Research On Motion Control Technique Of An Open-frame Underwater Vehicle

General Detection Remotely Operated Vehicle (GDROV) is a kind of remotely operated underwater vehicle designed for the inspection of cracks, crevices and other potential problems of the dams. It is designed as open-frame structure for the reason of being equipped with many different advanced inspecting sensors according to different tasks, which results in the difficulty in acquiring the precise model of it. Because of the particularity of open-frame structure, GDROV does not have streamlined geometry profile, so it is more difficult to control precisely. The problem has not yet been solved very well.In this dissertation, we aimed at the application of fuzzy neural network control in the motion control of the GDROV and designed the control system for the GDROV.According to the moving characters and the demand of the underwater vehicle, four-degree of freedom equation was set up at first, improving the simulation environment of GDROV. Then fuzzy neural network controller based on desired state programming was presented in this dissertation, which used the technique of fuzzy neural network. Finally, simulation experiments were conducted on GDROV. The results show that it has biggish improvement in response and precision, which merges the merits of traditional fuzzy control and neural network control, and has good robustness to the model's uncertainty and the disturbance. It has theoretical and practical value.