The Research Of Mineral Oil In Water Identification Technology And System Based On Fluorescence Detection

With the rapid development of industrial and agricultural production in China, thetransport and mining scale of oil in submarine and terrene continue enlarging, and the shipactivities become more and more frequent, these make the oil pollution condition of therivers, lakes, and coastal marine increases year by year, which impairs the waterecological balance and human health severely. Therefore, in order to master the dynamicchanging tendency of oil pollution, the water trace mineral oil concentration is neededonline test in real time, with the aim of management and control.Fluorescence spectroscopy is an important means for inspection by the petroleumcontaminants and hot research topics. The paper presents the research methods combinedwith fluorescence spectroscopy and parallel factor algorithm for the detection of mineraloil, and experiment to demonstrate it.Based on the principle that material can produce fluorescence by light excitation, thefeasibility of monitoring the content of mineral oil in water by fluorescence method isdemonstrated. It establishes the theoretical foundation for online monitoring to identify thedifferent categories of mineral oil in water.The paper made the analysis on the multi-component complex mixture withthree-dimensional fluorescence spectrometry, reconstructed the three-dimensionalfluorescence spectra and contour plots with Matlab then to extract feature information, andto determine the best excitation wavelength and emission wavelength of a variety ofmineral oil by characteristics of the crest of the three-dimensional fluorescence spectra,the peak number and the wavelength range.According to the Fluorescence characteristic and optical detection principle, we canboth design the optical system of excitation, collection and transmission and the weakfluorescence signal detection electronics system. The part of the optical design includesthe design of the light source, optical fiber and fiber optic and light source coupling. Theelectronics system mainly achieves the fluorescence signal detection and processing and todesign the circuit of the CCD spectrometer. The types of mineral oil in the mixture are analyzed by combining multivariatecalibration of parallel factor analysis and three-dimensional fluorescence spectral datamatrix. According to the spectral characteristics of the measured object, the number ofparallel factor is make base on the residual variance and nuclear consistent diagnosticmethod. With parallel factor analysis we can know the types of mineral oil in the mixture.After fitting the parallel factorization algorithm processing, spectra and experimentalspectra, the mineral oil content of the sample is estimated.