多重滞后系统的预测控制  

Predictive control to multi-time delayed systems

作　　者：胡品慧[1] 闫峰[1] 

机构地区：[1]石油大学(北京)自动化研究所,北京102249

出　　处：《控制与决策》2004年第11期1294-1297,共4页Control and Decision

基　　金：国家863计划项目(2001AA413120);中国石油集团公司石油科技中青年创新基金资助项目(04E7027).

摘　　要：研究了过程控制中常见的多重滞后系统的预测控制问题.同时考虑状态变量滞后和控制作用滞后,提出了一种改进的状态反馈预测控制算法,它假设在滞后时间内状态变量未来预测值保持其当前值不变,给出了其简化的递推算法.解决了因多重状态滞后问题的存在,使状态变量的未来值利用模型直接递推很难得到的问题.最后通过仿真实例表明了算法的有效性.该算法计算量小、易于实际工程应用.The common control problem of multi-time delayed systems in process control is studied. A modified state feedback predictive control (SFPC) method is presented for systems with both multi-time delayed states and multi-time delayed inputs. As the result of the multi-time delayed states, the predictive values of state variables are very difficult to be obtained recursively. In the delayed time interval, it is supposed that predictive states keep the current values invariant in the prediction horizon, the recursive approach is given simply. It is easy to apply to the industrial process. The simulation examples show efficiency of the method.

关 键 词：预测控制 多重滞后系统 状态滞后 控制滞后 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]