Multiple Mobile Robots In Indoor Environment Positioning Research

The intelligent era of Industry 4.0 is coming,and the positioning technology of mobile robots is developing rapidly.Research on positioning based on a single robot has made great progress in various aspects,but the current research directions of researchers are still mostly concentrated on a single machine.However,when a single robot performs positioning tasks,it has the shortcomings of limited available information,low efficiency,and easy accumulation of errors.It has been unable to meet the realistic needs of task stability.This requires the expansion of a single robot into a multi-robot system.The current research on multiple machines is still Relatively small.Although the multi-robot collaborative positioning technology can effectively solve the single-robot problem,it also faces many challenges.First,how does the robot determine where it is in the environment;second,how to establish the data association between the robots;and how to use the associated data for map fusion.Therefore,for the positioning of multiple mobile robots in an indoor environment with unknown a priori maps,the work in this paper is as follows:First of all,research on commonly used robot-related positioning technologies.In indoor scenes where global navigation satellite systems cannot meet the needs,the use of Simultaneous Localization and Mapping technology,an incremental mapping method,can be very good.To solve this problem.Through research and analysis,it is established that the distributed architecture is used to study the multi-mobile robot positioning system,the multi-robot sensor motion model is established,and the description model of the environment map is determined.Then,a scene-based recognition algorithm is used to complete the relative pose association of the robot.When a robot detects that it has reached a scene that another robot has been through,it shares data with each other and matches similar scenes through a bag-of-words model to establish connections with other robots,avoiding robots that cannot meet during movement.Happening.In addition,the traditional matching algorithm requires a long calculation time to find the corresponding point between the model point and the data point.To overcome this problem,this paper uses an accelerated iterative closest point algorithm based on height information to find the corresponding relationship by boundary search,and solves the problem of large amount of positioning calculation due to low matching efficiency.Then,based on the relative pose relationship between the robots,the pose constraint relationship between all nodes in the entire environment of the multi-robots is established,and the specific process framework for completing the map fusion is given.Through the constructed multi-robot global pose relationship graph,the least square algorithm is used to complete the multi-robot global graph optimization method.Finally,the proposed multi-robot scheme is tested.The results show that the scheme proposed in this paper can realize the fusion of multiple maps in an indoor environment,avoiding repeated exploration of the same location,and the effect of map fusion can reach expectations,and a reliable multi-machine coordinated positioning solution is proposed.