一种改进型广义预测控制系统研究及工程应用  

Research and Engineering Application of an Improved Generalized Predictive Control System

作　　者：李泉[1] 尹峰[1] 孙坚栋[1] 罗志浩[1] 

机构地区：[1]国网浙江省电力公司电力科学研究院,杭州310014

出　　处：《浙江电力》2017年第8期44-48,共5页Zhejiang Electric Power

基　　金：国网浙江省电力公司科技项目(5211DS1350JZ)

摘　　要：常规的广义预测控制算法在对象模型已知时,通过求解丢番图方程获得最优控制律,但是对模型精确度要求较高,不能适应工程需要,因此,提出一种改进型广义预测控制系统,当已知模型的参数变化时仍具有良好的控制性能。为了提高系统克服外扰的能力,还设计了扰动通道算法,形成带扰动模型的广义预测控制系统,将该系统应用于某600 MW超临界机组的脱硝控制系统中,通过建立系统对象模型和扰动模型来预测被控对象未来的变化趋势,当对象模型发生变化时,系统仍具有较强的鲁棒性。The optimal control law is obtained by solution of Diophantine equation in the case of a given object model. However, the requirement on model accuracy is so stringent that it can not cater to engineering use. This paper proposes an improved generalized predictive control system that owns control performance in the case of model parameter change, In order to improve the ability of the system to overcome the external disturbance, the disturbance channel algorithm is designed to form a generalized predictive control system with disturbance model. The system is applied to the denitrification control system of a 600 MW supercritical unit to predict the future trend of the controlled object through establishment of the system object model and the disturbance model; in the case of object model change, the system is still strongly robust.

关 键 词：广义预测 鲁棒 扰动 脱硝控制 模型 

分 类 号：TM621[电气工程—电力系统及自动化] TP273[自动化与计算机技术—检测技术与自动化装置]