Research On Multi-robot Collaborative Method Based On Vision

The demand for robots in today's society is not limited to repetitive tasks,but it is hoped that robots can have a certain degree of independent coordination.Therefore,more and more experts and scholars have begun to study multi-robot systems,and multi-robot collaboration has gradually become one of the research hotspots in recent years.Multi-robot systems have many advantages that are unmatched by single robots,such as reliability,parallelism,and scene distribution characteristics.In this paper,based on the apriltag visual reference system robot,the robot global positioning model is established,and the apriltag benchmark system is analyzed and improved.The monocular vision can be used to obtain the robot's own pose,which has simple resources,low cost and easy layout.Features.The concept of generalized virtual potential field is introduced,and the potential field model based on target gravity and obstacle repulsion is constructed.The path planning algorithm based on traditional potential field method is completed and simulation verification is realized.Aiming at the defects of the target unreachable and local minimum points in the traditional potential field method,an improved model of the repulsion potential field function is proposed for the target unreachable problem.The distance between the robot and the target point is introduced as the variable factor and the adjustment parameters are added to realize the potential field.Control;propose a detection method for the local minimum value problem and a multi-robot temporary sub-object escape strategy,solve the problem of the robot in the unknown environment,and carry out simulation verification.Based on the improved artificial potential field method,a multi-robot conflict resolution algorithm is proposed.The obstacle avoidance detection radius is determined according to the speed and priority.The robot conflict problem during formation formation is solved,and the high priority robot is guaranteed to achieve the target preferentially.Based on this,the leader formation method is introduced to control the formation of the robot.The improved artificial potential field method is used to design the planning strategy of the pilot robot.The follow-up strategy and obstacle avoidance method of the following robot are proposed,so that the formation avoidance obstacles are completed in multiple robots.At the same time,a collisionfree path from the starting point to the end point is planned.Finally,the consistency of the following robot is optimized.The control parameters of the following robot are defined by the consistency theory,and the validity is verified by simulation.Finally,the theory is applied to practice through the square robot platform and apriltag visual positioning.The experiments verify the formation of multi-robot formation,multi-robot conflict resolution and multi-robot collaborative formation theory.