Study On Oilborne Organic Detection Technology Based On Fluorescence Spectra

With the development of economy and scale of oil exploitation,the produced and living waste water also increased.The organic composition of the oil in the effluent is a serious threat to the balance of ecosystem and human health.Therefore,the rapid and accurate detection of oil pollutant composition and content,such as gasoline,diesel,kerosene,is of great significance for the environment monitoring and pollution control.The article is based on the theory of fluorescence luminescence and measurement technology,on the basis of deep research on the composition of petroleum organic compounds,optical fiber sensor and three-dimensional fluorescence spectroscopy,these composed of optical measurement section and signal processing section.Analyze the characteristics of the experimental instrument and the influence of the scattered light interference,take into account the sample excitation and emission measurement range.The multi-component concentration experiments were carried out on the gasoline,diesel,kerosene and other oil organic matter,and obtain the corresponding sample fluorescence spectra.The fluorescence spectra of the three kinds of petroleum samples are analyzed based on the number,location and intensity information of each fluorescence peak,verify the feasibility of detecting the organic pollutants of oil in water through fluorescence analysis.The mineral oil spectral data were measured by full-featured fluorescence spectrometer.The feature parameters are selected by MATLAB to compose the original feature vector.The principal component analysis(PCA)is used to reduce the input vector of the neural network.The Extension Neural Network method unified the training criterion and the classification criterion for qualitative analysis and identification of oil organic matter.The results show that PCA-ENN algorithm has higher recognition accuracy and efficiency,and this algorithm can also be used to identify other areas of organic substances.