System Modeling And Simulation Technology For Underwater High-Speed Vehicle

Supercavitating vehicle generates cavitation around its body to reduce the resistance,so as to realize ultra-high-speed underwater navigation. Because of the significant impacton the development of underwater vehicle and navy equipment, many countries areinvesting hugely in the research of supercavitating technology. This paper reports the studyof system modeling and simulation technology for underwater high-speed vehicle.First, according to its navigation characteristics, configuration scheme of theunderwater high-speed vehicle is designed and generation requirements of the cavitation arestudied. By introducing the basic coordinate systems, kinematic model of the vehicle isestablished. Then, detailed analysis of forces applied on the vehicle is provided, includingcavitator forces, sliding force and thrust, so as to build the kinetic model of the vehicle.Based on Newton’s Laws and kinetic theory, non-linear equations of the underwaterhigh-speed vehicle are implemented. Subsequently, the mathematical models are linearisedby introducing corresponding assumptions. Finally, simulation is completed with the help oflinear depth control models. Simulation results indicate that the derived models are able toreflect related motion principles of the vehicle.Secondly, introduction of basic process and methods of system identification isprovided. By applying methods of system identification, the depth control model ofunderwater high-speed vehicle system is established. This is accomplished by applyingexperimental data preprocessing, model structure and parameters identification, as well asmodel validation. Besides, the validity of the model is verified in the simulationenvironment with different inputs.Thirdly, the conventional PID control algorithm and BP neural network are combined,so as to design the neural network self-adaptive PID control algorithm, and a set of betterPID control parameters is achieved by online adjustment in the simulation environment. Itcan be seen from the simulation result that the neural network self-adaptive PID controlpossesses relatively ideal dynamic response characteristics, and it is able to well achieve thegoal of fixed depth control.Finally, structures and working principles of high-speed underwater vehicle semi-physical simulation system are analyzed, and the simulation software is completedaccording to system requirements. Besides, spectrum analysis is presented according to thedata obtained in the case of open-loop simulation, so as to determine whether the selectedsampling interval is appropriate. In addition, fixed depth control model and simulationidentification model are added to the system to make comparison, and the effectiveness ofthe model is verified. Thereafter, fixed depth control experiments are carried out, whereconventional PID control algorithm and neutral network self-adaptive PID algorithm areadded during the digital simulation, and different interference is also added to verify that thelatter algorithm has a better anti-jamming capability. Last but not least, component parts ofthe vehicle are verified, including controllers and sensors of the system. To achieve thepurpose of fixed depth control, experiments are implemented by adding external devicesone after another. Experimental results demonstrate that the semi-physical simulationsoftware can successfully realize the desired functionality, and the design is reasonable andeffective.