Variable Centroid Control for Low Speed AUV Based on Adaptive Neural Network

In order to overcome the low efficiency of rudder at low speed,a variable centroid control method for AUV (autonomous underwater vehicle) is investigated.The centroid of AUV system is changed by controlling the motion of three moving masses on three inertia axes of AUV,thus the hydrodynamic will be changed and then AUV can be controlled.The AUV dynamic model with four rigid bodies and nine degrees of freedom is derived based on the Kane's multiple rigid body systems dynamic theory,and the model lays the foundation for designing AUV variable centroid control system.Then the adaptive neural network control system is designed based on the uncertain MIMO(multiple input multiple output) nonlinear system theory,and the Lyapunov stability theory is used to analyze the stability of system.Simulation indicates the validity of the control method.