Research And Implementation Of Multi-robot Formation Control Technology

Nowadays,the robot field is developing from a single robot with a centralized function to multi-robot systems where robots cooperate with each other.Formation control is a fundamental problem in multi-robot systems.A great challenge is the development of algorithms to dynamically assign positions to robots and to form a desired formation,not relying on a priori assignment information.In this paper,to address this problem,we propose a dynamic role assignment formation control algorithm based on an improved Hungarian method that combines role assignment and formation synthesis together.This algorithm can synchronously finish role assignment and decide where to form the predefined formation with the lowest cost.And the improved Hungarian method guarantees that no matter the number of robots and the number of positions are equal or not the algorithm can always achieve the desired formation.Secondly,based on the formation control algorithm,this paper also proposes a formation algorithm selection mechanism,which can combine the advantages of various formation control algorithms,and reasonably select proper formation algorithm according to the motion state of robots and the task requirements.Finally,this paper implements the formation simulation prototype system.The system can meet the robot formation requirements through interface interaction,and visualize the formation process.Based on the prototype system,the performance of the formation control algorithm and algorithm selection mechanism is evaluated through a series of simulation experiments and comparison with other algorithms.The experimental results show that the formation algorithm and algorithm selection mechanism are accurate,reliable and efficient.