Research On The Method Of Motion Control For AUV

With the development of the activities in the sea, the application of Underwater Vehicles is widespread. However, owing to the nonlinearity and the unpredictable operating environment of Underwater Vehicles, many factors must be taken into consideration during the design of the control system of Underwater Vehicles.Up to now, intelligent control has been applied widely in the uncertain and nonlinear system. In this thesis, we aimed at the application of PID and neural network technology in the montion control and try to design a better system for Underwater Vehicles.In this dissertation, dynamic model of AUV is built, and based on the need of practice and the convenience of designing the controller the model is properly simplitied into the dynamic model with 6 directions of freedom (DOF). Basing on this, a controller with PID of Back-Propagation Neural Networks is presented. Then the structure, learning method of the PID and multilayer Back-Propagation Neural Networks Control were provided.This dissertation applies PID control method of Neural Networks to AUV. Combined with the dynamic model with 6 DOF, PID controller and PID controller of AUV based on Back-Propagation Neural Networks has been set up and carried on simulation experiment. The simulation showed that the methods are adapted to the control of the AUV.