基于约简鲁棒LSSVM铝酸钠浓度软测量

系统仿真学报 ›› 2015, Vol. 27 ›› Issue (9): 2203-2207.

基于约简鲁棒LSSVM铝酸钠浓度软测量

孙荣玲¹, 钱晓山^1,2

1.宜春学院物理科学与工程技术学院，江西宜春 336000;
2.中南大学信息科学与工程学院，湖南长沙 410083

收稿日期:2015-04-06 修回日期:2015-07-03 出版日期:2015-09-08 发布日期:2020-08-07
作者简介:孙荣玲(1967-),女,江西宜春,研究方向为智能控制；钱晓山(1980-),男,江西九江,博士,讲师,研究方向为复杂过程的建模、优化与控制。
基金资助:
国家自然科学基金项目(60634020, 60874069, 60804037, 51366013)；国家863项目(2006AA04Z181)；宜春学院科研课题(XJ1314)；

Soft Sensor of Concentration of Sodium Aluminate Solution Based on Reduction Robust LSSVM

Sun Rongling¹, Qian Xiaoshan^1,2

1. Physical Science and Technology College, Yichun University, Yichun 336000, China;
2. School of Information Science & Engineering, Central South University, Changsha 410083, China

Received:2015-04-06 Revised:2015-07-03 Online:2015-09-08 Published:2020-08-07

摘要/Abstract

摘要： 针对铝酸钠溶液浓度在线检测仪表稳定性差、具有放射性、维护保养成本高等不足及人工检测严重滞后的问题, 结合蒸发过程工艺机理分析,选取影响铝酸钠溶液浓度的参数为辅助变量,采用加权损失函数的最小二乘支持向量机(Least Squares Support Vector Machine,LSSVM)实现了铝酸钠溶液浓度的鲁棒软测量。并采用斯密特正交化方法约简核矩阵,降低计算复杂度。工业过程数据仿真结果表明,建立的软测量模型能够连续在线检测铝酸钠溶液浓度,并获得了比标准LSSVM、加权LSSVM及多核LSSVM更高的预测精度,完全满足工业要求。

关键词: 差分进化算法, 最小二乘支持向量机, 铝酸钠溶液, 鲁棒

Abstract: Taking into account of poor stability, radioactive, higher maintenance costs and lack of manual detection of serious lag issues for online testing instrument of sodium aluminate concentration, combined with the mechanism analysis of the evaporation process, selecting parameters that affect the concentration of sodium aluminate solution as the auxiliary variables, using weighted loss function of the least squares support vector machine (Least Squares Support Vector Machine, LSSVM), a robust soft measuring for the concentration of sodium aluminate solution was achieved. Schmidt method was used to simplify the matrix and reduce the computational complexity. Industrial process data simulation results show that the soft measurement model can detect continuously sodium aluminate solution concentration online, and receive more than standard LSSVM, weight LSSVM higher prediction accuracy, and fully meet the industrial requirements.

Key words: DE algorithm, least squares support vector machine, Sodium aluminate Solution, Robust

中图分类号:

TP18

孙荣玲, 钱晓山. 基于约简鲁棒LSSVM铝酸钠浓度软测量[J]. 系统仿真学报, 2015, 27(9): 2203-2207.

Sun Rongling, Qian Xiaoshan. Soft Sensor of Concentration of Sodium Aluminate Solution Based on Reduction Robust LSSVM[J]. Journal of System Simulation, 2015, 27(9): 2203-2207.

参考文献

[1] 李修亮, 苏宏业, 褚健. 基于在线聚类的多模型软测量建模方法 [J]. 化工学报, 2007, 58(11): 2834-2839.
(Li Xiuliang, Su Hongye, Zhu Jian. Multiple models soft-sensing technique based on online clustering arithmetic [J]. Journal of Chemical Industry and Engineering, 2007, 58(11): 2834-2839.)
[2] 陈定三, 杨慧中. 基于局部重构融合流形聚类的多模型软测量建模 [J].化工学报, 2011, 62(8): 2281-2285.
(Chen Dingsan, Yang Huizhong. Multiple model softsensor based on local reconstruction and fusion manifo-ld clustering [J]. Journal of Chemical Industry and Engineering, 2011, 62(8): 2281-2285.)
[3] Suykens J A K, Vandewalle J. Least squares supp-ort vector machine classifiers [J]. Neural Processing Letter (S1370-4621), 1999, 9(3): 293-300.
[4] 阎威武, 邵惠鹤. 支持向量机和最小二乘支持向量机的比较及应用研究 [J]. 控制与决策, 2003, 18(3): 358-360.
(Yan Weiwu, Shao Huihe. Application of support vect-or machines and least squares support vector machines to heart disease diagnoses [J]. Control and Decision, 2003, 18(3): 358-360.)
[5] Suykens J A K, De Brabanter J, Lukas L, et al. Weighted leasts quares support vector machines: robustness and sparse approximation [J]. Neuralcomputing(S0925-2312), 2002, 48: 85-105.
[6] Yu Runqiao, Dang Lei. Welding defect recognition based on gray correlation analysis in ultrasonic testing [J]. Chinese Journal of Scientific Instrucment(S0254-3087), 2006, 27(10): 1358-1360．
[7] Huang Liang, Hou Jianjun. Algorithm of selecting optimal test point for analog circuit based on gray relational entropy [J]. Journal of Electronic Measurement and Instrument(S1000-7105), 2009, 23(9): 27-32.
[8] Zhang Yingwei. Enhanced statistical analysis of non-linear processes using KPCA, KICA and SVM [J]. Chemical Engineering Science(S0009-2509), 2009, 64(5): 801-811.
[9] Valyon J, Horvath G. A sparse least squares suppor-t vector machine classifier [C]// Proceedings of the IEE-E International Joint Conference on Neural Networks. Piscataway, NJ, USA: IEEE, 2004: 543-548.
[10] 阳春华, 钱晓山, 桂卫华. 一种混沌差分进化和粒子群优化混合算法 [J]. 计算机应用研究, 2011, 28(2): 439-441.
(Yang Chunhua, Qian Xiaoshan, Gui Weihua. Hybrid algorithm of chaotic differential evolution and particle swarm optimization [J]. Application Research of Computers, 2011, 28(2): 439-441.)

[1]	王丽梅, 么红岩, 张康. 直接驱动H型平台鲁棒最优同步控制仿真研究[J]. 系统仿真学报, 2022, 34(3): 527-535.
[2]	路宏广, 赵树恩. 基于鲁棒模型预测的智能汽车轨迹跟踪控制研究[J]. 系统仿真学报, 2022, 34(1): 153-162.
[3]	余伟, 梁恒辉, 罗映. 一种改进差分进化算法的分数阶系统辨识研究[J]. 系统仿真学报, 2021, 33(5): 1157-1166.
[4]	邵诚, 孙宏远, 张林. 基于PSO-LSSVM的反舰导弹自控弹道仿真[J]. 系统仿真学报, 2021, 33(4): 918-926.
[5]	尤一琛, 王艳, 纪志成. 基于博弈论的柔性作业车间动态调度研究[J]. 系统仿真学报, 2021, 33(11): 2579-2588.
[6]	梅盼盼, 吴亮红, 张红强, 王慧莹. 含风电电力系统动态经济调度的差分纵横交叉优化[J]. 系统仿真学报, 2020, 32(6): 1179-1187.
[7]	钟麟, 佟明安, 张健, 梁建海. 基于效果作战的武器系统组合效费分析[J]. 系统仿真学报, 2020, 32(4): 718-726.
[8]	王哲, 李建华, 康东. 基于极大连通子图边界的复杂网络恢复研究[J]. 系统仿真学报, 2020, 32(12): 2306-2316.
[9]	王艳, 丁宇. 动态柔性作业车间优化调度与决策方法[J]. 系统仿真学报, 2020, 32(11): 2073-2083.
[10]	於万里, 王艳, 纪志成. 氨糖发酵过程建模与工艺参数优化研究[J]. 系统仿真学报, 2020, 32(10): 1895-1902.
[11]	张倩, 熊英, 何明珂, 张浩. 不确定需求生鲜电商配送路径规划多目标模型[J]. 系统仿真学报, 2019, 31(8): 1582-1590.
[12]	李从东, 李文博, 曹策俊, 王玉. 面向级联故障的相依网络鲁棒性分析[J]. 系统仿真学报, 2019, 31(3): 538-548.
[13]	林雨谷, 王艳. 离散车间能效数据挖掘及调度优化[J]. 系统仿真学报, 2019, 31(12): 2702-2711.
[14]	王杰, 陈彬, 袁鹏, 马亮, 邱晓刚. 数据驱动的最优互惠避碰模型偏好速度研究[J]. 系统仿真学报, 2019, 31(12): 2731-2739.
[15]	叶恒舟, 李陶深, 关云慧. 组合服务的时序鲁棒性仿真分析[J]. 系统仿真学报, 2018, 30(9): 3399-3403.