摘要
将最小二乘支持向量机法(LS_SVM)应用于中红外光谱分析,建立一种新的对常见废弃塑料进行分类的方法。依据不同类别的塑料在红外波段具有不同的特征吸收峰,采用LS_SVM方法对塑料的中红外光谱数据进行处理,并与全局相关法、系统比较法进行比较。实验结果表明,LS_SVM进行分类的正确率为0.92。与全局相关法和系统比较法相比,LS_SVM分类正确率较高,其解决小样本问题效果显著,可应用于常见废弃塑料的分类。
The least squares support vector machines method （LS_SVM） is used in the mid-infrared spectroscopy analysis of plastics to establish a new classification method of common waste plastics. According to different absorption peak positions of different plastic types in the mid-infrared bands, the infrared spectral data are processed by LS_SVM method. The result is compared with the results of the overall correlation method and the system comparison method. Experimental results show that classification accuracy rate of LS_SVM method is 0.92,which is higher than those of the other two metboces. And the effect of LS_SVM method solving the problem of small samples is remarkable. So it can be applied into the classification of common waste plastics.
出处
《北京信息科技大学学报（自然科学版）》
2011年第1期89-92,共4页
Journal of Beijing Information Science and Technology University
基金
现代测控技术教育部重点实验室资助项目
关键词
最小二乘支持向量机
塑料分类
中红外光谱分析
least squares support vector machines
classification method of common waste plastics
the mid-infrared spectroscopy analysis
