Design And Implementation Of Multi-Robot Collaborative Exploration Environment And Map Construction System

Multi-robot collaborative exploration and map construction in unknown environments is one of the hotspots in the field of mobile robots.This has great research value and application potential in the fields of space exploration,disaster area rescue,home service robots and so on.Considering the indoor working environment of mobile robots,the system framework of multi-robot collaborative exploration environment and map construction is designed,and the related technologies are deeply studied,and the software design and implementation for each module of the system are also carried out.The related study of multi-robot collaborative exploration environment,map construction and related technologies is summarized,and the difficulties of system design and implementation are analysed,and the requirement analysis is carried out.A multi-robot collaborative exploration environment and map construction system with flexibility and stability in indoor unknown environment is designed based on the Robot Operating System(ROS)and the system operation principle is elaborated in detail.Firstly,the related technologies of Simultaneous Localization And Mapping for mobile robot based on graph optimization theory are studied.The Monte Carlo Localization algorithm based on particle filter is used to realize the mutual localization of multi-robot in global coordinate system,and the relative initial poses of multi-robot are determined.Considering the requirement of global map fusion,a global map fusion module is designed based on the detailed discussion of coordinate system transformation and map fusion method in multi-robot system,which realizes the fusion of multiple local maps.Secondly,around the requirement of multi-robot collaborative in exploring environment,the improved rapidly exploring random tree algorithm is introduced into the search of unknown boundary points in the way of behavior-based strategy.The experimental test and analysis are carried out,and the boundary points search module based on the improved rapidly exploring random tree algorithm isdeveloped,which realizes the fast search of unknown boundary points in maps.Then the density-based clustering algorithm is discussed in detail,and the experiments are carried out to compare and analyze.A clustering and updating module of unknown boundary points set is developed,which realizes the clustering and updating of unknown boundary points set.Thirdly,the problem of multi-robot collaborative exploration task allocation is analyzed in detail.A dynamic task allocation strategy based on market method is designed.A module of task management and control is developed to realize the dynamic redistribution of multi-robot and target tasks.Considering the requirement of state monitoring of the multi-robot system,a human-computer interaction interface is designed and implemented based on Qt.Finally,combined with the indoor working environment of mobile robots,the test and evaluation of multi-robot collaborative exploration environment and map construction system and the overall operation demonstration are carried out by building simulation indoor environment and real indoor environment respectively,which verifies the rationality and effectiveness of the overall system architecture design and the development of each functional module.