电厂湿法烟气脱硫过程优化控制研究  

Research on Optimal Control of Power Plant Wet Flue Gas Desulphurization Process

作　　者：史运涛[1] 孙德辉[1] 李正熙[1] 李志军[1] 年四成[1] 高东杰[2] 

机构地区：[1]北方工业大学北京市现场总线与自动化重点实验室,北京100041 [2]中国科学院自动化研究所,北京100080

出　　处：《系统仿真学报》2007年第22期5287-5290,共4页Journal of System Simulation

基　　金：中国科学院知识创新工程重大项目(KGCX-SW-15)

摘　　要：深入研究了具有混杂系统特性的电厂湿法烟气脱硫(WFGD)过程的优化与控制问题;运用混合逻辑动态的建模方法(Mixed Logical Dynamic)建立了WFGD过程的MLD模型,基于MLD模型的预测控制方法把WFGD过程优化问题转化为实时求解混合整数规划(MIQP)问题,实时给出了WFGD过程优化运行的连续决策变量(吸收塔浆液的PH值设定)和离散决策变量(循环浆液泵开启台数)的最优解;用同样方法实现WFGD过程基础回路级中具有非线性动态的PH对象的控制;仿真表明基于MLD模型预测控制方法很好地解决了WFGD过程的优化与控制问题。The optimization and control problem of Wet Flue Gas Desulphurization in power plant was studied. Mixed logical dynamics model was built to describe hybrid phenomenon of the WFGD process. Based on the mixed logic dynamics model of WFGD, the predictive control optimization problem was solved using the Mixed Integer Quadratic program （MIQP） solver Therefore the optimal discrete decision variable （the number of LG pump） and the continuous variable decision variable （PH set-point） were obtained on line. The real-time optimization was implemented. An MLD model of PH was also developed based on which the predictive controller of PH was designed. The simulation results indicate that the predictive control based on MLD model scheme presented has good effect.

关 键 词：湿法烟气脱硫 混合逻辑动态模型 预测控制 混合整数规划(MIQP) 优化控制 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]