分布式MPC在网络化制造环境下SCM中的应用  

Applications of Distributed Model Predictive Control in Supply Chain Management under Networked Manufacturing

作　　者：董海[1] 王宛山[1] 李彦平[2] 巩亚东[1] 

机构地区：[1]东北大学机械工程与自动化学院 [2]沈阳大学信息学院,沈阳110044

出　　处：《系统仿真学报》2007年第6期1354-1357,共4页Journal of System Simulation

基　　金：国家自然科学基金(60274027);教育部高校博士点专项基金(0145015)

摘　　要：针对网络化制造环境下动态供应链管理特点,利用分布式模型预测控制技术解决供应链中动态优化控制问题。首先,阐述动态供应链运行机理,并以此提出完整的随机动态优化控制模型及其控制约束;其次,给出分布式模型预测控制算法,并构建了优化控制序列模型;最后,仿真实例结果表明,利用分布式模型预测控制技术计算供应链中各阶段订货率,可以预测企业各阶段理想库存,从而快速响应市场动态变化,有效控制供应链管理中的不确定性。Aiming at characteristics of dynamic supply chain management under networked manufacturing, distributed model predictive control was applied to solve dynamic optimization control in supply chain. Firstly, implementing mechanism of dynamic supply chain was expatiated, and complete, stochastic, dynamic optimization models as well as the control constraints were proposed. Secondly, a distributed model predictive control algorithm was proposed, and a model of the optimal control sequence was established. Finally, the simulation example results show that the order rate of every stage can be calculated by distributed model predictive control, and the ideal stock of every stage in the enterprise can be predicted based on it, such can quick response to the dynamic changes of the market, and effectively control uncertainty in supply chain management.

关 键 词：网络化制造 供应链管理 分布式模型预测控制 动态双向耦合 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]