Research On Underwater Recovery Of UUV Base On Active Disturbance Rejection Control

With the growth of the world’s population and the shortage of energy on land,since the beginning of the 21 st century,many powerful countries have focused their attention on marine development.The full development and utilization of the ocean must depend on a large number of high-tech marine equipment,and the Unmanned Underwater Vehicle(UUV)has received extensive research and attention due to its unique advantages.After the UUV completes its task,it needs such as maintenance,data exchange and energy supplement.If the UUV can be recycled automatically,its efficiency will be improved.Therefore,automated recycling is a long-term research hotspot and an important research direction.In this paper,a UUV is used as the research object and adopted the fork-carrying-pole recovery mode.In order to meet the control requirements and achieve the anti-interference effect,we divided the recovery process into sevel steps,which is based on Active Disturbance Rejection Control(ADRC).The specific research contents of this paper are as follows:Firstly,by analyzing the shortcomings of PID control,this paper introduces the improvement method of ADRC to the shortcomings of PID control,and then introduces the advantages of ADRC over PID control.By analyzing the influence of each parameter changes in every structure of the ADRC system,the debugging rules of ADRC has been summarized.After that,all of the system structures are combined to form a completed ADRC,and then the separability theorem of ADRC is introduced,and the general form and design method of ADRC are given.Secondly,it is briefly introduced the fork-carrying-pole recovery of UUV and the equipment used: recovery fork,foldable recycling pole of the UUV in the papper.According to the characteristics of fork-carrying-pole recovery the detailed recovery process is designed.The UUV recovery process is divided into three stages then established the fork-carrying-pole recovery strategy of UUV.Thirdly,the decoupled method for the UUV model is designed which accords to the decoupled method of ADRC.Then designs an ADRC controller for the decoupled model,and verify the availability of the designed controller by writing some MATLAB programs.Considering that UUV will be affected by ocean current and near-wall interference in the recovery process,the corresponding interference terms are added into the simulation,and the controller designed before was used for simulation.The simulation results after adding interference prove that ADRC has good anti-interference effect and is a very promising controller.Finally,this paper points out some shortcomings of ADRC in the case of the change of the controlled object is relatively large,and designs a method to compensate ADRC by using neural network.From the perspective of theory and design,the BP algorithm training function using gradient descent momentum and adaptive learning rate is introduced to avoid the local minimum value problem of neural network.According to this idea,an ADRC Controller based on the neural network is designed for the first stage of the horizontal plane in the recovery process.Finally,the neural network compensation is valid by MATLAB simulation.Through the discussion and simulation in this paper,it is verified that ADRC has good robustness and UUV whitch works in a strong interference environment can also be well controlled.It is a kind of controller with great practical value.