Design And Implementation Of Multi-Robot Cooperative Hunting System

The rapid development of economy has given birth to many heavy and complicated tasks.Because of the high cost and low efficiency of a single robot in handling such tasks,the multi-robot system has attracted the attention of experts and scholars.Compared with a single robot,multi-robot group has the advantages of strong robustness,good flexibility and high efficiency when dealing with complex tasks.Multi-robot collaborative hunting has become an important approach and platform to study multi-robot system because it involves several key technologies such as target search,collaborative task assignment,path planning,target tracking and collaborative formation control.Based on the mobile robot platform,this paper designs a system of multi-robot cooperative hunting and verifies the process of multi-robot cooperative hunting.The main work is as follows:?.Construction of multi-robot cooperative hunting system.The system consists of three layers: object layer,communication layer,monitor and display layer.The object layer mobile robot is designed with the concept of modularization,which mainly includes three modules: energy and power,central control and information perception.The communication network of the system is constructed based on socket.The communication protocol between the swarm robots is TCP protocol,and the communication protocol between the swarm robots and the background monitoring computer is UDP protocol.The monitor display interface is designed based on ros-gui,and the video stream received is decompressed and decoded.This paper will elaborate on the multi-robot cooperative hunting system.?.Laser and visual information perception.Mobile robot perception of global environmental information and the target of visual information is the premise and foundation to complete the task of hunting smoothly.In this paper,mobile robot perception information mainly includes three aspects.First,start up the lidar sensor and run the cartographer algorithm to build a map of the experimental area.Secondly,the camera is calibrated to obtain the parameters of the camera,and the calibrated camera is used to identify the target robot with a fixed drone.Thirdly,the position of the target robot is estimated by combining the position of the hunting robot(the coordinates of the hunting robot in the map)and the relative position of the robot and the target robot(the coordinates of the target robot in the camera coordinate system).?.Multi-mobile robot cooperative hunting.In this paper,multi-robot cooperative hunting is divided into three stages: target searching,target following and target surrounding.According to the characteristics of the experimental area,the target search strategy of the robot group was developed and realized by combining relevant technologies.Based on the information of the target robot in the camera,the tracking method of the hunting robot to the target robot is designed by PID control.The surrounding behavior is divided into two parts: the following of the hunting robot to the target robot and the formation control of the hunting robot group.The formation control of the hunting robot group is realized based on the formation theory.?.Program validation and result analysis.In the actual environment,based on the multi-robot cooperative hunting system,the design of the multi-mobile robot cooperative hunting scheme is verified.Experimental results show that the proposed scheme is reasonable and effective.