基于深度强化学习的平行企业资源计划

发布时间：2018-03-27 20:31

  本文选题：企业资源计划　切入点：深度强化学习　出处：《自动化学报》2017年09期

【摘要】：传统的企业资源计划(Enterprise resource planning,ERP)采用静态化的业务流程设计理念,忽略了人的关键作用,且很少涉及系统性的过程模型,因此难以应对现代企业资源计划的复杂性要求.为实现现代企业资源计划的新范式,本文在ACP(人工社会(Artificial societies)、计算实验(Computational experiments)、平行执行(Parallel execution))方法框架下,以大数据为驱动,融合深度强化学习方法,构建基于平行管理的企业ERP系统.首先基于多Agent构建ERP整体建模框架,然后针对企业ERP的整个流程建立序贯博弈模型,最后运用基于深度强化学习的神经网络寻找最优策略,解决复杂企业ERP所面临的不确定性、多样性和复杂性.
[Abstract]:Traditional Enterprise resource Planning (ERP) uses static business process design concepts, neglects the key role of people, and rarely involves systematic process models. Therefore, it is difficult to cope with the complexity requirement of modern enterprise resource planning. In order to realize the new paradigm of modern enterprise resource planning, this paper is driven by big data under the framework of ACP( artificial society), Computational experiment (Computational experiment), parallel execution of parallel execution. The enterprise ERP system based on parallel management is constructed by combining the deep reinforcement learning method. Firstly, the framework of ERP integrated modeling is constructed based on multiple Agent, and then the sequential game model is established for the whole process of enterprise ERP. Finally, the neural network based on deep reinforcement learning is used to find the optimal strategy to solve the uncertainty, diversity and complexity faced by the complex enterprise ERP.
【作者单位】： 中国科学院自动化研究所复杂系统管理与控制国家重点实验室;青岛智能产业技术研究院;中国科学院自动化研究所北京市智能化技术与系统工程技术研究中心;
【基金】：国家自然科学基金(71702182,71472174,71232006,61533019,61233001,71402178) 复杂系统管理与控制国家重点实验室优秀人才基金(Y6S9011F4E,Y6S9011F4H)资助~~
【分类号】：F272.9;TP18
，

本文编号：1673133