A Study Of Heading Control Of Unmanned Surface Vehicle Based On Fuzzy Neural Network

With the rapid development of science and technology,automation and intelligence have become the mainstream direction of the development of various industries.In navigation,there are often dangerous or unmanageable tasks to be performed,so the development of related automation and intelligent technologies has attracted much attention.As an intelligent surface robot,the unmanned ship has the ability to autonomously navigate.And through unmanned surface vehicle,the environment can be sensed and targets can be detected,which can replace humans on dangerous water surfaces to complete important tasks.So the applications of unmanned surface vehicle are widely concerned by countries all over the world.The development of unmanned surface vehicle has experienced the transformation from semi-automatic to intelligent,which has a great space to rise.Therefore,continuous innovation and improvement are needed to better promote the development.The control system of unmanned surface vehicle is very complicated,due to the nonlinearity,uncertainty and external environment of the system,and the control accuracy and speed of the system are required to be very high.Therefore,more intelligent and efficient control methods must be found on the basis of traditional heading control of unmanned surface vehicle.The research content of this article is as follows:(1)The mathematical model,hardware implementation and communication protocol of unmanned surface vehicle are introduced,and the system of unmanned surface vehicle is designed;(2)Fuzzy neural network(FNN)combining neural network and fuzzy logic is proposed to tune the PID parameters,but because of the slow convergence rate of the gradient descent method used in back propagation and the relatively complex network structure,The rapidity of control system is affected,so Fletcher-Reeves conjugate gradient method(FR)is used to replace the gradient descent method,so that the convergence speed of the network is optimized;(3)The fuzzy neural network model optimized based on FR method is established,the PID control parameters are adjusted by the optimized fuzzy neural network algorithm,and the heading controller is designed;(4)The navigation environment is complicated,so wind,wave and current factors are added to compare and analyze the anti-jamming capability of the heading control system.The control performance and anti-jamming capability of the fuzzy neural network heading controller optimized based on the FR method have been obviously enhanced.In the actual ship test phase,the remote end of the B/S architecture is introduced,and finally the heading control experiment is completed and analyzed.Through simulation experiments,the control performance and anti-interference ability of the unmanned surface vehicle heading control based on fuzzy neural network proposed in this paper have been significantly improved.And to a certain extent,it has promoted the development of unmanned surface vehicle heading control and provided theoretical support for the integration of unmanned surface vehicle with other disciplines and related theories.