| At present, because of resources diminishing, environment deteriorating, constraint of environmental policy increasing, steel production enterprise must achieve the targets of energy saving and emission reduction, further improve its adaptability and international competition ability, realize resource recycling and itself sustainable development. So, it is very necessary for steel production enterprise to research, analyze and optimize the logistics systems.Steel production is a complex system with more processes, more workstation, lager spatial span and more product categories. It is a key factor that affects the product categories, quality and yield in logistics system of Steel production. So, it is the key to understand logistics rules, formulate reasonable production plans and scheduling for guaranteeing production logistics unblocked, increasing production and quality, and reducing the cost.Because of its autonomy, social ability, reactivity, pro-activity, mobility and reasoning, agent can be used to realize the dynamic, uncertain environment, large-scale software systems. Multi-Agent System is defined as a system of a population of autonomous agents, which interact with each other to form the solution, according to environment and knowledge reasoning, learning etc. It can improve overall system performance and meet the system requirements of mobility, flexibility and openness.Agent-oriented methodology has provided a very natural method to understanding, modeling and development the complex system. It made the distributed structure easier, intelligent and robust, so it is recognized as a novel mythology for complex systems.The aims of this dissertation is to explore new simulation method not only describing the complex characteristics of logistics in steel production process, but also responding to the dynamic characteristics of logistics system and be widely used for different forms of steel production flow, which has a very important significance in both theory and practice. Aimed to the new request of flexibility, pliability and compatibility and the complex characteristics of logistics in steel production process, because of Agent technical superiority and Multi-Agent system merit, which could be used to overcome the shortcoming of the other's modelling method.so, simulation model of logistics in steel production process based on Multi-Agent was presented. In the model, the logistics system in steel production process has been defined as a Multi-Agent System. Based on the versatile principles, the technology of analyzing, designing and realizing for simulation system model was studied from the standpoint of software engineering. First of all, based on understanding and grasping the characteristics of complexity in steel production process, system function of simulation model of logistics in steel production process based on Multi-Agent has been abstracted and defined by introducing Multi-Agent technology in modelling process.Secondly, interactive models of agents have been established by analyzing the types and structures. Internal and external action of agents has been described by extending the structure and function of Petri Net and developing behavior theory of agent and interactive behavior theory of agents.Next, because of the special requirement of flexibility and mobility for logistics system in steel production process, the transportation system has played an important role in the whole logistics system, so the transportation system was combined with the whole logistics system in steel production process by exploring the operation mechanism of crane. Finally, to ensure the effectiveness and correctness of simulation model based on Multi-Agent, the logistics simulation system has been realized by extending the method of Multi-Agent System Engineering, and the system model has been designed and verified by the Agent Tool.According to actual production conditions of No.3 steel-making plant in Kunming Iron and Steel Co., LTD,the simulation model of logistics system in steel production process based on Multi-Agent was established and the simulation results was compared with actual data, so the results show:The simulation model of logistics system in steel production process based on Multi-Agent is correct and effective, according to the actual scale requirement, which was flexibly established by different processes and workstation. In similar input conditions, logistics average circulation time from converter to continuous casting was compared with simulation results, and the results further show:the simulation model of logistics system in steel production process based on Multi-Agent is consistent with actual system, which can correctly reflect production logistics system in steel production process.The simulation model of logistics system in steel production process based on Multi-Agent can build up model for different steel-making process, and simulation can disclosure bottleneck of different steel-making process under different productions, such as desulfurization process is bottleneck in steel-making process with the mixer furnace process, as for steel-making process without mixer furnace process, desulfurization process is bottleneck when the arrival of the hot metal from blast furnace is three ladles percent sixty minute, but when the arrival of the hot metal from blast furnace is more than five ladles percent sixty minute, converter process is bottleneck.Simulation analysis using this model is help to make decision for improving production efficiency and continuous eating of different steel-making process. Through comparing the simulation results of steel-making process with mixer furnace process and the one without mixer furnace process, it is shown that when the arrival of the hot metal from blast furnace is three ladles percent sixty minute, it is good for improving production efficiency to accelerate the arrival of the hot metal from blast furnace and the production time of converter process, for steel-making process without mixer furnace process the production time of converter process should be longer in order to improve casting. The arrival of the hot metal from blast furnace is almost independent on production efficiency and continuous casting when it exceeds five ladles percent sixty minute, it is benefit for improve production efficiency to accelerate the production time of converter process for steel-making process without mixer furnace process.The simulation model of logistics system in steel production process based on Multi-Agent has good generality, practicability and flexibility, which can be used to simulate logistics characteristics of metallurgical production and diagnose and rehearse logistics system by different production process. According to the system simulation results of simulation model, production organization and production process in steel production process was optimized and decision support for the whole continuous casting production management and logistics control was provided. The simulation results show that simulation model of logistics system in steel production process based on Multi-Agent is convenience and validity to express the characteristics of logistics in steel production process and disclosure the rule of complex logistics, has flexibility and adaptability for modeling and simulating in different steel production process, and can better satisfy the current requirement of distributed complex system modeling. This model method provides a new tool and route for simulation logistics in complex steel manufacturing process. |
PDF Full Text Request