焦炉集气管压力状态空间建模与多模型预测控制  

State Space Modelling and Multi-model Predictive Control for Gas Collector Pressure of Coke Oven

作　　者：刘昕明[1] 吕亮 王威 LIU Xin-ming;LU Liang;WANG Wei(College of Electrical and Control Engineering,Liaoning Technical University,Huludao 125105,China;Third Institute,China Aerospace Science&Industry Corp,Beijing 100013,China)

机构地区：[1]辽宁工程技术大学电气与控制工程学院,辽宁葫芦岛125105 [2]中国航天科工集团公司第三研究所,北京100013

出　　处：《辽宁工业大学学报（自然科学版）》2020年第5期281-285,共5页Journal of Liaoning University of Technology(Natural Science Edition)

基　　金：辽宁省教育厅科学研究一般项目(LJYL013);辽宁省重点研发指导计划(2019JH8/10100050)。

摘　　要：依据焦炉集气管压力系统物理特性建立集气管压力的机理模型,将整个焦炉集气过程分为结焦初期落煤、结焦中期和结焦末期3种工况,然后非线性机理模型在这3个平衡点处线性化。再将线性化模型转化为状态空间预测模型,最后进行多模型预测控制器的设计。通过对模型的动态过程进行仿真验证,证明模型合理性。控制仿真实验表明,在焦炉生产的不同工况,采用不同的预测模型明显强于目前采用的单一控制率方法。According to the physical characteristics of the coke oven gas collection pressure system,the mechanism model of gas collection pressure is established.According to the production process,the whole coke oven gas collection process can be divided into three working conditions:coal falling in the early stage,middle stage and end stage of coking.The nonlinear mechanism model is linearized at these three equilibrium points.The linearized model is transformed into a state space prediction model,and then a multi-model predictive controller is designed.The simulation tests of the dynamic processes of the models are made to prove reasonableness of the model.The results demonstrate that using different predictive models has more favorable dynamic characteristics than the current control method of a single rate under the different working conditions of the production of coke oven.

关 键 词：焦炉集气管压力 状态空间模型 多模型预测控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]