摘要
研究了闭环供应链网络动态系统的模型预测控制算法。通过建立系统的预测模型,给出状态估计和预测的计算公式。推导了多阶段最优预测模型的表达式,给出了目标函数和反馈校正方法。推导了基于二次优化的约束优化算法,并通过一个算例给出了这种算法的一般步骤。这种算法的本质是通过滚动优化和滚动实施,不断适时调整订货量、生产量等控制变量,使系统在稳定的前提下得到良好的动态性能,使牛鞭效应得到抑制。这种方法为处理复杂非线性的闭环供应链网络的稳定和协同控制,提供了一种新的方法。
A model predictive control algorithm of closed-loop supply chain networks dy-namic system is studied. Based on the established system prediction model,the calculation formula of state estimation and prediction are given;the multi-stage optimal prediction mod-el expressions are deduced;the objective function and the feedback correction method are given;the constrained optimization algorithm based on quadratic optimization is derived;and the general steps of this algorithm are given by a numerical example. The essence of this al-gorithm is that through the rolling optimization and continuous implementations,adjusting the order quantity,production control variables,the system has good dynamic performances in the premise of stability,the bullwhip effect is restrained. In order to get the stability and coordination control method of the complex nonlinear closed-loop supply chain networks,a new method is presented.
出处
《沈阳理工大学学报》
CAS
2017年第5期1-4,33,共5页
Journal of Shenyang Ligong University
基金
国家自然科学基金资助项目(70702036)
关键词
模型预测控制
闭环供应链网络
牛鞭效应
closed-loop supply chain networks
model predictive control
bullwhip effect
