Research On Intelligent Control Techniques For Motion Control Of Underwater Vehicles

Presently, intelligent control is the hotspot in the research field of motion control of Underwater Vehicles (UVs), and it is the research direction in the future to a great extent. This dissertation aims at the application of intelligent control techniques to motion control of UVs, and trying to design a control system with nice performance.First, the motion simulation model of the UV is analyzed, including current and wave effects, then the simulation system in the dissertation is constructed on the basis of this model.Fuzzy control, neural network control and expert control are mainly studied. A kind of self-adjusting fuzzy control is studied, by adjusting fuzzy control rules expressed in analytical form through intelligent weight function, not the defect of establishing satisfactory fuzzy control rules through experience is overcame, but the robustness and adaptability of the fuzzy control system are improved. Neural network feed-forward inverse control for the UV is proposed, this feed-forward/feedback compound control strategy trains the neural network feed-forward controller with the instruction of motion will of the UV to be the inverse dynamic model of the UV, it does not need identification of the UV and the stability of the control system is guaranteed. S plane control method is a simple but practical control method for motion control of UVs. Expert S plane intelligent control method is proposed in the dissertation. S plane controller's parameters are adjusted on-line using expert controller for improvement of S plane controller's performance, and man-imitating intelligent integral is introduced for improvement of S plane controller's steady-state accuracy.Controllers of each control method studied are designed respectively, and simulation experiments are carried out correspondingly, which validate the feasibility of the control methods studied in the dissertation.