Study Of The Identification And Analysis System For Mineral Oil In Water Based On Fluorescence Mechanism

As demand for the amount of oil rising, making for oil exploration, extraction,transportation, processing, and other activities become more frequent, which results in a large number of leaks and emission into coastal waters that poses a serious threat to the ecological balance. Trace mineral oil for the type and concentration of real-time online testing has become a key of environmental regulation and governance.The article regards fluorescence spectral analysis as a major research tool, and by the research for complexity components of mineral oil, completes the analysis and feasibility studies to species identification and concentration measurement of mineral oil by the combination method of fluorescence spectral analysis and the parallel factor analysis and uses wavelet entropy threshold denoising algorithms to undo noise.On the basis of mineral oil on the analysis of its structure, it demonstrates the feasibility of the detection to mineral oil in water by fluorescence method, and by considering the impact of various environmental factors in fluorescence intensity, it lays a theoretical foundation for fluorescence spectroscopy detection and recognition of different types of mineral oil in water.It studies a fiber optic detection system which contains a fiber optic sensing system that can be used to detect fluorescent, and a full electronics system that can complete for detecting and processing weak fluorescence signal, and displays the measurement results via computer, and then completes the accurate measurement of mineral oil in water.By measuring three-dimensional fluorescence map mineral oil samples, it studies and analyzes the characteristics of fluorescence spectral. By studying the principle of multivariate calibration and parallel factor, combined with spectral characteristics of mineral oil samples, it determines the number of parallel factors, and then it applies the parallel factor analysis to identify and quantitative measurement of spectral features of mineral oil in water. Through testing, it completes the composition prediction and content measurement of mineral oil in water.By wavelet threshold denoising analysis which is based on multi-resolution wavelettransform and maximum entropy principle of entropy, and by selecting the best wavelet function, it can undo noise and make a simulation comparison with wavelet entropy threshold denoising algorithms and traditional analog low-pass filter denoising algorithms.The results show that after fluorescence spectra is processed by wavelet entropy threshold denoising algorithms, the edge and details information of original signal it can be more reserved and achieves better noise cancellation effect which is beneficial to enhance the accuracy of the system.