| Tea is susceptible to dozens of diseases and insect pests such as tea aphid(Toxoptera aurantii Boyer de Fonscolombe)and tea lesser leafhopper(Empoasca pirisuga Matumura)during its growth,and therefore the use of pesticides plays an important role in the prevention and control of diseases and insect pests,stable yield and increase of income.However,improper use and abuse of pesticides in tea production not only endanger consumer health,induce environmental pollution and ecological safety risks,but also have a major impact on the tea consumer market and the import and export trade of tea.Therefore,it is of great significance to establish a rapid,simple and on-the-spot detection method of pesticide residue.Conventional rapid detection technology,such as the enzyme inhibition principle of detection technology,is suitable for on-site rapid detection and screening of pesticide,but only to achieve qualitative or semi-quantitative detection.Chromatography,mass spectrometry and capillary electrophoresis have the advantages of high detection precision,good sensitivity and low detection limit,but they also have the disadvantages of high cost,cumbersome operation and poor portability.Immunological assays are highly sensitive and the device is easy to miniaturize,but the antibody is high cost and the detection object is relatively single.In recent years,the sensor method has drawn more and more attention in the research of pesticide detection due to its advantages of easy to use,miniaturization and high degree of intelligence of equipment.Sensor method can realize material analysis based on various detection methods such as optics and electrochemistry.Among them,the sensor method based on photochemical colorimetric detection has the advantages of simple operation,rapid response,large amount of information,and low detection cost.In recent years,it has developed rapidly in the research of pesticide detection.The photochemical colorimetric sensor uses the color change of the reaction system as a measure of the reaction degree,and with the help of the powerful computer color management system and digital image processing technology,image colorimetric analysis has shown unique advantages in the qualitative and quantitative detection of the substance.In this study,the prepared plant esterase and commercial acetylcholinesterase were used as specific sensitive materials,two kinds of esterase-colorimetric arrays were designed and constructed with image colorimetric analysis as the detection means.The detection of pesticide residues in tea had been realized based on computer digital image processing and pattern recognition algorithm,and it was expected to provide ideas and references for the rapid detection of pesticide.The main work of this thesis is as follows:1)Preparation of plant esterase and its susceptibility to pesticidesFirstly,the plant esterase of Chongqing Erba Cowpea had higher activity and specific activity by screening test,and maybe,it can be used as an alternate enzyme for pesticide detection,and then the optimization of extraction process of the enzyme and the optimization of enzyme-catalyzed reaction condition were studied.Purification of plant esterase was performed by salting out,ion exchange chromatography and molecular sieve chromatography.The activity of the plant esterase was largely unaffected by the common metal ions and the composition of tea,and further studies had found that the stability of the plant esterase was better than that of the commercialized acetylcholinesterase.The plant esterase was more sensitive to ten kinds of organophosphorus and carbamate pesticides,and it indicated that the plant esterase could be used for the detection of two kinds of pesticides and be used as the specific sensitive material for the subsequent construction of the colorimetric sensor array.The possible mechanism of the sensitivity difference of plant esterase to different pesticides was preliminarily studied by molecular docking.2)The design and construction of esterase colorimetric sensor array systemThe idea and method of detecting specific components in a mixed system based on image colorimetric analysis were studied,and based on this,a single-esterase colorimetric array system by choosing acetylcholine esterase as a specific sensitive material and a double-esterase colorimetric array system by choosing plant esterase and acetylcholinesterase as specific sensitive materials were designed and constructed.To achieve image colorimetric analysis,the former was based on the change of color intensity of single color reaction.The latter was a colorimetric sensor array composed of six specific color reactions with cross-response mechanism,which could realize image colorimetric analysis through the different maps of array.The workflows of two kinds of esterase colorimetric arrays and the methods of image acquisition,image preprocessing,the extraction of characteristic value of image were established.3)Quantitative detection of pesticide in tea based on single esterase colorimetric sensor array combined with pattern recognition algorithmWith different concentrations of pesticides,the color difference of the chromogenic system of the array was visible.Based on this,the corresponding relationship between the concentration of pesticides and the color intensity of the CMYK color space was established through image processing techniques.The quantitative analysis model was established on the base of Levenberg-Marquardt back propagation neural network(LMBP).Based on the array and constructed LMBP model,the detection of pesticide samples was realized.The decolorized extracts of tea containing different pesticides were respectively analyzed based on monoesterase colorimetric array and image colorimetric analysis,and there was no significant difference between the results obtained by LMBP model and gas chromatography(GC)(p>0.05).The experimental results of accuracy,precision and reproducibility showed that the established method had good stability,and it could provide a reference for the rapid detection of pesticides in real samples.4)Qualitative and semi-quantitative detection of pesticide in tea based on two-esterase colorimetric sensor array combined with pattern recognition algorithmThere were visible differences in the differential maps of the array of different kinds of pesticides and different concentrations of the same pesticides.Based on the characteristic difference map,the qualitative analysis of the samples of different concentration of same pesticide and the different types of pesticides were achieved by image colorimetric analysis combined with pattern recognition algorithms.There was no significant difference in the color difference between the standard sample of pesticide and three kinds of tea decolorization extracts containing pesticides(P>0.05),and the above two kinds of samples could be clustered into a class by Euclidean distance.Based on pesticide concentration and its correspondingly different map of array,the automatic classifier of the combination of genetic algorithm with neural network(GA-BP)and the combination of genetic algorithm with support vector machine(GA-SVM)had been designed and built through parameter optimization for pesticide detection,and the results showed that the two classifiers had good robustness and generalization ability,which can achieve qualitative and semi-quantitative detection of pesticide in tea.5)Quantitative detection of pesticide in tea based on double-esterase colorimetric sensor array combined with pattern recognition algorithmBased on double esterase colorimetric sensor array and image processing technology,the relationship between the concentration of pesticide standard and the Euclidean distance of the difference map of array was explored,and based on the above relationship,three kinds of quantitative analysis model were established,and the comparison results showed that the predictive performance of quantitative model of generalized regression neural network(GRNN)was obviously better than that of LMBP model and linear regression model by parameter optimization,and the Pearson's correlation coefficient between the predicted value and the actual value of the pesticide based on the quantitative model of GRNN reached 0.9954.The decolored extracts of tea containing organic phosphorus and carbamate pesticide were detected by the colorimetric array and the GRNN model,and there was no significant difference in comparison with the detection results of GC(p>0.05).It showed high accuracy,precision and good reproducibility for detection of pesticide by image colorimetric analysis based on dual esterase colorimetric array and pattern recognition algorithm.It could provide a reference for the rapid detection of pesticides. |
PDF Full Text Request