基于LSSVM的真空玻璃传热过程建模  

Modeling of heat transfer process for vacuum glazing based on LSSVM

作　　者：王元麒[1,2] 王磊[1] 李桂香[1,3] 

机构地区：[1]海南大学信息科学技术学院海南省特种玻璃重点实验室,海南海口570228 [2]米兰理工大学,意大利米兰20133 [3]广东省医用电子仪器及高分子材料制品重点实验室,广东广州510500

出　　处：《真空》2015年第2期57-60,共4页Vacuum

基　　金：国家自然科学基金(61463011);国家科技支撑计划课题(2012BAA10B03);广州市科技计划项目(201334500076)

摘　　要：真空玻璃传热过程隔热性能和质量评价关键指标-传热系数(U值)受真空度影响,难以精确检测。为实现真空玻璃隔热性能和质量的快速评定,建立了基于LSSVM的真空玻璃传热过程智能模型,应用MATLAB软件和实验数据,对真空玻璃传热过程进行了模拟,在线预测了真空玻璃传热后侧(非热源侧)的中心温度。结果表明:模型理论正确,预测迅速(时间<1 s),误差小(相对误差<1.2%),为真空玻璃隔热性能和真空玻璃质量的快速评定提供了一种新的技术手段。同时,也为后续研究能将温度参数预测转化为传热系数预测奠定一定的理论和应用基础。The key indicator-heat transfer coefficient （U value ） used to evaluate the heat insulation performance and quality of vacuum glass is hard to detect accurately, for influenced by vacuum degree. To realize a quick assessment of the vacuum glass heat insulation performance and quality, an intelligent model of vacuum glass heat transfer process based on LSSVM was established. And MATLAB software and the experimental data was used to simulate the vacuum glass heat transfer process and predict the center temperature in the back side （non-heat source side ） of vacuum glass online. Results show that the model is correct, the predict speed is fast （〈1 s） and the predict error is small （the relative error 〈1.2%）. This provides a new method to evaluate the vacuum glass heat insulation performance and the quality of vacuum glass. Meanwhile, a certain theoretical and application basis was provided for researching the vacuum glass heat insulation detector which can convert temperature to U value.

关 键 词：真空玻璃 传热过程 LSSVM 智能建模 性能检测 

分 类 号：TB771[一般工业技术—真空技术]