Research On Rapid Identification And Classification Of Common Explosives Based On Infrared And Raman Spectroscopy

With the increasingly frequent activities of international terrorist organizations,domestic and foreign terrorist incidents have occurred repeatedly.Explosives have become the preferred attack tool for terrorists due to their strong destructive power,significant harm,and wide impact range.Due to the great harm of such incidents,explosive detection and identification technology has been highly valued by governments and relevant institutions in various countries,and related research has become one of the research focuses of current international attention.For the detection of explosives,spectral detection technology has attracted widespread attention due to its advantages such as fast detection speed,high accuracy,low cost,and nondestructive testing.This article focuses on six common explosives and explores the application of infrared and Raman spectroscopy in the identification and classification of explosives.The research content and main achievements are as follows:(1)Analyzed and established a fast identification and classification model for common explosives based on a single infrared spectrum.Three preprocessing methods,first derivative(FD),standard normal transformation(SNV),and multiple scattering correction(MSC),were used to process the collected infrared spectra,and then the preprocessed infrared spectra were processed by successive projections algorithm(SPA)and competitive adaptive reweighted sampling(CARS)to extract feature variables,and finally a classification model was built using support vector machine(SVM).The results show that the SNV-SPA model has the best prediction performance,with a test set accuracy of 100%.(2)Analyzed and established a rapid identification and classification model for common explosives based on a single Raman spectrum.The asymmetric least squares method(ALS)was used to perform baseline correction on the collected Raman spectra to eliminate background noise,and the moving average(MA)and Savitzky-Golay(SG)smoothing algorithms were used to smooth the spectra,and then the preprocessed Raman spectra were processed by SPA and CARS algorithms to extract feature variables,and finally uses support vector machine(SVM)to establish a classification model.The results showed that the ALS-SG-SPA model has the best prediction performance,with a test set accuracy of 98.33%.(3)Analyzed and established a rapid identification and classification model for common explosives based on data fusion technology.Among the models established by a single spectrum,only the infrared SNV-SPA model has a prediction accuracy of 100%,while the other models have lower prediction accuracy.In order to improve the accuracy and generalization ability of the model,this paper establishes a multi-source spectral fusion prediction model for explosive species recognition based on infrared spectroscopy and Raman spectroscopy at the data and feature layers.The results show that data fusion technology has a significant effect on improving the prediction accuracy of explosive classification models,especially for explosive SVM classification models established at the feature layer level,all models have a prediction accuracy of 100%.On the basis of single spectral detection,this paper applies data fusion technology to explosive identification and classification,achieving good results and enhancing the accuracy and robustness of model prediction,which is of great significance.