this hand cooperates with the robot arm. It can be applied widely in the field of exploiting the resources under the sea.