Distributed Multi-robot Collabortive Task Allocation Mechanism For Smart Factory

With the continuous development of “Made in China 2025”,through close cooperation around key areas of key manufacturing areas,China has launched a new generation of information technology and intelligent manufacturing equipment integration development and innovation.At present,many businessmen have built smart factories,such as a large number of intelligent manufacturing high-end enterprises,car companies,production equipment manufacturers,etc.With the full automation and intelligence of the manufacturing process in the tradition factory,many kinds of robots are introduced into the manufacturing process to work together for manufacturing and processing tasks.Therefore,the multi-robot system(MRS)that performs manufacturing and manufacturing tasks in smart factories will become extremely massive,and more complicated.How to effectively allocate the manufacturing tasks and making the smart factory multi-robot system operate efficiently becomes an urgent problem.In this thesis,the problem of multi-robot collaborative task allocation in smart factories with static environment is studied.Firstly,the research status of Multi-robot Task Allocation(MRTA)at home and abroad,and the related theories of multi-robot task decomposition are studied.The task allocation classification method is used to classify MRTA problems for various multi-robot systems,and the corresponding solutions of MRTA problems are analyzed in turn.Then,two kinds of MRTA problems commonly encountered in intelligent factories are studied.Aiming at the MRTA problem of simple type in intelligent factory,a distributed multi-robot communication network is established.Considering the requirements of cooperative tasks and the constraints of execution ability of robots,the multi-robot task allocation problem is defined.Based on the decomposition of the task,a distributed auction algorithm is designed to solve this problem.Moreover,the algorithm does not require the control center to store the relevant auction information.Each robot communicates with its neighbors in the multi-robot network to update the local price information.Then,this thesis studies complex type MRTA problem in smart factories.According to recent research progress,the method of forming the multi-robot alliance is used to solve such problems.By combining the energy residual function,the task cost function and the fault tolerance function,the objective function of the multi-robot alliance is established.Combined with the objective function,the improved binary particle swarm optimization algorithm(IBPSO)is used to solve the alliance formation problem.On the basis of the formation of multi-robot alliance,a single task distributed task allocation mechanism is designed,and the algorithm and experimental simulation of the alliance formation algorithm are carried out.Finally,this thesis develops a multi-robot task distribution system based on the WEB.The system has production planning modules,robot management modules and robot monitoring modules,and We can complete the mapping relationship between production planning and robot task allocation through these modules.And the simulation experiment of task allocation in the system verifies the task allocation mechanism proposed in this paper is reasonable.