Research On Path Planning,Tracking And Hunting Control Of Multi-mobile Sensor System

In real life,if the agent does not have the ability to plan independently,it will be difficult to achieve formation control,and round-up control,as a form of formation control,requires the agent to form a formation and plan a path autonomously.Therefore,the in-depth study of the path planning,tracking and round-up control problems of multi-agent systems not only has important theoretical significance,but also has practical application value.This dissertation mainly studies the path planning and round-up control of multiple mobile sensors based on the multi-agent system.The research content is as follows:(1)To solve the control problem of stationary multi-target rounding up,a path planning and rounding up algorithm based on RRT and OSPA is proposed.Firstly,the modal judgment is performed,and the rapid search random tree algorithm(RRT)is used to plan the path and keep the leader-follow formation cruising movement when it is not in the round-up range.When the leader enters the round-up range,through OSPA distance calculation,the round-up targets are assigned to each agent and RRT path planning is performed respectively.When encountering a static obstacle,the obstacle avoidance algorithm is used to make the agent avoid the obstacle,and finally realizes the round-up.(2)To solve the problem of target rounding up and path planning in a dynamic obstacle environment,a path planning algorithm based on potential field is proposed.Divide the agent system into two types(type I and type II),both of which are modeled by unicycle kinematics,where type I represents rounding up the agents,sharing their respective states,and calculating the distance between each other and whether they are in contact with obstacles at all times objects,class II represents moving obstacles.There is a repulsive force between type I and type II,and there is a gravitational force with the target.The speed and direction are adjusted according to the magnitude of the gravitational force and the repulsion force,and finally the round-up control is realized.(3)To solve the path planning and round-up control problem of first-order nonlinear multiple dynamic targets,an algorithm based on event trigger strategy is proposed.In the path planning process,the Kalman tracking algorithm is used to obtain the target state estimation,and then the multi-agent is tracked according to the target estimation by designing the controller,and finally the path is generated.This article uses matrix theory,graph theory and Lyapunov stability methods to obtain the conditions for system stability.Through the event trigger mechanism,the entire round-up process is divided into three stages.The first is the cruise stage,where multiple agents maintain leader-follower the formation to cruise and move;the second is the pre-round-up stage,when the distance between the agent and the target is less than a certain value,the event is triggered at any time,and the pre-round-up is carried out;the third is the final round-up phase,which triggers the event after maintaining the pre-round-up formation for a certain period of time,reduces the round-up range,and finally realizes the round-up control.