化工过程强非线性系统的变模型自适应预测控制  

Varying Model Based Adaptive Predictive Control of Highly Nonlinear Process in Chemical Process

作　　者：许锋[1] 张艺[1] 罗雄麟[1] 季德伟 陈育昆 

机构地区：[1]石油大学自动化研究所,北京昌平102249 [2]中国石油大港石化分公司,天津300280

出　　处：《化工自动化及仪表》2004年第6期32-35,共4页Control and Instruments in Chemical Industry

摘　　要：提出一种变模型自适应预测控制算法 ;基于非线性状态空间模型 ,通过每步在当前工作点 (非平衡点 )线性化获得线性化子模型 ,以此进行状态反馈预测控制 ,线性化子模型随工作点变化 ,且不限于平衡点。通过pH值控制的对比仿真实验 ,证明其对强非线性过程的控制效果优于传统的多模型预测控制。最后分析讨论了该控制算法存在的几个重要问题 。A varying model based adaptive predictive control algorithm is proposed. A nonlinear state space model is linearized on each nonequilibrium operating position every step, the acquired linearized submodel is applied to state feedback predictive control, and the linearized submodel is automatically adjusted according to the move of operating position, which is not limited to the equilibrium position. Through the simulation of pH control, compared with traditional multiple model predictive control, its better effect on highly nonlinear processes is showed. Finally several problems of this method are analyzed and discussed, and the concerned future research aspects are proposed.

关 键 词：强非线性 变模型 自适应预测控制 线性化 PH值控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]