Font Size: a A A

Research Of Polycyclic Aromatic Hydrocarbons Detection Based On Fluorescence And Signal Processing

Posted on:2017-01-15Degree:MasterType:Thesis
Country:ChinaCandidate:X L WangFull Text:PDF
GTID:2271330503982468Subject:Instrumentation engineering
Abstract/Summary:
Environmental pollution has always been a global problem. China bears the burnt of the pollution since it is under rapid development of economy. Water pollution, soil pollution and air pollution are primary problems in environmental pollution. It is hard to count the kinds of substance which caused the environmental pollution. Among them, PAHs is great harmful to the health of plants, animals and human. Being difficult to decompose and easy to damage the structure of the animal cell even to lead to the occurrence of cancer results from its physical and chemical properties as well as its structure. PAHs is widespread due to industrial products,by-products and the processing. A rapid and accurate detection of polycyclic aromatic hydrocarbons can not only protect human health, but also can promote environmental protection. Hence, to find out it becomes a top priority. PAHs contains many classes of substance, many of them are isomers which result in a number of them have similar physical and chemical properties and therefore it won’t be easily detected.Since the spectrum of the mixture will overlap, this issue will do a quantitative research on the mixture by the methods combined with fluorescence and chemometrics. And the final data can prove the accuracy of the method.First, we do a deep study on the latest research results of PAHs by consulting lots of information. There are severe spectral overlapping phenomenons in spectrum of the mixture, not simply an addition of each substance’s spectral. It is targeted to predict the concentration of the mixture by using the methods that combines the ensemble empirical mode decomposition and the partial least squares regression.Useful information will be preserved after de-noise processing which can not only eliminate noise interference but also reduce data dimensions to avoid data redundancy. Import the data before and after denoising into the algorithm and compare the results, from which it can be learnt that the data prediction result after denoising is more precise and rapid. The final result also proves the correctness and feasiblity of this method.
Keywords/Search Tags:PAHs, Fluorescence, Ensemble empirical mode decomposition, Partial least squares regression
Related items