Study Of The Spectral Data Processing In Laser Induced Breakdown Spectroscopy Technology

Laser-induced breakdown spectroscopy (LIBS), is based on emission spectroscopy by the atomic and laser plasma, which can analysis the sample ingredient by the data, and it also has many advantages, such as:easy, rapid, real-time., it can analyze multiple elements simultaneously, and don’t need for sample pretreatment which is the greatest feature. It is widely used in extraterrestrial probe material composition analysis, environmental pollution detection, cultural identification, biomedical,,as well as many other areas. In recent decades, the focus of LIBS technique research is still mainly focused on the experimental basis theory, optimizing experimental parameters, the different characteristics of different samples and the analytical accuracy and other issues. However, other more sophisticated spectroscopic techniques have also been confirmed that the spectral data processing is an effective means to improve the accuracy of spectral analysis. If we can achieve a better spectral signal preprocessing.On one hand, it can help to improve the technical specifications, reduce the complexity and cost of the system; on the other hand, you can also solve some technical problems can not be solved through optimizing the hardware, expansion the system suitability test under different environments. Therefore, as an important development direction of LIBS technology has attracted great attention of researchers.This article carry out in-depth theoretical and experimental studies for LIBS technique of spectral data processing techniques, focusing on the spectral de-noising, baseline correction of spectral image, substance classification and so on. The main work is completed as follows:Firstly, constructed experimental platform, to obtain a large number of spectral data of sample.Then, based on segmented spectral feature extraction and wavelet transform segmentation, respectively, for the spectral noise and baseline differences such two main factors affected segment (LIBS) signal quality, carried out pretreatment of spectral signal. Based on the laboratory LIBS experimental device, verified by experiments, the wavelet transform algorithm has draw the optimal algorithm parameters and eliminates the effect of noise, at the same time, the position of peak unchanged, does not affect the accuracy of analysis. According to background differences in the multi-channel spectrometer, put forward a method called segmented spectral feature extraction, effectively improve the baseline flatness LIBS spectroscopy signal.On the basis of the above work, artificial neural network based error back propagation is adopted to identify spectral line of the copper and stainless steel sample successfully.All the results illustrate that the utilization of multiple data processing method for spectral signal processing in LIBS technique can improve the quality of line’s analysis and recognition.