本文選題：農(nóng)藥殘留 + 恒光程��； 參考：《江蘇大學(xué)》2017年碩士論文

【摘要】：目前主流的農(nóng)藥殘留檢測(cè)手段所涉及的檢測(cè)設(shè)備普遍存在著體積龐大、檢測(cè)精度低、穩(wěn)定性差的缺陷。為此,課題基于微流控芯片搭建了便攜式農(nóng)藥殘留檢測(cè)裝置。然而所搭建的便攜化農(nóng)藥殘留檢測(cè)裝置存在下述兩個(gè)問(wèn)題:(1)存在著恒程誤差的干擾;(2)缺少相應(yīng)的硬件或軟件誤差補(bǔ)償模型。針對(duì)上述問(wèn)題,課題從以下幾個(gè)方面展開(kāi)工作:(1)基于朗伯比爾定律建立了恒程誤差σ與光程長(zhǎng)l之間的關(guān)系模型,并從影響其關(guān)系模型的噪聲角度對(duì)所建立的關(guān)系模型進(jìn)行了修正;(2)利用CorelDraw設(shè)計(jì)了旋轉(zhuǎn)蝶式微流控芯片模型并借助COMSOL對(duì)其進(jìn)行了仿真,確定了芯片的材料和制作方法,搭建了農(nóng)藥殘留光度檢測(cè)裝置并優(yōu)化了實(shí)驗(yàn)溫度、光源光強(qiáng)、光源波長(zhǎng)三個(gè)參數(shù),利用農(nóng)藥殘留光度檢測(cè)裝置驗(yàn)證了均勻指數(shù)仿真模型并建立恒程誤差硬件補(bǔ)償模型;(3)利用最小二乘法、BP神經(jīng)網(wǎng)絡(luò)以及支持向量機(jī)三種算法分別建立了相應(yīng)的恒程誤差軟件補(bǔ)償模型,選取最優(yōu)模型作為最終恒程誤差軟件補(bǔ)償模型;(4)搭建了基于微流控芯片便攜式農(nóng)藥殘留檢測(cè)裝置,介紹了便攜式農(nóng)藥殘留檢測(cè)裝置硬件電路設(shè)計(jì),優(yōu)化了底物的濃度值、酶抑制時(shí)間以及溶液的pH值三個(gè)參數(shù)。借助便攜式農(nóng)藥殘留檢測(cè)裝置并應(yīng)用所建立的硬件和軟件補(bǔ)償模型對(duì)草莓和西紅柿中敵百蟲(chóng)、呋喃丹、久效磷3種常見(jiàn)的有機(jī)磷或氨基甲酸酯類農(nóng)藥殘留進(jìn)行檢測(cè)以驗(yàn)證所建立補(bǔ)償模型的正確性。實(shí)驗(yàn)結(jié)果表明:經(jīng)硬件和軟件補(bǔ)償模型補(bǔ)償后的吸光度均方根誤差RMSE分別為0.3901、0.0883,基本實(shí)現(xiàn)了農(nóng)藥殘留檢測(cè)恒程誤差的補(bǔ)償;與傳統(tǒng)微流控農(nóng)殘光度檢測(cè)方法相比,課題所述方法使草莓和西紅柿中的敵百蟲(chóng)等3種常見(jiàn)有機(jī)磷或氨基甲酸酯類農(nóng)藥農(nóng)殘檢測(cè)誤差降低24.3%以上,基本實(shí)現(xiàn)較為精確的農(nóng)藥殘留檢測(cè);在應(yīng)用所建立的補(bǔ)償模型后,課題所述的農(nóng)藥殘留方法對(duì)草莓和西紅柿中農(nóng)藥殘留檢測(cè)值與農(nóng)藥加標(biāo)濃度值的相對(duì)誤差分別為3.99%、4.58%、4.03%,基本驗(yàn)證了所建立補(bǔ)償模型的正確性。
[Abstract]:At present, the detection equipment involved in the mainstream pesticide residue detection methods generally has the defects of large volume, low detection precision and poor stability. Therefore, a portable pesticide residue detection device based on microfluidic chip is built. However, the portable pesticide residue detection device has the following two problems: 1) there is constant range error interference / 2) there is a lack of hardware or software error compensation model. In order to solve the above problems, the following work is done: 1) based on Lambert's law, a model of the relationship between constant path error 蟽 and optical path length l is established. At the same time, the relationship model is modified from the angle of noise which affects its relation model. The model of rotating butterfly microfluidic chip is designed by CorelDraw and simulated by COMSOL. The material and fabrication method of the chip are determined. The experimental temperature, light intensity and wavelength of the light source were optimized. The uniform exponent simulation model and the hardware compensation model of constant range error are established by using the pesticide residue photometric detection device.) the corresponding constant range is established by using the least square BP neural network and the support vector machine respectively. Error software compensation model, Selecting the optimal model as the final constant range error software compensation model, the portable pesticide residue detection device based on microfluidic chip is built. The hardware circuit design of the portable pesticide residue detection device is introduced, and the concentration value of the substrate is optimized. The enzyme inhibition time and pH value of the solution were three parameters. With the help of portable pesticide residue detection devices and using the established hardware and software compensation models, furan, trichlorfon and furan in strawberries and tomatoes, Three common organophosphorus or carbamate pesticide residues were detected to verify the correctness of the compensation model. The experimental results show that the root mean square error (RMSE) of absorbance after compensating by hardware and software compensation model is 0.3901 / 0.0883respectively, which basically realizes the compensation of the constant path error of pesticide residue detection, and compared with the traditional microfluidic method, the RMSE of RMSE is 0.3901 / 0.0883.Compared with the traditional microfluidic method, The method can reduce the detection error of pesticide residues of three common organophosphorus or carbamate pesticides, such as trichlorfon in strawberry and tomato, by more than 24.3%. The relative error between the detection value of pesticide residue in strawberry and tomato and the concentration of pesticide in strawberry and tomato is 3.99 and 4.58 and 4.03, respectively, which basically verifies the correctness of the compensation model.
【學(xué)位授予單位】：江蘇大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：S481.8;TP18

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 楊寧;張榮標(biāo);徐佩鋒;郭建江;成立;邵世和;;基于微流控芯片的乳膠免疫凝集光電檢測(cè)方法[J];儀器儀表學(xué)報(bào);2013年06期

2 陳思;柯福順;樂(lè)永康;;光柵光譜儀的標(biāo)定[J];物理實(shí)驗(yàn);2012年03期

3 冷翔鵬;孫欣;房經(jīng)貴;宋長(zhǎng)年;;波爾多液作用機(jī)理及其在果樹(shù)生產(chǎn)上的應(yīng)用與相應(yīng)藥害研究進(jìn)展[J];江蘇農(nóng)業(yè)科學(xué);2012年02期

4 陳清海;黎海文;吳一輝;郝鵬;;生化分析儀的誤差分析與建模[J];西安理工大學(xué)學(xué)報(bào);2011年01期

5 程定璽;趙亞鵬;高曼;;熒光法快速檢測(cè)有機(jī)磷農(nóng)藥殘留總量[J];分析試驗(yàn)室;2010年02期

6 干寧;王峰;楊欣;李天華;王軍;;采用納米修飾雙酶電極生物傳感器檢測(cè)有機(jī)膦與氨基甲酸酯類農(nóng)藥[J];農(nóng)藥學(xué)學(xué)報(bào);2008年03期

7 王玉田;崔立超;王冬生;李艷春;;基于同步-導(dǎo)數(shù)熒光光譜法的多組分農(nóng)藥殘留測(cè)定的研究[J];光譜學(xué)與光譜分析;2006年11期

8 何穎,張濤,康天放;蔬菜中有機(jī)磷農(nóng)藥殘留的分光光度法快速檢測(cè)[J];環(huán)境化學(xué);2005年06期

9 陳光,任志良,孫海柱;最小二乘曲線擬合及Matlab實(shí)現(xiàn)[J];兵工自動(dòng)化;2005年03期

10 汪V，

本文編號(hào)：2047975