分布异构环境下基于多Agent的复杂供应链网络计算实验方法研究

Computational experiment for complex supply chain networks has arisen as an active research topic in the field of computational experiment's applied research. Current research in this area focuses on case studies of computational experiment for complex supply chain networks. In-depth and systematic research has yet to be conducted on the theory and methodology of multi-agent based computational experiment for complex supply chain networks. After addressing limitations of computational experiment for complex supply chain networks in existing literature, this proposal aims to investigate multi-agent based computational experiment modeling approach for complex supply chain networks under distributed and heterogeneous environments. The applicants will establish a multi-layered local and global computational experiment modeling approach based on the structure and function of supply chain networks under distributed and heterogeneous environments and an approach for ontology based correctness and consistency test of computational experiment models. Building upon the modeling research, the applicant will propose an approach and the corresponding key technologies for multi-agent based computational experiment implementation for complex supply chain networks under distributed and heterogeneous environments. After research above, the applicant will carry out research on multi-agent based distributed computational experiment platform for complex supply chain networks. The architecture of the platform will be established and the models, methods and key technologies to solve key issues in the platform design and development will be proposed. Finally, these new computational experiment approaches and technologies will be applied to and calibrated with complex and multi-facet real-world case studies. It is expected that this proposed research will further develop and refine multi-agent based computational experiment theory and methodology, expand and improve computational experiment research for supply chain networks, and play a significant role in providing a scientific framework for computational experiment for supply chain networks and enhancing the quality and efficiency of computational experiment.

复杂供应链网络计算实验问题是计算实验应用研究领域的一个热点问题，但目前研究主要关注供应链网络计算实验实例问题，而对分布异构环境下基于多Agent的复杂供应链网络计算实验理论和方法缺乏深入研究。本项目针对现有研究的局限，开展分布异构环境下基于多Agent的复杂供应链网络计算实验建模方法研究，提出分布异构环境下基于结构和功能的多层次的局部/全局计算实验建模方法和基于本体的计算实验模型正确性和一致性验证方法；进一步地，提出分布异构环境下基于多Agent的复杂供应链网络计算实验实现方法和关键技术；并研究基于多Agent的复杂供应链网络分布计算实验平台的构建问题，提出该平台的体系结构，针对关键问题提出相应的模型、方法与关键技术；最后开展相关应用研究。其研究成果对进一步发展和完善基于Agent的计算实验理论与方法，拓展供应链网络计算实验研究体系，提高供应链网络计算实验科学化及其质量与效率具有重要意义。

复杂供应链网络计算实验问题是计算实验应用研究领域的一个热点问题，但目前研究主要关注供应链网络计算实验实例问题，而对分布异构环境下基于多Agent的复杂供应链网络计算实验理论和方法缺乏深入研究。本项目针对现有研究的局限，首先开展分布异构环境下基于多Agent的复杂供应链网络计算实验建模方法研究，提出基于结构和功能、融合多Agent系统、流和过程等知识体系的多层次的计算实验建模方法，建立基于多Agent系统的供应链网络结构表达方法和集合流和过程的供应链网络功能表达方法；进一步，结合供应链网络领域知识和多Agent系统特性，提出分布异构环境下基于多Agent的复杂供应链网络计算实验实现方法，建立基于流视角下标准过程的供应链网络情景与Agent的建模与实现方法；并融合多Agent系统、流和过程等知识体系，提出分布异构环境下基于多Agent的复杂供应链网络分布计算实验平台关键技术，解决多Agent系统、流和过程等不同方法论的统一问题；最后开展相关验证与应用研究。其研究成果对进一步发展和完善基于Agent的计算实验理论与方法，拓展供应链网络计算实验研究体系，提高供应链网络计算实验科学化及其质量与效率具有重要意义。