Research Of Oil Identificaton Method Based On Three-dimensional Fluorescence Spectroscopy

Oil pollution has become a main pollutant source of water safety, which has many devastating impacts on the whole environment, it can not only cause aquatic organisms death, fishery damage, rare species extinction, but also influent the atmospheric circulation and the ecology blance. All of them can make the world governments and people suffer from huge economic losses. Therefore, it is important for the water quality management and pollution prevention to identify the type, source and character of pollution oil with high accuracy in a short time. After analying the development status of conventional fluorescence technology, we propose an oil identification method based on multi-feature recognition theory of three- dimentional(3D) fluorescence spectroscopy, and also a series of experiments and algorithm analysis were carried out.First, on the basis of literature reading, some exploratory experiments were taken out to understand the fluorescent principle of mineral oil. Base on the workings, we confirmed the feasibilty to identify oil type with fluorescent technology and designed the specific project schemes.And then, a lot of 3D fluorescence data of oil samples were gotten by using cary eclipse spectrometer according to the test schemes. Based on image statictical theory, we abtained some characteristic information of the 3D data, and sugest a new method to identify oil with contour feature spectra.Finally, in order to has a full understanding of oil compisition,we established the parallel factor analysis(PARAFAC) model of different oils. With the loading and score vectors, we got the estimate spectra of pure components and a new oil characterize method. During the reseach, we analysed many oil samples form different gas stations and oil refinery, and can distinguish oil successfully with contour feature spectra and PARAFAC model. Estimate results proved that the new method has high validity for oil identification. This new method in this paper is not only simple, low cost, analyis fast and accurate, but also has a good ability to distinguish oil with simlar component. The PARAFAC model provide a effective basis for both water quality monitoring and micro-fine chemicals.