基于最小二乘支持向量机的三段膜分离模型
作者:李桂香1,王  磊13,王元麒2,李继定3
单位: 1海南大学信息科学技术学院, 海口 570228;2大连育明高中, 大连 116023;3清华大学化学工程系化学工程国家重点联合实验室, 北京 100084
关键词: 气体膜分离技术;最小二乘支持向量机;三段;在线检测;实时优化控制
出版年,卷(期):页码: 2013,33(6):71-77

摘要:
提出基于三段膜分离过程的智能模型,并应用它在线分析炼厂气氢回收过程中的关键性能参数。首先,应用网格搜索和交叉验证,结合贝叶斯估计得到最小二乘支持向量机的sig2和gam参数的最优值;然后,建立基于最小二乘支持向量机的三段氢回收膜分离过程模型;最后,基于Matlab2010a软件平台和现场数据编程建模,对炼厂气氢回收过程中的关键性能参数进行在线预测分析。仿真结果表明,模型正确合理、预测速度快,其预测值和实际测量值基本吻合,误差小,可以很好地反映出三段膜膜组件良好的分离性能,对气体膜分离过程中的参数在线检测和过程实时优化控制具有一定的指导意义。
A three-stage intelligent model of gas membrane separation process was proposed, and was applied to analysis the key performance parameters of hydrogen recovery membrane separation process in real time. Firstly, combined grid search and cross validation with bayes estimation were used to obtain the optimal value of two important parameters (i.e.,sig2 and gam ) of least squares support vector machine; then, three-stage model of hydrogen recovery membrane separation process based on least squares support vector machine was built. Finally, modeling program was wrote based on Matlab2010a and field data, and the key performance parameters of hydrogen recovery membrane separation process was predicted and analysis on-line. The simulation results show that the model is reasonable, its convergence speed is very fast, and the prediction results of the model are in good agreement with the measurement values with reasonable errors. It well reflects the good separation performance of the membrane module of the three-stage membrane process. This study has a great significance for the research of on-line detection of important performance parameters and its optimal control in the gas membrane separation process.
李桂香(1988-),女,湖南永州人,硕士, 研究方向:智能检测。通讯作者:王磊(1966-),男, 辽宁大连人,博士, 高级工程师, 研究方向:绿色分离过程与优化控制系统、低碳节能技术及产业化。E-mail: wanglei0520@126.com

参考文献:
[1] Koros W J. Some opportunities & challenges for our membrane community to consider-with an Emphasis on gas separations [J]. Membrane Science and Technology, 2006, 26(4): 1~5.
[2] 张广信, 郑邦婞, 于京鑫, 等. 膜分离技术用于气体脱湿的研究现状[J]. 化工科技, 2010, 18(5): 73~76.
[3] 陈麟风. 膜分离技术概述[J]. 能源与环境, 2011, 2: 91.
[4] 王建宏, 陈家庆, 曹建树. 加油站膜分离烃类VOCs回收技术分析[J]. 膜科学与技术, 2009, 29(3): 93~97.
[5] 孙翀, 李洁, 孙丽艳, 等. 气体膜分离混合气中二氧化碳的研究进展[J]. 现代化工, 2011, 31(1): 19~23.
[6] 张菀乔, 张雷, 廖礼, 等. 气体分离膜技术的应用[J]. 天津化工, 2008, 22(3): 21~22.
[7]  Wang Lei, Shao Cheng, Wang Hai,et al. Radial Basis Function Neural Networks-Based Modeling of the Membrane Separation Process: Hydrogen Recovery from Refinery Gases [J]. Journal of Natural Gas Chemistry, 2006, 15(3): 230~234.
[8] 李桂香, 王磊, 李继定, 等. 基于主元分析的气体膜分离过程RBFNN建模[J]. 系统仿真学报, 2012, 24(9): 2003~2006.
[9] 王磊. 膜分离过程的优化与控制方法研究[D]. 大连: 大连理工大学, 2006.
[10] Suykens J A K, Vandewalle,J. Least squares support vector machine classifiers, Neural processing letters, 1999, 9: 293~300.
[11] 陶少辉. 最小二乘支持向量机的改进及其在化学化工中的应用[D]. 浙江: 浙江大学, 2006.
[12] 黄银蓉, 张绍德. MIMO最小二乘支持向量机污水处理在线软测量研究[J]. 自动化与仪器仪表, 2010, 4: 15~17.
[13] 张宏建, 孙志强. 现代检测技术[M]. 北京: 化学工业出版社, 2009: 59~68.
[14] Shang Tianfeng, Geng Zhiqiang. Online Operation Optimization based on GA-RBF Network in Ethylene Cracking Furnace [C]// IEEE, 2010. USA: IEEE, 2012: 624~627.

服务与反馈:
文章下载】【加入收藏

《膜科学与技术》编辑部 地址:北京市朝阳区北三环东路19号蓝星大厦 邮政编码:100029 电话:010-64426130/64433466 传真:010-80485372邮箱:mkxyjs@163.com

京公网安备11011302000819号