Research On Shared Control Of Autonomous And Remotely Operated Underwater Vehicle

With the development of underwater vehicles,a new type of autonomous and remotely operated underwater vehicle(ARV)which combines the features of autonomous underwater vehicle(AUV)and remotely operated vehicle(ROV)comes up.ARV integrates the advantages of the AUV and ROV,it could perform underwater jobs in autonomous,semi-autonomous and remote-operated ways,respectively.According to the features of ARV,shared control modes and methods of ARV should be investigated.The thesis is supported by a Chinese high-tech research and development program(863 program)-Arctic Pole Autonomous Remotely operated Vehicle,in which systematic research efforts have been done on shared control of ARV.The main contents of this thesis are as follows:1)First,the concept of ARV is elaborated,and the literature on recent developments and researches of ARV both in China and abroad is reviewed.And then,some researches and applications of shared control in the field of robotics are presented and discussed.2)Three main control system architectures primarily employed in underwater robots are described,and their advantages and disadvantages are analyzed.And the shared control system architecture that is applied in the robotics field is also analyzed.Finally,based on the features of shared control of ARV,a shared control system based on the MOOS(Mission Orientated Operated Suite)is put forward,and its operating principle is designed.3)Based on the analysis of requirements of shared control of ARV,two Shared control modes are proposed,which are shared control of arbitration type which switches between autonomous and remotely-control,and shared control type with real-time fusion feature,respectively.The real-time fusion shared control includes two methods,one is a shared control method based on artificial potential field,and the other one is a shared control method based on behavior-based control.In the shared control method of arbitration,the working principle and working process of this method is designed.In the shared control method based on artificial potential field,an operator's control instruction is abstracted to a vector of ARV in the potential field,with this method,the intelligence of operator and the ARV's machine intelligence integrate in real time.In shared control method based on behavior based control,an algorithm based on efficiency-weighted is proposed.In the algorithm,the control weights and efficiency in the process of control are measured and calculated in real time.Shared control can be used in many application scenarios during the process of ARV working.This thesis mainly focus on researching specific algorithm based on the important and classic applications of obstacle avoidance and navigation,underwater robot,to testify the shared control method and algorithm proposed in this thesis.4)Based on the characteristics of the proposed ARV shared control algorithms,a computer simulation environment which is suitable for ARV shared control study is developed.Then simulation evaluation of the proposed control methods and algorithms are performed on the simulation environment.And Qualitative and quantitative analysis of the simulation results are carried out,which validate.the proposed methods and algorithms in this thesis and demonstrate their superiority.