Research On Logistics Business Process Composable Modeling And Simulation Based On Multi-Agent

With the advancement of global economic integration, the prosperous development of internet of things technology, the growing of cloud logistics, the constant emergence of novel IT technology, the connotation and definition of logistics business processes have been continuously extended, which shows its uncertainty, discrete randomness and complicacy. The traditional business processes modeling cannot be applicable for such complicate and changeable requirements, so a new method is urgently needed for this kind of model. Thus, building an effective, flexible process model with low cost is urgently required.This study analyzed the characteristics and requirements of current business process based on the complex systems. Then a multiple Agent business process modeling portfolio was developed according to the complexity and uncertainty of the process, and the optimization was proposed based on the simulation results, aiming to realize the online control and offline decision, guide to the intelligent and dynamic flexible modeling, thus to provide pioneering idea and method for the business process modeling and simulation.This research topic was concluded from the practical project and the approaches have already been used in practice. Firstly, the research status, main theories and technologies were analyzed and compared. And then the structure and composition of business process were dissected in different ways. A four-dimensional area analysis-oriented business process model was built, which was integrated with Agent modeling technology; and a subsequent multidimensional Agent-process-oriented modeling was proposed. A decomposition strategy based on the process exploiting technology and a module-oriented public process library and multi-objective combined strategy were designed. Several intelligent, self-adjusted and demand-combined decision model are obtained, including:DEVS-based atom model, public module library-oriented multilevel model and distribution-based combination model. Finally, the multidimensional Agent business process-oriented modeling method was applied to the LongChang warehouse to improve its business process, including: introducing the intelligence of RFID internet of things technology, designing a new business process and complete the modeling and simulation for the improved business process. After the comparison of the former and latest simulation results, business process integration was obtained, which strongly support the feasibility and affectivity of the proposed model.Issues solved by this study and the main innovative points:(1) Based on the perspective of the area analysis-oriented model, a four-dimensional multilevel model was proposed and was integrated with Agent modeling technology, thus a behavior-oriented modeling method was built.(2)A decomposition strategy based on the process exploiting technology and a module-oriented public process library and multi-objective combined strategy were designed, then several intelligent, self-adjusted, reusable models were obtained, and dynamic flexible simulation system was proposed. This introduced a novel idea for this research field, and might be the most valuable achievement in this study.(3) Warehouse was taken as an example to be rebuilt and optimized. The obtained intelligence and upgrade strongly support the extensive application prospects of the proposed model.