Rapid detection and recognition of phosphors using laser-induced breakdown spectroscopy and principal component analysis method-back propagation neural network algorithm

Rapid detection and quality monitoring of phosphor materials have always been a difficult problem in phosphor materials market. In this work, an independently proposed method based on principal component analysis method-error back propagation neural network algorithm-laser induced breakdown spectroscopy (PCA-BPNN-LIBS) was used for the detection and recognition of phosphors. Firstly, spectroscopic study was carried out on phosphor material samples, and the composition of phosphor elements was analyzed according to the full emission spectrum. Spectral data with different element characteristics detected by LIBS were used as training data sets for further identification. Then PCA method and BPNN algorithm were applied to identify 4 types phosphor samples (P11, P20, P43, P46). A very clear distinction graph was obtained, and the classification accuracy of 99.93% was verified. Allresults show that the proposed PCA-BPNN-LIBS method is an effective method for rapid analysis and recognition of phosphors.