球磨机多模型控制系统

提出了一种用于多变量模型不确定性控制的方法，克服了常规解耦设计没有考虑模型的不确定性，当对象模型摄动时解耦效果可能退化的缺陷。其要点是把整个工作范围按工作特性划分为若干个子区间，并以确定性模型表示。运用该概念，从对角优势化的定义出发，提出一种基于多模型的多变量控制系统逆奈奎斯特阵列设计方法，并应用于火电厂球磨机控制系统设计。从各模型对应的Gershgorin带及以单位阶跃作为输入的仿真结果可看出，所设计的多变量解耦控制系统具有较强的鲁棒性和满意的调节品质；且在不增加设备投资的情况下，可用现有设备实现所设计的系统。

MULTI-MODEL CONTROL IN BALL MILL PROCESS

This paper proposes a novel idea for controlling an object with multi-variable and uncertain model. It overcomes theshortcoming of conventional methods for designing decoupled networks that can't ensure good decoupling results when themodel is upset. Its main point is to divide the whole working region into several sub--regions. each of which is expressed witha certain model. Based on the above--mentioned concept and definition of diagonally dominant, an inverse Nyquist array designmethod is reckoned and used to design a ball mill control system for power plant. The Gershgorin belts of every model and thesimulation curves by the unit--step inputs show that the designed multi-variable decoupling control system has good robuststability and satisfactory regulation quality, and the control system can be designed with the equipment in hand.