Quantitative & qualitative analysis of the LIBS spectrum using artificial neural networks

摘要：

LIBS analysis by measurement of the optical emission from a laser-induced plasma, is a method of monitoring chemical composition of materials. The assignment of different atomic and ionic lines, which is signatures of a particular element, is the basis of identification of the species present in plasma.In traditional calibration methods we need to have exact information of the element concentration. This is a limitation for unknown samples. In this article we are going to show the potential of Neural Networks in classification and prediction of the concentration of elements from the spectrum signal. It has been shown that if a spectrum of unknown sample is presented to the network, the elemental composition is identified in less than a few msec.

关键词： Artificial Neural Networks Calibration Curves LIBS Plasma Spectrum

作者: Asiyeh Moosavi Hossein Saghafifar Zahra Zahedi

作者单位: Mobin Petrochemical Company, Assaluyeh, Iran Department of Laser & Optics, Malek Ashtar University o Department of Laser & Optics, Malek Ashtar University of Technology, Isfahan, Iran Department of Laser & Optics, Malek Ashtar University of Technology, Isfahan, Iran SA Iran Electro o

会议类型: 国际会议

会议名称: 2011亚洲光纤通信与光电国际会议

会议地点: 上海

会议语种:英文

页码: 1-3

在线出版日期: 2011-11-13（万方平台首次上网日期，不代表论文的发表时间）