| Ocean monitoring is the basis of ocean exploration and exploitation,in which the autonomous underwater vehicle(AUV)has become an indispensable and important equipment.The spatial distribution and time parallelism provided by multi-AUV cooperation can greatly improve the efficiency of ocean monitoring tasks,and increase the response speed to the task requirements and hostile target intrusions,which makes it a hot research topic in the field of underwater robotics nowadays.The flow field is an important parameter that affects the marine ecological environment and the trajectory of the AUV,and the target monitoring is closely related to practical applications such as marine confrontation and disaster rescue.Flow field estimation and underwater target enclosing as representative application scenarios in ocean monitoring by multiple AUVs have important research significance.However,the particularity of the marine environment brings new challenges to the theoretical study of cooperative estimation and control.Therefore,this thesis conducts a series of researches on two key issues for cooperative flow field estimation and distributed target enclosing by multiple AUVs.The main research contents of this thesis are given as follows:(1)Considering the high complexity of the flow field and the lack of direct measurements,a cooperative flow field estimation method based on motion-integration errors is proposed.Firstly,the concept of the motion-integration error is defined as the trajectory deviation affected by the flow field.We establish error constraints between the unknown discrete flow field and unknown trajectories by domain gridding and the flow field estimation problem is transformed into a problem by solving these constraints.Then,a row-iterative cooperative estimation algorithm is designed,which does not require measuring the local flow velocity to reconstruct the flow field.And the coordination of relative position between neighbors is introduced to achieve the cooperative flow field estimation.(2)Considering the relative position measurements required in the domain gridding method are insufficient,a cooperative flow field estimation based on the parametric layered model is proposed.We utilize the physical properties of the flow field to establish the smooth flow field as a parametric layered model.It incorporates the concept of incompressibility and the characteristic of negligible vertical flow velocity to provide a physical property,which relaxes the limitation on the movement of the vehicles.Then,we establish the relation between the flow model and unknown trajectories.To reconstruct the smooth flow field,an iterative algorithm is designed to estimate the model parameters.(3)Considering the unknown absolute positions of AUVs and some vehicles cannot detect the target,a target enclosing control method based on local information is proposed.Based on the kinematic and dynamic models of multiple AUVs,we propose a cascade-based distributed control law,which uses local relative position measurement to realize a desired circumnavigation formation around the static target.Then,the distributed auxiliary system is added to realize the circumnavigation of multiple AUVs around the moving target.Furthermore,this control law enables all vehicles to achieve the target-enclosing formation through the radius determinant mechanism and the appropriate controller parameter selection.(4)Considering trajectories of AUVs are affected by the flow field during the underwater movement,a multi-target enclosing controller based on auxiliary variables is proposed.To extend the single-target enclosing control to the multi-target situation,the proposed controller allows the vehicle team to handle a group of targets with coordination.The designed auxiliary variables are introduced into the proposed controller to eliminate the influence of the flow field on trajectories,and analyze the curve of the designed auxiliary variables.The system stability has been proved in detail and through the proposed method,the multi-target enclosing formation under the flow field is achieved.The proposed methods solve the problems of cooperative estimation and control in the marine environment from multiple perspectives,and are expected to lay the necessary theoretical foundation for ocean monitoring using multiple autonomous underwater vehicles. |
