Research On UUV Path Following Control Method For Depth-keeping Cruise Operation

With the increasing demands of countries around the world in marine exploration and resource development,the research interest in underwater intelligent and unmanned equipment continues to rise.Unmanned Underwater Vehicle(UUV)has the advantages of simple operation,superior performance and high safety,and has become an underwater equipment that has attracted much attention from academia and industry.UUV can realize military and civilian tasks such as underwater intelligence search,anti-submarine warfare,environmental monitoring,and resource exploration.Depthfixed cruise is one of the basic operating conditions for UUVs to perform the above tasks,but UUVs are easily affected by obstacle constraints,external environment interference,model uncertainty,and physical constraints of actuators under this working condition.Accuracy and robustness put forward higher requirements.Therefore,it is of great significance to study the path tracking control of UUV under depth-keeping cruise conditions.The main work of this paper includes:(1)Aiming at the sudden change of UUV path tracking control input caused by obstacles,an adaptive obstacle avoidance tracking control method based on barrier lyapunov function is proposed.First,considering the influence of static and dynamic obstacles on UUV navigation safety in the marine environment,a mathematical model of obstacles is established,and the navigation danger zone,collision avoidance zone and controlled safety zone are defined.Secondly,a UUV collision risk assessment model is established according to the minimum meeting time and minimum meeting distance between obstacles and UUVs.Then,the obstacle avoidance controller is designed by introducing the obstacle Lyapunov function and backstepping method.The weight parameter of barrier function is constructed by the collision risk to meet the UUV obstacle avoidance and stability requirements.The simulation results show that the method proposed in this paper can effectively avoid collision in the case of encountering static and dynamic obstacles,and to a certain extent reduces the sudden change of the control quantity caused by the gradient change of the barrier lyapunov function.(2)Aiming at the low tracking accuracy and slow convergence speed of UUV caused by environmental disturbance and model uncertainty,a backstepping control method based on Linear Extended State Observer(LESO)is proposed.First,based on the nominal UUV mathematical model,LESO is designed to achieve online estimation of interference.On this basis,a path tracking controller is designed by combining the virtual guide method and the backstepping method,and an expanded state drift angle observer is designed to ensure the convergence of the heading error.Secondly,the dynamic surface technology is introduced to design the dynamic controller,and the unknown part of the system in the controller is estimated and compensated by LESO.Finally,based on the Lyapunov stability theory,it is proved that the error of the closedloop system is uniform and eventually bounded.The simulation results show that under the influence of environmental disturbance and model uncertainty,the method proposed in this paper has higher control accuracy and stronger anti-disturbance performance.(3)Aiming at the performance degradation of UUV path following control under input constraints caused by physical limitations of actuators,an adaptive anti-saturation control method based on orthogonal neural network is proposed.This method takes into account the influence of the physical limitations of the actuator on the performance of the control system,and establishes a UUV dynamic model considering the dynamic characteristics of the actuator.Based on the backstepping method,a tangential sigmoid function-based following differentiator is introduced to deal with the sudden change of control input caused by the path,and the unknown dynamic parameters of the actuator are estimated by the adaptive law.Secondly,an anti-saturation compensator is designed based on an orthogonal neural network,and it is proved that all errors of the closedloop system are uniform and eventually bounded.The simulation results show that,under the physical constraints of the actuator,the proposed method effectively reduces the performance loss of the control system when the actuator is saturated.