Path Tracking Control Based On Adaptive Sliding Mode For Unmanned Surface Vehicl

With the application and development of the unmanned equipment at sea,the unmanned surface vehicle(USV)has been listed as an important military and civilian project by many countries because of its obvious advantages such as small size,high sensitivity,and capability of working in rugged environment.The motion control system of an USV is the core part of its research and development,and also the basic premise of completing various tasks.Therefore,this topic studies the path tracking control problem of an underdriven USV in disturbed environment.The major study contents include:(a)The characteristics of the underactuated USV and the complexity of the operating environment were analyzed,and various motion models of the underactuated USV were established,including the global hydrodynamic model,the separate hydrodynamic model and the response model,which laid the foundation for the design of the control algorithm in the future.(b)The course control algorithm of an USV is designed and implemented.Considering the comprehensive disturbance of the system,two kinds of course controllers based on adaptive sliding mode control are studied.Firstly,the model is optimized based on the USV’s holistic model,and the corresponding wave compensation method is raised.For the unknown disturbance of the system,a course controller based on parameter self-adjustment and a controller based on radial basis function(RBF)neural network are presented.Finally,the stability theory analysis and system simulation of the two control systems are carried out respectively,and the validity of the two control strategies is proved.By comparing the two algorithms,the control method based on neural network is more adaptive and intelligent.(c)The path tracking control algorithm of an USV is studied and an adaptive sliding mode controller based on dynamic surface control(DSC)and simplified RBF is proposed.Firstly,the virtual control law of the system is designed by the virtual ship guidance algorithm,and the complicated problem of derivation of the virtual control law is overcome by DSC method.In the dynamic loop,an adaptive sliding mode controller based on simplified RBF is designed.The controller approximates the uncertainty of the system through RBF neural network,and combines sliding mode control to suppress the comprehensive disturbance of the system.On this basis,instead of the weight vectors of network,the adaptive laws of their norm are designed,and then the calculation of the controller is simplified.Furthermore,the stability of the system is proved,and the great performance of the controller is certified by a case study.(d)To further improve the control performance of path tracking control,an adaptive sliding mode controller based on extreme learning machine(ELM)and barrier function(BF)was proposed.In the kinematic loop,the guidance algorithm based on virtual ship is also selected,and DSC technology is combined.In the dynamic loop,the ELM is adopted instead of the RBF neural network used above to approximate the uncertainty of the model.Furthermore,on the basis of sliding mode control,a switching adaptive law based on BF is designed.The design can estimate the systematic errors and constrain the sliding variable inside the predefined range,without causing the control gain to be overestimated.Then,the stability theory demonstrate that the system is stable.The consequences of simulation prove that the proposed controller is efficient,and it has great performance in robustness and applicability.