Path-Following Control of Autonomous Underwater Vehicles Subject to Velocity and Input Constraints via Neurodynamic Optimization

A design method is presented for path-following control of under-actuated autonomous underwater vehicles subject to velocity and input constraints as well as internal and external disturbances. In the guidance loop, a kinematic control law of desired surge speed and pitch rate is derived based on a backstepping technique and a line-of-sight guidance principle. In the control loop, an extended state observer is developed to estimate the extended state composed of unknown internal dynamics and external disturbances. Then, a disturbance rejection control law is constructed using the extended state observer. To bridge the guidance loop and control loop, a reference governor is proposed for computing optimal guidance signals within the velocity and input constraints. The reference governor is formulated as a quadratically constrained optimization problem. A projection neural network is employed for solving the optimization problem in real-time. Simulation results illustrate the effectiveness of the proposed method for path-following control of autonomous underwater vehicles subject to constraints and disturbances simultaneously in the vertical plane.