Study on PCA-LDA for Fast Identifying the Type of Coal Mine Water with LIF Technology

Abstract

Identifying the type of coal mine water is the foundation of coal mine hydrogeological research and has great significance for coal mine safe production. Considering the time consuming of the conventional method, we propose a new Laser Induced Fluorescence (LIF) combining with Principal Component Analysis and Linear Discriminant Analysis (PCA-LDA) method for coal source identification. Firstly, in order to obtain the valid bands from spectra results, LIF system is used to stimulate 405nm laser for the 400-800nm fluorescence spectra of tested water. Then, PCA is applied to reduce the dimension of spectral data. Finally, according to the different number of principal component with different pre-treatments the LDA identifications are implemented. Experiment results indicate that after valid bands selection, wavelet pre-processing, and PCA dimension reduction, the identification effect of spectra data with LDA method can be excellent as the number of principal component sets 6, and the correct recognition rate can reach to 100%. Thus, PCA-LDA algorithm with LIF tech is an effective identification method for quickly identification of the type of coal mine water.