基于加权最小二乘支持向量机的温室小气候建模与仿真  

Modeling and Simulation of Greenhouse Microclimate System Based on Weighted Least Squares Support Vector Machines

作　　者：李晋[1] 秦琳琳[1] 吴刚[1] 苑媛[1] 岳大志[1] 

机构地区：[1]中国科学技术大学自动化系,合肥230027

出　　处：《系统仿真学报》2008年第16期4232-4236,共5页Journal of System Simulation

基　　金：国家高技术研究发展计划(2006AA10Z253);新世纪优秀人才支持计划(NCET-04-0564);安徽省优秀青年科技基金资助项目(08040106910);中国科学院自动化研究所;中国科学技术大学智能科学与技术联合实验室自主研究课题:"温室小气候环境建模"

摘　　要：分析了温室小气候系统结构,采用加权最小二乘支持向量机回归方法在线建立温室小气候模型,并进行仿真研究,取得了较好的效果。最小二乘支持向量机中引入加权因子,使其回归估计对非高斯分布噪声及野点数据具有较好的鲁棒性。最后将此方法和带有智能监督级的渐消记忆递推增广最小二乘方法的在线建模及仿真结果进行了对比分析。The mechanism of the greenhouse microclimate system was analyzed, and then on-line modeling and simulation of the system were implemented via Weighted Least Squares Support Vector Machines Regression （WLS-SVMR）, which leads to satisfying results. By introducing weights factor to LS-SVMR, robust estimation for regression was obtained in the case of outliers and heavy tailed non-Gaussian error distributions. Finally, the results obtained here were compared with those by Gradually Oblivious Recursive Extended Least Squares （RELS） method with Intelligent Supervisory Segment.

关 键 词：温室 小气候 系统建模 支持向量机 加权最小二乘 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术] S625.5[自动化与计算机技术—计算机科学与技术]