Research On The System Of Nonlinear Observer Design Method For AUV Control

Nowadays,with a high broad prospect in both military and civil applications,autonomous underwater vehicle(AUV)has become a hot research at home and abroad.The expansion in application scope makes it more and more complicated for AUV to carry out tasks and requires higher control performance.Yet,the characteristics of AUV itself,such as strong nonlinearity,strong coupling,model parameter perturbation and interference vulnerable to the external environment(such as waves and currents)will harden the design of AUV control system.Therefore,it is necessary to adopt efficient control strategies to meet the demands for high precision control performance of AUV.As an important part of the control system,nonlinear observer is an effective method to improve the nonlinear,coupling and disturbance the system robustness.It can not only observe the unpredictable state information according to the effective measurement information,but also can get the observation for environment interference and filter the high-frequency measurement noise.This paper mainly has a deeply research on the design of nonlinear AUV observer for constant speed,direction and depth control.Firstly,we have a deeply study on the dynamics and kinematics model for AUV.The AUV mathematical model with six freedom degrees is established,including current and environmental interference.Then,the mathematical model is decoupled into the two slight coupling subsystems in horizontal and vertical surface mode.And the simulation confirms the correctness of the coupling horizontal and vertical surface model.Secondly,a nonlinear observer called nonlinear extended state observer(NESO)is introduced in details which does not depend on the standard model of controlled object.Based on the AUV horizontal model,we study the NESO design in speed and heading control.NESO can realize the observation for position,velocity and total interference of the controlled object.Therefore,it can improve the robustness of controlled system on the interference of environment through disturbance compensation.The simulation experiment validates the availability of the design of NESO for AUV constant speed and heading control system.Thirdly,a nonlinear observer called nonlinear luenberger state observer(NLSO)is introduced which depends on the standard model of controlled object.We have also briefly analyzed its stability.NLSO can reconstruct the states with effective position information so as to estimate the external environment effecting on the system.What's more,it can also reduce the high measurement noise introduced by the measurement elements.The function ofthe designed NLSO is verified by the simulation experiment of the AUV motion model of horizontal control.Finally,we study the design of a nonlinear AUV observer for motion controlling in vertical surface.The AUV vertical surface model is simplified to a depth control model.Based on the model,we analyze and verify the deficiencies of NESO.At the same time,the NESO is improved based on the application of saturation function for interference estimation in NESO.The NLSO and NESO are combined together to constitute the nonlinear observer for AUV in deep control which compensates for their deficiencies.Simulation verifies that the NLSO-NESO method can meet the design requirements and is conducive to improve the dynamic performance of the control system.