预粉磨系统的智能建模与复合控制  

Intelligent modeling and compound control of pre-grinding system

作　　者：刘志鹏[1] 颜文俊[1] 

机构地区：[1]浙江大学电气工程学院,浙江杭州310027

出　　处：《浙江大学学报（工学版）》2012年第8期1506-1511,共6页Journal of Zhejiang University：Engineering Science

基　　金：国家自然科学基金资助项目(60574079);浙江省自然科学基金资助项目(601112)

摘　　要：为了提高水泥生产过程中预粉磨系统的运行效率,减少水泥生产过程的能源消耗.针对预粉磨系统的特点,对预粉磨系统进行深入分析,设计一种智能建模与复合控制系统来稳定和优化预粉磨系统的运行状况.将预粉磨系统分为2个不同的子系统,一个子系统通过模糊逻辑控制器(FLC)进行控制,另一个子系统通过基于线性矩阵不等式的模型预测控制方法(LMI-based MPC)进行控制,预测模型则是通过最小二乘支持向量机(LS-SVM)获得.最后通过OPC,DCS和C++等相关软件技术的支持实现了控制系统,并在工业现场应用,对预粉磨系统的运行进行控制.结果表明,控制系统可实现对预粉磨系统的优化控制,预粉磨系统运行的稳定性有了很大的提高.In order to improve the operational efficiency and energy saving in cement plants. Detailed analysis on the pre-grinding system according its features was given. An intelligent modeling and compoud control system was built to stablize and optimize the pre-grinding system in this paper. The whole system was devided into two sub-systems, one is controlled by fuzzy logic control （FLC）, and the other was handled by linear matrix inequality based model predictive control （LMI-based MPC）, and the model was acquired by using least square support vector machine （LS-SVM） regression. The control system was implanted with the support of OPC, DCS and C＋＋ techniques. The control system has been deployed in the field to control the pre-grinding system. The result shows that the pre-grinding system gets a good control from the control system, and pre-grinding system is more stable than what it was before.

关 键 词：预粉磨系统 模糊逻辑控制 最小二乘支持向量机 线性矩阵不等式 模型预测控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]