水泥分解炉出口温度HDP优化控制研究  

Research on HDP-based optimal control in calciner outlet temperature

作　　者：史长城 田森平[2] 

机构地区：[1]汉江师范学院汽车与电子工程系,湖北十堰442000 [2]华南理工大学自动化科学与工程学院,广东广州510640

出　　处：《湖北大学学报（自然科学版）》2017年第5期558-562,共5页Journal of Hubei University：Natural Science

基　　金：国家自然科学基金(60964002);湖北省教育厅科学技术研究计划优秀中青年人才项目(Q20105001)资助

摘　　要：水泥分解炉出口温度是一个典型的非线性、多输入、强耦合过程,它直接影响水泥生产的产量、质量和能耗.本文中以这一复杂工业控制过程为研究对象,研究水泥分解炉出口温度的优化控制问题.先建立分解炉出口温度的径向基函数神经网络(radial basis function neural network,RBFNN)模型;然后在此基础上设计分解炉出口温度启发式动态规划(heuristic dynamic programming,HDP)控制器,并在MATLAB环境下对所设计的HDP控制器进行了仿真实验,实验结果表明该HDP控制器实现分解炉出口温度的稳定控制.该控制算法适用于其他类似的工业控制过程,具有一定的参考借鉴意义.Calciner outlet temperature,which features a typically nonlinear process with multiple input and strong coupling,directly affects the production of cement production,quality and energy consumption. We choosed the complex industry control process as a researching object through probing into optimized control of calciner outlet temperature. In the present paper,therefore,firstly contructs modeling to calciner outlet temperature based on RBFNN. Secondly,the authors attempt to achieve algorithmic derivation and program realization for applying RBFNN to designed calciner outlet temperature controller which as well works under the direction of Heuristic Dynamic Programming（ HDP） optimization theory. In the light of the simulation experiments by MATLAB,it is made to indicate that the controller based on the algorithm tends to be effective and the controller can reach to stability control of calciner outlet temperature. Eventually,we manages to denote that the suggested system and the used algorithm are proved to be feasible with respect to offering assistance for relevant complex industrial process.

关 键 词：径向基函数神经网络 分解炉出口温度 启发式动态规划(HDP) 

分 类 号：TG146.21[一般工业技术—材料科学与工程]