Research On The Task Allocation And Collision Avoidance Methods Of Multi-robot Co-detection

With the development in robotics system, the capability of robot is improved ceaselessly and its application field has been extended. Robots are being expected to do more complex task which can't be accomplished by a single robot. Obviously, these tasks need coordination and cooperation among multiple robots to finish. So it is important to study multi-robots cooperation system. The detecting of multi-targets by multi-robots has extensive use.This dissertation takes the task of detecting multi-targets cooperatively with multi-robots as center. Firstly, introduce the architecture of multi-robot system, and then mainly research the task allocation and collision avoidance method of multi-robots. The main contribution and work are described as following:In the research of the task allocation method of multi-robots, first we can use the combinational auction to gain the efficient target sets, use the improved ant system algorithm to detect those target sets. So I propose a task allocation method of multi-robots based on the improved ant system algorithm. The proposed methods include multi-robots parallel searching simultaneously, renew the pheromone, and improve the local search algorithm. At last, realizes the task allocation method of multi-targets detecting by multi-robots through simulation. The results show the validity and correctness of the proposed method.In the process of robots perform task, there may be exist the phenomenon of crossing path, and the crossing path is the main reason for the collision happens in multi-robots. In order to solve the collision avoidance problem of multi-robots, this dissertation propose a coordination collision avoidance method of multi-robots based on the movement planning. By problem statement, model building, the design of the traffic rules and the design of priority, etc, to realize the resource allocation in conflict region, that is, the collision avoidance problem of multi-robots. The simulation results show the proposed method is feasible.