苏南地区外墙保温系统综合评价

摘要

以苏南地区外墙保温系统为背景，以技术、经济、施工、环境四个方面16个指标建立了外墙保温系统综合评价指标体系。采用有效解决小样本、非线性、基于结构风险最小化的支持向量机（Support Vector Machine，简称 SVM）评价分类模型，对样本训练集进行学习，使用网格交叉验证法选择最优模型参数，对测试样本进行测试，结果具有较高的精确度，研究成果在外墙保温系统中具有很好的应用价值。%Took the external wall thermal insulation system in Southern Jiangsu as the background, the comprehensive evaluation index system of exterior wall thermal insulation system was established with 16 indicators from aspects of technology,economy,construction and environment being taken into account. Through the evaluation classification model of nonlinear Support Vector Machine (SVM) based on structural risk minimization that can effectively solve the small samples,the author studied the sample training set, selected the optimal model parameters through the grid cross validation method,and tested the samples. The results were of high accuracy. Thus,the research results have very good application value in exterior wall thermal insulation system.