Research On Heading Control And Visual Simulation Of Unmanned Surface Vehicle

As an intelligent marine vehicle,Unmanned Surface Vehicle（USV）can complete autonomous navigation and planning,and perform a variety of tasks.Compared with conventional ships,USV has excellent maneuverability due to its small size,which has a wide range of applications.The motion control is the prerequisite for USV to achieve autonomous navigation,and heading control is the basis of the entire motion control,the research of heading control is of great significance.This dissertation takes the problem of heading control of USV under the interference of complicated marine environment as the research object,and researches on the design and implementation of visual simulation system and heading control algorithm,the specific research work is as follows:Firstly,in order to solve the heading control of USV under the steering of large rudder angle,a second-order nonlinear response model is established to consider the wind wave current environment interference,the particle swarm optimization（PSO）algorithm and fuzzy control algorithm are respectively used to realize offline tuning and online self-tuning of proportional-integral-derivative（PID）controller parameters.Through the computer simulation platform in the laboratory,the PID,PSO-PID and fuzzy PID are used to simulate the heading control respectively.The results show that the fuzzy PID and PSO-PID have higher control accuracy and stronger antiinterference than the PID controller,both can effectively improve the effect of PID control.Secondly,a radial basic function（RBF）neural network is used to design the heading controller of USV,and then the stability of the control system is proved by Lyapunov stability theory.When the RBF neural network structure is determined,the selection of network parameters has a great impact on its performance.Aiming at the problem that the center,width of the hidden layer nodes and connection weights of the RBF neural network are not easy to identify,the Beetle Swarm Optimization（BSO）algorithm is used to train the parameters of the RBF neural network.Compared with PSO,the convergence speed of BSO is faster and the optimization result is better,effectively improve the learning speed and performance of the RBF neural network.Finally,the effectiveness and superiority of the proposed improved RBF neural network control algorithm are verified by simulation.Finally,the visual simulation system for USV heading control has been designed and implemented.Unity3 D combined with 3DS MAX is used to develop a virtual USV sailing scene,UGUI is used for the design and production of the interaction interface,using Visual Studio 2017 to write heading control algorithm program,the control parameters and USV model parameters can be flexibly modified online.The visual simulation system for USV heading control has good real-time interactivity,realize the simulation of the attitude of USV in the process of heading control,further verify the feasibility of the designed heading control method.The designed and implemented visual simulation system can be used to reduce the cost and risk of research and simultaneously improve the research efficiency of USV and the authenticity of simulation.