Influence Of Formation Topology Structure Of Multi-AUVs On The Performance Of The Coordinated Control System

Affected by the flocking,foraging and some other behavior of swarm intelligence which includes fish,birds and other organisms in nature.Multi-agent coordination problem has attracted more and more public attention.It can be summarized as that,agents who have the ability to press a certain degree of autonomy law constitute a network,and use a local control algorithm to adjust the control input in order to exchange the key state information more effectively with neighbor to complete collaborative tasks.Compare with a single agent,multi-agent system can adapt to the complex environment and has certain advantages in solving large,complex practical tasks.Therefore,it has certain practical significance to research cooperative control and navigation problem.In this paper,we study an autonomous underwater vehicle(AUV)which can maintain a certain team in the collaborative tasks and overall perceive environmental information across the unknown region and has strong adaptability and robustness.Complete tasks with multi-agent may lead to improve the system efficiency and shorten the time.Redundant and complementary resources within multi-AUV can also increase the likelihood of successful realization of the task,it is helpful to solve some complex and challenging practical problems.Cooperative control algorithm is necessary safeguard for multi-AUV system and interactive information between the members is necessary precondition for cooperative control.In order to get other member's formation for every agent,we use a distributed cooperative control strategy based on the interaction of local information.The paper analyzes the impact of changes in general topology network on cooperative control and navigation performance.The main contents are as follows:(1)Regard the members and connection between two members as points and edge respectively.Based on a distributed cooperative control strategy,get the model with a leader-follower structure and undirected topology between followers and followers.(2)Give the necessary and sufficient conditions of its stability.(3)Derive the necessary and sufficient conditions to reach a maximum speed of convergence and analyzes the impact of changes in general topology network on the maximum convergence rate and design the experiment to verify the conclusion.(4)First,build the collaborative navigation model.Secondly,the paper discusses application of EKF,UKF and PF in collaborative navigation for multi-AUV in order to eliminate measurement errors.Combine UKF and PF,we study UPF algorithm.Finally,in order to reduce network complexity and energy consumption,introduce consensus algorithmand propose a collaborative navigation algorithm based on the consensus and UPF with multi-leader and analyze its navigation performance and the influence of topology structure on the performance of collaborative navigation and positioning accuracy.