基于DHP方法的锅炉燃烧系统优化控制研究  

Research on optimal control in boiler combustion system based on DHP

作　　者：史长城[1] 吴晓莉[1] 

机构地区：[1]郧阳师范高等专科学校物理与电子工程系,湖北丹江口442700

出　　处：《湖北大学学报（自然科学版）》2011年第4期460-466,共7页Journal of Hubei University：Natural Science

基　　金：湖北省教育厅科研项目(Q20105001)资助

摘　　要：电厂锅炉燃烧过程是一个典型的强非线性、多输入、多输出、强耦合过程,以这一直接影响机组安全经济运行的复杂过程控制为研究对象,研究火电厂锅炉燃烧系统的优化控制.首先,采用径向基函数神经网络(radial basis function neural network,RBFNN)对电厂锅炉燃烧系统进行了建模和模型测试.然后,用RBFNN设计了基于双启发式动态规划(dual heuristic programming,DHP)的电厂锅炉燃烧控制器,并在MATLAB环境下对所设计的DHP控制器进行了仿真试验,仿真结果表明利用这种算法设计的控制器实现了锅炉的稳定燃烧控制,且具有强鲁棒性.所给出的控制方法同样可以应用于其他复杂工业过程.Boiler combustion process for thermal power plant,which displays the features of strong non-linearity,multi-input,multi-output and close coupling control,is a typical researching object.Since the complex process control directly affects the security and economical operation of the units,this paper choosed the process as a researching object through probing into optimization control of boiler combustion system of thermal power plant.First,modeling to boiler combustion system based on RBFNN,and tested the model.Secondly,tried to achieve algorithmic derivation and program realization for applying RBFNN to design boiler combustion controller which was also under the direction of dual heuristic programming（DHP）optimization theory.Finally,the simulation experiments by MATLAB were made to indicate that the controller based on the algorithm was effective and the controller could achieve optimum ideal of combustion in the boiler with high efficiency and strong robustness.Eventually,the suggested system and used algorithm in the paper were also suitful to offer assistance for others complex industrial process.

关 键 词：径向基函数神经网络(RBFNN) 锅炉燃烧 双启发式动态规划(DHP) 

分 类 号：TG146.21[一般工业技术—材料科学与工程]