Research On Formation Control Of Multiple Autonomous Underwater Vehicles Based On Structured Artificial Potential Field

The Autonomous underwater vehicle(AUV)has the characteristics of low cost,flexible operation,miniaturization,and intelligence.Compared with traditional marine forces,it is performing related tasks such as environmental information collection and target detection.With the increase in the complexity of operating tasks,it is difficult for a single AUV to meet its needs,and the further development of AUVs’ flocking has also been focused on.At present,unmanned flocking is a research hotspot in various industries.The multi-agents collaborative system is a flexible system composed of multiple agents.The agent in the system can combine autonomous execution and collaborative execution for tasks to meet the increasing need.However,the control problems of multi-AUV systems are often restricted by the limitations of underwater communication distance,speed,bandwidth,etc.,and there is no perfect solution can be obtained.This paper focuses on the problem of multi-AUV cooperative formation with limited communication.The main research of the paper are summarised as follows:(1)First of all,a multi-AUV formation motion control system framework with limited communication is established.The relationship between AUV and agent is established through line of sight tracking method,combined with the heading and speed adaptive PID control method of AUV,the AUV motion control organization layer is abstracted,and the initial algorithm framework of this paper is obtained,which is based on Olfati-Saber’s artificial potential field concept.The algorithm is an AUV cluster control method that does not rely on underwater communication.The flocking control of the AUV is completed under the condition of only relying on sensor perception information.Compared with the general communicationbased control method,this method can complete Large-scale decentralizati-on effectively,and it avoids the collapse of a centralized structure due to communication limitations,also has a high degree of scalability.(2)Secondly,the AUV formation control method that does not rely on communication is established.In consideration of task requirements such as AUV information collection,a multiAUV system needs to be able to arrange any special formations and moving.On the basis of the agent structure formation formed by the AUV flocking control method,a formation control algorithm based on a structured improved artificial potential field is proposed.By adding virtual agent neighbors at the AUV motion control perception layer,the expected formation can be gradually formed.In the case of limited communication,the comparison with the virtual structure method verifies the superiority of the structured improved artificial potential field algorithm in decentralization and formation control.(3)Finally,the obstacle avoidance control algorithm of AUV formation is established.The formation control algorithm based on structured improved artificial potential field is a method based on graph theory.Although this method can complete obstacle avoidance and formation reorganization,it has poor performance in running time,path length,obstacle avoidance cost,formation maintenance,etc.,this article adds a switching obstacle avoidance control strategy based on the switching topology on this basis,which can sensitively react to the environment with obstacles,and maintains stability in the process,avoiding mutual conflicts between AUVs.Under the obstacle control method,the AUV performs corresponding movement adjustments based on its own perception information to realize decentralized distributed control,and defines the formation safety zone,which can distinguish different environments and accomplish the task.