阶梯式多变量DMC分散优化算法及工程化实现  

Stair-like Multivariable Decentralized Optimization DMC Algorithm and Engineering Implement

作　　者：薛福珍[1] 崔连喜[1] 赵灿[1] 李世剑[1] 唐雷[1] 

机构地区：[1]中国科学技术大学自动化系,合肥230027

出　　处：《化工自动化及仪表》2007年第4期30-33,共4页Control and Instruments in Chemical Industry

基　　金：安徽省发改高技项目(2005661)

摘　　要：提出基于状态空间模型的阶梯式多变量动态矩阵控制分散优化算法,使计算量大大减少,并在自行研制的先进控制平台上,实现了这种算法。该算法的工程化实现方法,使用户可以方便地构建自动控制系统并实施多变量动态矩阵控制。对双容水箱的控制实验表明,该控制算法对具有多变量、时滞、耦合和不确定性的复杂对象,具有良好的控制效果。A stair-like decentralized optimization dynamic matrix control （DMC） algorithm tor multivariable systems based on state-space models was proposed. The new algorithm can reduce the calculation significantly and was implemented in the advanced control platform self-developed. The algorithm implement method in engineering allows user can easily build an automatic control system, which operates on stair-like multivariable DMC （StairMDMC）. The control experiments of tank with double columns show that the StairMDMC has excellent effects to the complicated plants with multivariable, time delay, coupling and uncertainty.

关 键 词：阶梯式 MDMC 分散优化 先进控制平台 动态链接库 控制回路模板 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]