Design Of Neural Network Control Course System For Unmanned Surface Vehicle

Unmanned Surface Vessel(USV),as a new type of vehicle,has the characteristics of high flexibility,strong maneuverability,and hidden action.These features can help large ships and users to complete many difficult and dangerous Work has played an important role in ocean exploration and military missions.Considering the complexity and danger of its marine environment and the difficulties of maneuvering and model perturbation of the unmanned boat itself,improving the accuracy of the unmanned boat’s heading control has become a hot issue in current research.Neural network control,as a hot field of current research,has the advantages of self-adaption and self-learning,which have certain inspiration and influence on the design of control systems.By designing a neural network control algorithm to study the course control of unmanned boats and constructing and solving Lyapunov inequality according to the LDI equations output by the neural network,and then judging the stability of the neural network.And based on different sea conditions,perturbations,etc.,the design and simulation of heading neural network control system.First,the mathematical model of the plane motion of the unmanned boat is built and the corresponding nomoto response mathematical model is given.Combined with the principle and design requirements of the steering gear,a nonlinear steering gear response model was built,and the mathematical model of the external disturbance of the unmanned boat such as sea breeze and sea waves were established according to the specific environment of its use.It was the design and test of the unmanned boat neural network course control system.The analysis laid the foundation.Secondly,combining the neural network and the control system design,the unmanned boat PID heading control system model was established as a reference,and the offline training neural network method was used to design the heading neural network control system.Control identification system;improve the design of the heading control system by quantifying external disturbances and perturbation of its own parameters.Thirdly,by referring to the current international field of neural network stability theory and the content of Lyapunov function,the output and linearized representation of the unmanned boat heading neural network control system are derived;the Lyapunov inequality is constructed and the linear matrix solution The method to prove the stability of the inequality,combined with the LMI toolbox in Matlab for simulation calculation,gives the results that can prove the stability of the system.Finally,the simulation experiment of the course control of the unmanned boat under complex sea conditions and parameter perturbation is completed;the statistical results of the heading angle and rudder angle under steady state conditions are obtained using statistical principles,and the comparison conclusion is given.Simulation experiment results show that the unmanned surface boat heading control system based on neural network has better control accuracy and robustness.