Multi-robot Collaborative Environment Detection Research

Detecting unknown environments is a fundamental problem in the field of mobile robots,which is a necessary condition for robots to perform other tasks in an unknown environment.Compared with a single robot,multi-robot system has better robustness,adaptability,flexibility and scalability.It is more suitable for detecting unknown environment,but there are still many problems in map fusion and task allocation.Map fusion directly determines the accuracy of multi-robot detection unknown environment,and task allocation affects the efficiency of multi-robot completing environmental detection.In this paper,we studied two important contents of multi-robot cooperative detection in the unknown environment of map fusion and task allocation.Aiming at the technical difficulties in map fusion and task allocation,this paper proposed solutions to improve the detection efficiency.The main work of this paper is as follows:Firstly,a geometric-topological hybrid map fusion method is proposed.The topological nodes and their geometrical features are taken as the basic unit of map matching,and the similarity of topological nodes' geometrical features is taken as the same judgment criterion of topological nodes.The Euclidean distance between the topological nodes and the neighboring nodes is added to the calculation of node similarity.An effective map fusion index and map fusion order are proposed.Firstly,the map fusion method based on geometric-topology hybrid map is studied.In order to increase the reliability of map fusion,this paper takes the topological nodes as the unit,the topological nodes contain the geometric figures as the judgment basis for map fusion.A hybrid map representation method based on point,line geometry and topological nodes is proposed,which makes better use of geometric map and topological map.This paper introduces how to select the angle as the basis of fusion judgment,analyzes the selection method of fusion reference,and designs a simple and convenient map fusion method.At the same time,the map fusion is analyzed,and the best fusion result is selected as the final fusion map according to different fusion reference objects.Secondly,a multi-robot task allocation strategy based on market law is proposed.By using the auction idea of the economic market and using the contract network protocol,the second auction method is used to allocate the tasks according to the state of the robot.In order to prevent the robot from concentrating too much on the same area,the common bid calculation method is improved,and the bidding calculation is made by rejecting the pheromone,which reduces the repeated exploration of the same area by the robot.Finally,the experiment of the map fusion algorithm and task assignment algorithm proposed in this paper is carried out in the simulation experiment platform,and the effectiveness of the method is verified by setting different numbers of robots and environment.