Studies On Detection Method Of Carbamate Pesticides Mixture Based On Fluorescence Spectroscopy

Carbamate pesticides,as one of the most common pesticides in agricultural production in China,are widely used in farmland.However,pesticide residues in water will cause great harm to the environment and human body.Therefore,traditional detection methods are unable to accurately and rapidly determine pesticides.Fluorescence spectrometry has the advantages of high sensitivity,multiple characteristic parameters and simple operation,which can be used for the detection of carbamate pesticides and is of great significance for the monitoring and treatment of environmental problems.The purpose of this paper is to solve the traditional testing instrument category and content of the pesticide pollution in the aquatic environment for precise measurement and identify problems,based on the mechanism of fluorescence spectroscopy combined with chemometrics in three kinds of second order correction methods in laboratory configured carbamate pesticide mixture solution qualitative quantitative detection,then,the two kinds of neural network algorithm was used to make quantitative prediction on the analysis,and the experimental results for further discussion.The main work of this paper is as follows:(1)The theory of fluorescence spectrometry,especially the principles of photoluminescence and three-dimensional fluorescence spectrometry,the hardware structure of the fluorescence detection system,and the feasibility of detecting carbamate pesticides by fluorescence spectrometry were described.(2)Selecting three pesticides,carbamate pesticide carbaryl,tsumacide and organophosphate insecticide triazopho as samples,in a laboratory simulation nature present situation of the existence of various pesticide residues in the aquatic environment,using fluorescence spectrometer to fluorescence detection of configured pesticide mixture solution,and scattering of three-dimensional fluorescence spectrum data to eliminate and spectrum correction of pretreatment operation,analysis the three dimensional fluorescence spectrum features.(3)Mixed pesticides for the excitation and emission spectra overlapping degree high and hard to distinguish between problems,combined with chemometrics second-order correction method theory,the application of parallel factor analysis method,alternating trilinear decomposition algorithm and the weighted alternating trilinear decomposition algorithm respectively and quantitative analysis on three kinds of pesticide mixture solution,by comparing the experimental results,select the prediction effect,the optimal algorithm.(4)By analyzing the BP neural network algorithm and combining it with the genetic algorithm,the GA-BP neural network algorithm model was obtained and optimized,and the mixture solution of two kinds of carbamate pesticides was detected and quantitatively analyzed,verifying the feasibility of using the neural network algorithm to predict pesticide concentration.