Research On Job Shop Scheduling Technology And System With Multi-resource Constrains Based On Multi-Agent System

With the continuous advancement of “Made in China 2025”,companies have increasingly strong demand for automation and intelligence in the manufacturing process.Job-shop scheduling is the key to the manufacturing process.The current job-shop scheduling system cannot complete the scheduling process combining manufacturing resources such as tools and materials.The single resource scheduling ability of the system has become a bottleneck restricting the intelligence of the job-shop,and the production efficiency of the enterprise cannot be improved.At the same time,due to the complex and changeable manufacturing environment,interference and abnormalities may occur in the production process at any time,which puts forward higher requirements on the abnormal response and handling capacity of the scheduling system.Therefore,the research on the multiple resource scheduling capabilities and the response processing to abnormalities of the scheduling system has important practical significance.Aiming at the production line’s demand for tool scheduling and rapid response to abnormalities,this paper constructs a mathematical model based on the theory of multi-agents,and reasonably divides the functions of the system and designs four types of agent: production line management agent,work order management agent,execution management agent group and resource management agent.The distributed structure of the scheduling system is determined by combining the work process and collaboration methods of each agent.The scheduling cooperation strategy between the agents of the multi-agent system based on the tool constrain is designed,including the multi-agent static scheduling strategy and dynamic rescheduling strategy.The static scheduling strategy is based on current resource information to statically schedule tasks,using the characteristics of multi-agent collaborative planning.It can combine the two resources of machines and tools to complete the scheduling of work order tasks;The dynamic rescheduling strategy is the optimization of the static planning scheme and the response and rescheduling to interference and abnormalities in a dynamic environment.The scheduling algorithm uses improved genetic algorithm to solve and plan the production line scheduling.The coding method,the cross method and mutation method of the genetic algorithm is improved.It combines with the variable neighborhood algorithm to improve the quality of the solution.Experiments have proved the effectiveness of this algorithm.Finally,the architecture design and software development of the multi-agent flexible workshop scheduling system were completed based on the micro-service model.Experiments verified the effectiveness of the scheduling system in static resource scheduling and dynamic abnormal scheduling strategies,which maintains the production efficiency and stability of the system.