基于近红外光谱的材料粗糙度变化检测研究  

Detection of material roughness change based on near infrared spectroscopy

作　　者：鲁芬[1] 郁伯铭[2] LU Fen;YU Boming(Institute of Artificud,Intelligence,Wuchang Institute of Technologyt Wuhan 430000,China;School of Physics,Huazhong University of Science and Technology,Wuhan 430074,China)

机构地区：[1]武昌工学院人工智能学院,武汉430000 [2]华中科技大学物理学院,武汉430074

出　　处：《激光杂志》2022年第5期70-74,共5页Laser Journal

基　　金：湖北省教育厅科学研究项目(No.B2018323)。

摘　　要：为提高材料粗糙度变化检测的准确性,提出了一种基于近红外光谱的材料粗糙度变化检测方法。测量热红外偏振度,采用主成分分析方法降维处理。利用双温方程对光子照射到材料表面的条件下光子与电子及离子的能量转化情况分析,以对基体材料吸收率计算,等效处理材料表面粗糙度。通过双向分布函数构建粗糙度计算模型,完成材料粗糙度变化检测。实验将检测准确性与检测时间作为对比指标,结果表明,此次研究的检测方法有效提高了检测的准确度,平均为91%,并减少了检测的时间,平均为2.2 min。In order to improve the accuracy of material roughness change detection,a method of material roughness change detection based on near infrared spectroscopy is proposed. The polarization degree of thermal infrared was measured,and the dimension was reduced by principal component analysis. The energy conversion of photons,electrons and ions under the condition that photons irradiate the material surface is analyzed by using the two-temperature equation,so as to calculate the absorption rate of the matrix material and treat the material surface roughness equivalently. The roughness calculation model is constructed by bidirectional distribution function,and the change detection of material roughness is completed. The experiment takes the detection accuracy and detection time as comparative indexes. The results show that the detection method in this study can effectively improve the detection accuracy,with an average of 91%,and reduce the detection time,with an average of 2. 2 min.

关 键 词：近红外光谱 材料粗糙度 变化检测 等效处理 

分 类 号：TN929[电子电信—通信与信息系统]