A multi-sample labeled agricultural product pesticide residue detection method and system

Abstract

The application discloses a kind of multi-sample marking agricultural product pesticide residue detection method and system, it is related to pesticide residue detection technical field, the application includes extraction state setting, mixed liquid state setting and pesticide residual analysis, the method first takes the parallel test sample of target batch agricultural product, obtains the pressurized solvent extraction process of adaptation this batch by temperature and pressure gradient test optimization;Again, through the state data analysis of multiple groups of mixed formula, the target formula is screened and prepared to be detected mixed liquid;Finally, through the potential-time-current three-dimensional dynamic sequence analysis calculation pesticide residue concentration, complete multi-sample marking and grading early warning of overproof sample.The method can adapt to the matrix characteristics of different batches of agricultural products, which helps to improve the consistency of batch sample detection results and low concentration residue detection capability, suitable for batch rapid screening of agricultural product pesticide residue.

Description

Technical Field

[0001] This invention relates to the field of pesticide residue detection technology, specifically to a method and system for detecting pesticide residues in agricultural products with multi-sample labeling. Background Technology

[0002] Pesticide residue detection is a crucial step in the quality control of agricultural products. The accuracy of the detection results, the efficiency and stability of batch testing directly affect the effectiveness of agricultural product quality screening. Therefore, a multi-sample labeled pesticide residue detection method and system for agricultural products is needed.

[0003] Currently, pesticide residue detection technologies for agricultural products are mainly divided into two categories: laboratory precision instrument detection and on-site rapid detection. Among them, laboratory precision instrument detection methods, such as gas chromatography and liquid chromatography-mass spectrometry, have high detection accuracy, but are limited by complex pretreatment processes, long detection cycles, high equipment costs, and inability to adapt to rapid batch screening on-site. On the other hand, on-site rapid detection methods, such as spectroscopic detection and electrochemical detection, have become the mainstream technology for on-site screening of agricultural products due to their advantages of fast detection speed, portable equipment, and batch processing capability.

[0004] In existing rapid detection technologies, pressurized solvent extraction is a commonly used sample pretreatment method that can effectively improve the extraction efficiency of pesticide components. However, existing technologies mostly use fixed extraction temperature and pressure parameters, without personalized optimization for the matrix characteristics of different batches of agricultural products. This can easily lead to differences in extraction efficiency among different samples within the same batch, resulting in deviations in batch detection results and affecting the stability of the detection results. Furthermore, the pretreatment reagent mixture formulations of existing detection methods are mostly fixed ratios, without adaptation and optimization for matrix interference in the extracts from different batches of agricultural products. This can easily lead to a low signal-to-noise ratio in the detection system, insufficient detection capability for low concentrations of pesticide residues, and a certain risk of false negatives.

[0005] Existing rapid electrochemical detection methods mostly use a single voltammetric scanning mode to acquire detection data, lacking multi-dimensional dynamic sequence analysis. The accuracy of identifying pesticide characteristic peaks is easily affected by matrix interference, resulting in a high false positive rate. At the same time, in batch multi-sample detection scenarios, existing technologies lack a sound labeling mechanism for samples exceeding standards and a graded early warning system, which cannot provide sufficient data support for the quality graded control of batch samples and is difficult to adapt to the application needs of large-scale batch screening of agricultural products. Summary of the Invention

[0006] To address the aforementioned technical shortcomings, the present invention aims to provide a method and system for detecting pesticide residues in agricultural products using multi-sample labeling.

[0007] To solve the above technical problems, the present invention adopts the following technical solution: The present invention provides a method for detecting pesticide residues in agricultural products with multiple sample labeling, including the following steps: Step 1, extraction state setting: Take parallel test samples of each agricultural product sample of the target batch, first test the temperature of the sample to obtain the target temperature, and test the pressure of the parallel test samples of the same batch to obtain the target pressure. With the target temperature and target pressure as the core parameters, set the pressurized solvent extraction process of the target batch.

[0008] Step 2, Mixture State Setting: According to the pressurized solvent extraction process of the target batch, extract each agricultural product sample of the target batch to obtain the corresponding sample extract. According to the preset multiple mixing formulas, mix the extract of each agricultural product sample with the matching reagents to obtain the sample mixture corresponding to each mixing formula. Collect and obtain the state data of the sample mixture corresponding to each mixing formula; analyze the state data of each mixing formula to obtain the target formula.

[0009] Step 3: Pesticide Residue Analysis: The extracts of each agricultural product sample in the target batch are mixed with the target formula to obtain the test mixture corresponding to each sample. Three-dimensional dynamic sequence analysis is performed on each test mixture. The pesticide residue concentration is calculated based on the analysis results. Multiple samples exceeding the standard are labeled, and graded early warning is implemented based on the labeling results.

[0010] Preferably, the specific process for obtaining the target temperature is as follows: From the spectral reflectance images corresponding to each temperature change segment, extract the reflectance value of each pixel at a wavelength of 550nm, and construct a T×M... 2 A 3D temperature reflectance matrix, where T is the number of temperature variation segments and M is the pixel side length of the spectral reflectance image.

[0011] The standard deviation of the temperature reflectivity matrix along the T dimension is calculated to obtain M. 2 A spatial standard deviation vector; the positions of elements in the spatial standard deviation vector whose absolute value is greater than a preset threshold are marked as high-response regions.

[0012] The number of pixels in the high-response region corresponding to each temperature change segment is counted to form a T-dimensional response intensity vector. The response intensity vector is normalized so that the sum of its elements is 1. The temperature change segment number corresponding to the maximum value in the normalized response intensity vector is used as the target temperature identifier, and the end temperature of the temperature change segment corresponding to the target temperature identifier is the target temperature.

[0013] Preferably, the pressure test of the parallel test samples in the same batch is carried out as follows: a fluorescent tracer with the same extraction kinetics as the target pesticide is mixed into the parallel test samples in the same batch to obtain each pressure analysis sample.

[0014] Each pressure analysis sample is placed on a microporous polytetrafluoroethylene membrane inside a sealed extraction chamber. The vacuum-positive pressure dual-mode pressure generator is activated to set a gradient extraction pressure for each pressure analysis sample. At the target temperature, each pressure analysis sample is subjected to pressure extraction testing.

[0015] Starting from the moment of pressure extraction, the fluorescence intensity values ​​of the characteristic peaks of the fluorescent tracer in the extract at the corresponding pressure are collected at preset sampling intervals to obtain the time-series change sequence of fluorescence intensity for each pressure analysis sample.

[0016] Preferably, the acquisition of the target pressure is specifically carried out as follows: a standard change curve of fluorescence intensity under the condition of complete extraction of the target pesticide is preset, the sampling time points of the standard change curve correspond one-to-one with the sampling time in the pressure test, and the fluorescence intensity values ​​of each sampling point in the standard change curve are arranged in time sequence to obtain the basic fluorescence intensity sequence vector.

[0017] From the time-series changes in fluorescence intensity of each pressure analysis sample, the fluorescence intensity values ​​of each sampling point that are strictly aligned with the sampling time points of the standard change curve are extracted, and the measured fluorescence intensity sequence vector of each pressure analysis sample is constructed.

[0018] The cosine similarity between the basic fluorescence intensity sequence vector and the measured fluorescence intensity sequence vector of each pressure analysis sample is calculated to obtain the similarity between each pressure analysis sample and the standard variation curve; the extraction pressure of the pressure analysis sample corresponding to the maximum similarity is denoted as the target pressure.

[0019] Preferably, the analysis of the state data of each mixed formulation is carried out as follows: baseline correction is performed on each absorption spectrum curve using Savitzky-Golay smoothing filtering; the absorbance value at a wavelength of 280 nm is extracted from the corrected spectrum as the characteristic response value of the corresponding mixed formulation; at the same time, the standard deviation of absorbance in the 200 nm-230 nm non-characteristic absorption band is extracted as the baseline noise value.

[0020] The ratio of the characteristic response value to the baseline noise value is recorded as the signal-to-noise ratio of each mixture formulation. The signal-to-noise ratio is corrected by combining the slope of the UV absorption standard curve of the preset target pesticide: the slope of the UV absorption standard curve of the target pesticide is multiplied by the signal-to-noise ratio of each mixture formulation to obtain the corrected signal-to-noise ratio estimate of each mixture formulation. The mixture formulation corresponding to the maximum corrected signal-to-noise ratio estimate is selected as the target formulation.

[0021] On the other hand, the present invention provides a multi-sample labeled pesticide residue detection system for agricultural products, including the following modules: an extraction state setting module, used to take parallel test samples of each agricultural product sample of the target batch, first test the temperature of the sample to obtain the target temperature, then test the pressure of the parallel test samples of the same batch to obtain the target pressure, and set the pressurized solvent extraction process of the target batch with the target temperature and target pressure as the core parameters.

[0022] The mixed liquid state setting module is used to extract each agricultural product sample of the target batch according to the pressurized solvent extraction process of the target batch, to obtain the extract of the corresponding sample, and to mix the extract of each agricultural product sample with the matching reagent according to multiple preset mixing formulas to obtain the sample mixture corresponding to each mixing formula. The module collects and obtains the state data of the sample mixture corresponding to each mixing formula; and analyzes the state data of each mixing formula to obtain the target formula.

[0023] The pesticide residue analysis module is used to mix the extracts of each agricultural product sample in the target batch with the target formula to obtain the test mixture corresponding to each sample, perform three-dimensional dynamic sequence analysis on each test mixture, calculate the pesticide residue concentration based on the analysis results, label the samples that exceed the standard, and execute graded early warning based on the labeling results.

[0024] The beneficial effects of the present invention are as follows: 1. The present invention optimizes the target temperature and target pressure that are suitable for the matrix characteristics of the target batch of agricultural products by gradient testing of parallel test samples of the target batch. Based on this, the pressurized solvent extraction process is set. Compared with the extraction method with fixed parameters, it helps to improve the consistency of pesticide extraction efficiency of different agricultural product samples in the same batch. It can reduce the result deviation of batch multi-sample detection to a certain extent and improve the stability of the detection results.

[0025] 2. This invention collects and analyzes state data of multiple mixed formulations, and combines signal-to-noise ratio and standard curve slope correction to screen out the target formulation suitable for the batch of extract. This can reduce the interference of agricultural product matrix on the detection results to a certain extent, improve the signal-to-noise ratio of the detection system, help improve the detection capability of low-concentration pesticide residues, and reduce the risk of false negatives.

[0026] 3. This invention constructs a three-dimensional dynamic sequence matrix with potential, time, and current response dimensions through a three-segment voltage program. Combined with multi-dimensional detection data from linear scanning voltammetry and differential pulse voltammetry, compared with detection methods using a single scanning mode, this helps improve the accuracy of pesticide characteristic peak identification, reduce the probability of false positives caused by matrix interference, and improve the reliability of detection results.

[0027] 4. This invention can complete the labeling of multiple samples that exceed the standard during the batch detection process, and construct a three-level graded early warning mechanism by combining the proportion of samples exceeding the standard and the multiple of the exceeding concentration. It can adapt to the application needs of large-scale batch rapid screening of agricultural products, and can provide multi-dimensional detection data support for the graded control of agricultural product quality. It has good practicality and promotion potential. Attached Figure Description

[0028] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0029] Figure 1 This is a schematic diagram of the implementation steps of the method of the present invention.

[0030] Figure 2 This is a schematic diagram of the system structure connection of the present invention. Detailed Implementation

[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] according to Figure 1 As shown, the present invention provides a method for detecting pesticide residues in agricultural products with multiple sample labeling, including the following steps: Step 1, extraction state setting: Take parallel test samples of each agricultural product sample of the target batch, first test the temperature of the sample to obtain the target temperature, and then test the pressure of the parallel test samples of the same batch to obtain the target pressure. Using the target temperature and target pressure as core parameters, set the pressurized solvent extraction process for the target batch.

[0033] In one specific embodiment, the temperature test of the sample is carried out as follows: the parallel test sample is placed on the Peltier temperature control stage, the Peltier temperature control stage is started, and gradient heating is performed according to the preset heating program: starting from the preset reference temperature, the temperature is increased uniformly at the preset heating rate. Each increase of the preset temperature gradient value is recorded as a temperature change node and the corresponding timestamp is recorded. The heating process is divided according to each timestamp to obtain multiple continuous temperature change segments.

[0034] During the isothermal holding phase corresponding to each temperature change node, spectral reflectance images in the 400nm–700nm band are acquired using a hyperspectral imager to obtain spectral reflectance images corresponding to each temperature change segment.

[0035] It should be noted that the 400nm–700nm visible light band was chosen for hyperspectral imaging because this band covers the characteristic reflectance range of chlorophyll and cell structure in agricultural product matrices. Furthermore, the target pesticide components exhibit captureable spectral response differences with temperature changes within this band. The Peltier temperature-controlled stage enables high-precision linear heating and isothermal control, ensuring that the temperature control error for each temperature change segment does not exceed ±0.5℃, thus avoiding interference from temperature fluctuations on the spectral acquisition results. The gradient heating design aims to simulate the state changes of the sample matrix at different extraction temperatures. Through non-contact hyperspectral imaging, the precipitation and extraction response characteristics of pesticide components at different temperatures can be obtained non-destructively without the need for additional detection reagents, avoiding interference with subsequent pressure testing.

[0036] In one specific embodiment, the acquisition of the target temperature is specifically performed as follows: From the spectral reflectance images corresponding to each temperature change segment, the reflectance value of each pixel at a wavelength of 550nm is extracted, and a T×M array is constructed. 2 A 3D temperature reflectance matrix, where T is the number of temperature variation segments and M is the pixel side length of the spectral reflectance image.

[0037] The standard deviation of the temperature reflectivity matrix along the T dimension is calculated to obtain M. 2 A spatial standard deviation vector; the positions of elements in the spatial standard deviation vector whose absolute value is greater than a preset threshold are marked as high-response regions.

[0038] The number of pixels in the high-response region corresponding to each temperature change segment is counted to form a T-dimensional response intensity vector. The response intensity vector is normalized so that the sum of its elements is 1. The temperature change segment number corresponding to the maximum value in the normalized response intensity vector is used as the target temperature identifier, and the end temperature of the temperature change segment corresponding to the target temperature identifier is the target temperature.

[0039] It should be noted that the reflectance value at a wavelength of 550nm was chosen as the basis for calculation because this wavelength is the characteristic reflectance peak position of green agricultural products. Temperature changes causing damage to the sample cell structure and the release of pesticide components will directly lead to a significant change in the reflectance at this wavelength, which can accurately reflect the change law of extraction efficiency with temperature. The standard deviation is calculated along the T dimension to quantify the degree of fluctuation of the reflectance of each pixel with temperature changes. The larger the standard deviation, the more sensitive the position is to temperature changes, corresponding to a high-response region enriched with pesticide components. The normalization process eliminates the order of magnitude difference in the number of pixels in different temperature ranges, which allows for an objective comparison of the extraction response intensity of each temperature range. The final target temperature is the optimal temperature with the highest pesticide extraction efficiency under the matrix of this batch of agricultural products, which can ensure the consistency of extraction efficiency of all samples in the same batch.

[0040] In one specific embodiment, the pressure test of the parallel test samples in the same batch is carried out as follows: a fluorescent tracer with the same extraction kinetics as the target pesticide is mixed into the parallel test samples in the same batch to obtain each pressure analysis sample.

[0041] Each pressure analysis sample is placed on a microporous polytetrafluoroethylene membrane inside a sealed extraction chamber. The vacuum-positive pressure dual-mode pressure generator is activated to set a gradient extraction pressure for each pressure analysis sample. At the target temperature, each pressure analysis sample is subjected to pressure extraction testing.

[0042] Starting from the moment of pressure extraction, the fluorescence intensity values ​​of the characteristic peaks of the fluorescent tracer in the extract at the corresponding pressure are collected at preset sampling intervals to obtain the time-series change sequence of fluorescence intensity for each pressure analysis sample.

[0043] It should be noted that by selecting a fluorescent tracer with the same extraction kinetics as the target pesticide, the extraction kinetics process under different pressures can be monitored in real time and non-destructively through fluorescence signals without introducing interference from the target pesticide or damaging the sample matrix, eliminating the need for destructive testing of the extract. The microporous polytetrafluoroethylene membrane can filter solid matrix residues in the sample, allowing only the extract containing the tracer to pass through, thus avoiding interference from matrix impurities on the fluorescence detection results. Pressure testing at the determined target temperature can eliminate the influence of temperature variables on the extraction process, ensuring the accuracy of the pressure optimization results. The gradient extraction pressure setting range covers the working range of 0.5MPa to 20MPa for conventional pressurized solvent extraction, comprehensively covering the suitable pressure range for different agricultural product matrices.

[0044] For example, fluorescein diacetate is suitable for moderately polar organophosphorus pesticides, coumarin-3-carboxylic acid ethyl ester is suitable for strongly polar organophosphorus pesticides, and 9,10-diphenylanthracene is suitable for highly lipid-soluble pyrethroid pesticides.

[0045] In one specific embodiment, the acquisition of the target pressure is specifically carried out as follows: a standard change curve of fluorescence intensity under the condition of complete extraction of the target pesticide is preset, the sampling time points of the standard change curve correspond one-to-one with the sampling time in the pressure test, and the fluorescence intensity values ​​of each sampling point in the standard change curve are arranged in time sequence to obtain the basic fluorescence intensity sequence vector.

[0046] From the time-series changes in fluorescence intensity of each pressure analysis sample, the fluorescence intensity values ​​of each sampling point that are strictly aligned with the sampling time points of the standard change curve are extracted, and the measured fluorescence intensity sequence vector of each pressure analysis sample is constructed.

[0047] The cosine similarity between the basic fluorescence intensity sequence vector and the measured fluorescence intensity sequence vector of each pressure analysis sample is calculated to obtain the similarity between each pressure analysis sample and the standard variation curve; the extraction pressure of the pressure analysis sample corresponding to the maximum similarity is denoted as the target pressure.

[0048] It should be noted that the standard variation curve of fluorescence intensity under the complete extraction state of the target pesticide was obtained through a prior complete extraction verification experiment, which can reflect the entire extraction kinetics of the target pesticide under optimal conditions. The cosine similarity calculation can quantify the degree of fit between the measured extraction process and the complete extraction standard process. The closer the similarity is to 1, the more consistent the extraction rate and extraction completion degree are with the complete extraction state under that pressure. The target pressure determined in this way can ensure that the target pesticide in the batch of agricultural product samples is fully extracted, avoiding the problem of low detection results and false negatives due to insufficient extraction.

[0049] In one specific embodiment, the pressurized solvent extraction process for the target batch is set with target temperature and target pressure as core parameters. The specific process is as follows: Take samples of agricultural products to be tested from the target batch, crush and homogenize them with a high-speed pulverizer, accurately weigh a preset mass of homogenized sample, and put it into a pressurized extraction tank made of stainless steel. Both ends of the sample are filled with quartz wool to avoid sample powder clogging the flow path.

[0050] Chromatographically pure acetonitrile was used as the extraction solvent. The isothermal parameters of the extraction system were set to the target temperature and the system pressure parameters were set to the target pressure. The static extraction cycle was set to 1-3 times, the holding time for a single static extraction was 5-15 min, the rinsing volume was set to 1.5 times the volume of the extraction cell, and the nitrogen purging time was set to 60 s.

[0051] Start the pressurized solvent extraction system, preheat the extraction tank, and after the tank temperature stabilizes at the target temperature and is maintained at that temperature for 30 seconds, inject the extraction solvent into the extraction tank and pressurize it to the target pressure. Complete the preset static extraction cycle under the target temperature and target pressure conditions.

[0052] After extraction, the system automatically depressurizes and cools to room temperature, collects all the extract eluent, filters it under vacuum through a 0.22μm organic phase microporous membrane, and transfers it to a volumetric flask. The volume is then adjusted to a preset volume using the extraction solvent to obtain the clarified extract corresponding to each agricultural product sample in this batch, which is used for subsequent mixture preparation and detection.

[0053] Step 2, Mixture State Setting: According to the pressurized solvent extraction process of the target batch, extract each agricultural product sample of the target batch to obtain the corresponding sample extract. According to the preset multiple mixing formulas, mix the extract of each agricultural product sample with the matching reagents to obtain the sample mixture corresponding to each mixing formula. Collect and obtain the state data of the sample mixture corresponding to each mixing formula; analyze the state data of each mixing formula to obtain the target formula.

[0054] In one specific embodiment, the extraction liquid of each agricultural product sample is mixed with the matching reagents. The specific mixing process is as follows: the extraction liquid of the target batch of agricultural product samples is injected into the micro-mixing chamber with acetonitrile, phosphate buffer and sodium dodecyl sulfate solution according to the preset volume ratio of each mixing formula to obtain the sample mixture corresponding to each mixing formula.

[0055] For each mixed formulation, a micro stirrer was started under a preset pressure. After stirring, the mixture was allowed to stand and separate into layers. Then, a UV-Vis spectrometer was triggered to collect UV absorption spectra in the 200nm–400nm band. Each mixed formulation corresponds to an absorption spectrum curve. The absorbance values ​​of each absorption spectrum curve at each wavelength were recorded as the state data of the agricultural product sample mixture of the corresponding mixed formulation.

[0056] It should be noted that the acetonitrile used in this step is a commonly used extraction and dilution reagent in the field of pesticide residue detection, which can ensure the stable dissolution of the target pesticide in the mixed system; phosphate buffer is used to precisely control the pH value of the mixture, avoiding pH fluctuations from interfering with subsequent ultraviolet spectroscopy and electrochemical detection results; sodium dodecyl sulfate is an anionic surfactant that can disperse large matrix impurities in the mixed system, reduce impurity aggregation, reduce matrix interference, and improve the dispersibility of the target pesticide in the system; the micro-mixing chamber and micro-stirrer work together to achieve rapid and uniform mixing of the extract and reagents, ensuring that the mixing reaction conditions of different samples in the same batch are completely consistent, and improving the repeatability of batch detection results; 200nm–400nm is the ultraviolet characteristic absorption band of most commonly used organophosphorus, pyrethroid, and carbamate pesticides, which can cover the detection needs of most conventional pesticides.

[0057] In one specific embodiment, the analysis of the state data of each mixed formulation is carried out as follows: baseline correction is performed on each absorption spectrum curve using Savitzky-Golay smoothing filtering; the absorbance value at a wavelength of 280 nm is extracted from the corrected spectrum as the characteristic response value of the corresponding mixed formulation; at the same time, the standard deviation of absorbance in the 200 nm-230 nm non-characteristic absorption band is extracted as the baseline noise value.

[0058] The ratio of the characteristic response value to the baseline noise value is recorded as the signal-to-noise ratio of each mixture formulation. The signal-to-noise ratio is corrected by combining the slope of the UV absorption standard curve of the preset target pesticide: the slope of the UV absorption standard curve of the target pesticide is multiplied by the signal-to-noise ratio of each mixture formulation to obtain the corrected signal-to-noise ratio estimate of each mixture formulation. The mixture formulation corresponding to the maximum corrected signal-to-noise ratio estimate is selected as the target formulation.

[0059] It should be noted that Savitzky-Golay smoothing filtering can effectively remove random noise during the spectral acquisition process while preserving the characteristic peak signals of the spectrum, thus improving the accuracy of baseline correction. 280nm is the characteristic absorption wavelength of most commonly used pesticides, and using it as a characteristic response value can directly reflect the detection response intensity of the target pesticide in the mixture. The standard deviation of absorbance in the 200nm-230nm non-characteristic absorption band can objectively quantify the baseline noise level of the mixture, unaffected by the characteristic peaks of the target pesticide. By correcting the signal-to-noise ratio through the slope of the standard curve, both the response sensitivity and noise level of the detection system can be considered. The higher the corrected value, the stronger the detection ability of the detection system for the target pesticide under this formulation. The selected target formulation can maximize the reduction of interference from agricultural product matrix and improve the detection limit of low-concentration pesticide residues.

[0060] Step 3: Pesticide Residue Analysis: The extracts of each agricultural product sample in the target batch are mixed with the target formula to obtain the test mixture corresponding to each sample. Three-dimensional dynamic sequence analysis is performed on each test mixture. The pesticide residue concentration is calculated based on the analysis results. Multiple samples exceeding the standard are labeled, and graded early warning is implemented based on the labeling results.

[0061] In one specific embodiment, the three-dimensional dynamic sequence analysis of each sample of the target batch of agricultural products is performed as follows: the sample of each batch of agricultural products is injected into the detection cell of the electrochemical microfluidic chip, wherein the detection cell contains a platinum wire counter electrode, a silver chloride reference electrode and a carbon nanotube modified glassy carbon working electrode.

[0062] The electrochemical workstation applies voltage according to a three-stage procedure: the first stage is the enrichment stage, where a preset constant potential is maintained for a preset enrichment time to allow the target pesticide to be enriched on the surface of the working electrode; the second stage is the initial screening stage, which uses linear scanning voltammetry to perform a uniform scan at a preset scan rate within a preset potential scan range, simultaneously acquiring current-potential response data to obtain a linear scanning voltammetry curve; the third stage is the precision detection stage, which uses differential pulse voltammetry to apply pulse potentials according to preset pulse parameters within a preset potential detection range, simultaneously acquiring peak current response data corresponding to each pulse potential to obtain a differential pulse voltammetry curve.

[0063] A three-dimensional dynamic sequence matrix was constructed using the potential, time, and current response dimensions to complete the three-dimensional dynamic sequence analysis of the mixed liquid samples of each agricultural product in the target batch.

[0064] It should be noted that in the three-electrode system used in this step, the platinum wire counter electrode has stable conductivity and strong electrochemical inertness, ensuring the stability of the detection circuit; the silver chloride reference electrode provides a stable potential reference, avoiding potential drift during detection; the carbon nanotube-modified glassy carbon working electrode can significantly increase the electrode specific surface area, improve the enrichment efficiency and electron transfer efficiency of the target pesticide, and significantly enhance the sensitivity of the detection response signal; the design logic of the three-stage voltage program is as follows: in the enrichment stage, constant potential enrichment allows the target pesticide molecules to be directionally adsorbed on the surface of the working electrode, improving the response intensity of subsequent detection; the linear scanning voltammetry initial screening can quickly locate the characteristic oxidation peak potential range of the target pesticide, narrowing the range of subsequent accurate detection and eliminating interference from irrelevant potential ranges; the differential pulse voltammetry can effectively eliminate charging current interference during the detection process, significantly improving the accuracy of peak current detection at low concentrations; the three-dimensional dynamic sequence matrix integrates the full-process detection data of potential, time, and current, completely recording the entire response from enrichment to detection, providing multi-dimensional data support for subsequent characteristic peak identification and concentration calculation, and reducing the detection error caused by single curve analysis.

[0065] In one specific embodiment, the execution of graded early warning is carried out as follows: baseline correction is performed on the linear scanning voltammetric curve and the differential pulse voltammetric curve in the three-dimensional dynamic sequence matrix, and the characteristic oxidation peak potential and peak current value corresponding to the target pesticide are extracted as feature detection parameters; the feature detection parameters are fitted with the preset target pesticide standard curve to calculate the residual concentration of the target pesticide in the corresponding sample.

[0066] The pesticide residue concentration of each sample is compared with the preset maximum residue limit for the corresponding agricultural product; if the residue concentration is greater than or equal to the maximum residue limit, the sample is determined to be an excessive sample.

[0067] Based on the percentage of samples exceeding the standard and the multiple of the exceeding concentration, a three-level warning system is implemented: a level one warning is triggered when the concentration of a single sample exceeds the preset first maximum residue limit range, or when the percentage of samples exceeding the standard is less than or equal to the first percentage; a level two warning is triggered when the concentration of a single sample exceeds the second maximum residue limit range, or when the percentage of samples exceeding the standard is greater than the first percentage but less than the second percentage; and a level three warning is triggered when the concentration of a single sample exceeds the preset first maximum residue limit range, or when the percentage of samples exceeding the standard is the second percentage.

[0068] It should be noted that the baseline correction is as follows: the linear scanning voltammetry curve and the differential pulse voltammetry curve are corrected using an iterative polynomial fitting baseline method. The specific process is as follows: first, the voltammetry curve is fitted with a polynomial for 5 orders to obtain an initial fitted baseline; points in the curve whose difference from the fitted baseline is greater than 3 times the standard deviation are marked as characteristic peaks. After removing the characteristic peaks, the remaining data points are fitted with a polynomial again, and this process is repeated 3-5 times to obtain the final baseline; the baseline current value at the corresponding potential is subtracted from the current value of the original voltammetry curve to complete the baseline correction and eliminate background current interference during the detection process.

[0069] The fitting process with the preset target pesticide standard curve is as follows: At least five target pesticide standard solutions with different concentration gradients are pre-prepared. Using the same pretreatment and detection procedures as the sample to be tested, differential pulse voltammetry curves corresponding to each concentration standard solution are obtained, and the peak current values ​​of the characteristic oxidation peaks are extracted. With the target pesticide concentration as the abscissa and the corresponding peak current values ​​as the ordinate, a linear fit is performed using the least squares method to obtain the target pesticide standard curve. The expression for the standard curve is: Where I is the corrected peak current value, C is the pesticide concentration, k is the slope of the standard curve, and b is the intercept.

[0070] The calculation yields the residual concentration of the target pesticide in the corresponding sample. The specific calculation process is as follows: Substitute the peak current value corresponding to the characteristic oxidation peak potential in the corrected differential pulse voltammetry curve of the sample into the standard curve expression of the target pesticide to calculate the concentration of the target pesticide in the mixed solution to be tested. Then, combining the extraction solution volume, dilution factor, and sample mass during sample pretreatment, the actual residual concentration of the target pesticide in the agricultural product sample is calculated using the following formula: Where C_sample is the pesticide residue concentration in the agricultural product sample, C_mix is ​​the pesticide concentration calculated from the mixed solution to be tested, V is the volume of the extract, n is the dilution factor in the pretreatment process, and m is the sampling mass of the agricultural product sample.

[0071] For example, the three-tiered early warning system classifies different risk levels based on the extent to which batch samples exceed the limits. The first maximum residue limit (MRL) range can be set to 1 ≤ exceeding concentration < 2 times the MRL; the second MRL range can be set to 2 ≤ exceeding concentration < 5 times the MRL; and the third MRL range can be set to exceeding concentration ≥ 5 times the MRL. The first percentage can be set to 5%, the second percentage to 20%, and different warning levels correspond to different handling procedures: Level 1 warning corresponds to sample retesting, Level 2 warning corresponds to batch isolation and assessment, and Level 3 warning corresponds to prohibiting the batch from entering the market, thereby improving the efficiency of batch sample quality control. according to Figure 2 As shown, the present invention provides a multi-sample labeled pesticide residue detection system for agricultural products, including the following modules: an extraction state setting module, a mixed liquid state setting module, and a pesticide residue analysis module.

[0072] The mixture state setting module is connected to the extraction state setting module and the pesticide residue analysis module, respectively.

[0073] The extraction state setting module is used to take parallel test samples of each agricultural product sample in the target batch, first test the temperature of the sample to obtain the target temperature, and then test the pressure of the parallel test samples in the same batch to obtain the target pressure. Using the target temperature and target pressure as core parameters, the pressurized solvent extraction process of the target batch is set.

[0074] The mixed liquid state setting module is used to extract each agricultural product sample of the target batch according to the pressurized solvent extraction process of the target batch, to obtain the extract of the corresponding sample, and to mix the extract of each agricultural product sample with the matching reagent according to multiple preset mixing formulas to obtain the sample mixture corresponding to each mixing formula. The module collects and obtains the state data of the sample mixture corresponding to each mixing formula; and analyzes the state data of each mixing formula to obtain the target formula.

[0075] The pesticide residue analysis module is used to mix the extracts of each agricultural product sample in the target batch with the target formula to obtain the test mixture corresponding to each sample, perform three-dimensional dynamic sequence analysis on each test mixture, calculate the pesticide residue concentration based on the analysis results, label the samples that exceed the standard, and execute graded early warning based on the labeling results.

[0076] In this embodiment, the following preset values ​​can be set according to actual business needs: preset reference temperature, preset heating rate, preset temperature gradient value, preset sampling duration, preset enrichment duration, preset first constant potential, preset potential scanning interval, preset scanning rate, preset potential detection interval, preset pulse parameters, preset first maximum residue limit value interval, preset second maximum residue limit value interval, preset third maximum residue limit value interval, preset first proportion, and preset second proportion. For example, for the detection of organophosphorus pesticides in leafy vegetables, the preset reference temperature can be set to 25℃, the preset heating rate can be set to 5℃ / min, the preset temperature gradient value can be set to 10℃, the preset sampling duration can be set to 30s, and the preset enrichment duration can be set to... The preset time is 60 seconds. The preset first constant potential can be set to -0.8V, the preset potential scan range can be set to -0.8V to 1.2V, the preset scan rate can be set to 100mV / s, the preset potential detection range can be set to 0V to 1.0V, the preset pulse parameters can be set to pulse amplitude 50mV, pulse width 40ms, pulse period 200ms, the preset first maximum residue limit range can be set to 1 ≤ exceedance concentration < 2 times the maximum residue limit, the preset second maximum residue limit range can be set to 2 ≤ exceedance concentration < 5 times the maximum residue limit, the preset third maximum residue limit range can be set to exceedance concentration ≥ 5 times the maximum residue limit, the preset first proportion can be set to 5%, and the preset second proportion can be set to 20%.

[0077] The Peltier temperature-controlled stage, hyperspectral imager, vacuum-positive pressure dual-mode pressure generator, fluorescence spectrometer, ultraviolet-visible spectrometer, electrochemical microfluidic chip, electrochemical workstation, platinum wire counter electrode, silver chloride reference electrode, and carbon nanotube-modified glassy carbon working electrode described in this embodiment are all commercially available equipment and components that can be purchased from commercial channels. The pressure solvent extraction technology, linear scanning voltammetry, differential pulse voltammetry, Savitzky-Golay smoothing filtering algorithm, iterative polynomial baseline fitting method, least squares linear fitting, and cosine similarity calculation method are all existing mature technologies in the fields of analytical detection and data processing, and can be found in publicly available academic literature and industry standards. The fluorescent tracer with the same extraction kinetics as the target pesticide can be obtained by screening existing commercially available fluorescent reagents based on the physicochemical properties of the target pesticide. The relevant screening methods are existing technologies and will not be described in detail here.

[0078] The examples described in this invention are not limited to the specific embodiments listed above. The examples are merely illustrative to facilitate understanding of the invention and do not constitute a limitation on the scope of protection of this invention. Any modifications, equivalent substitutions, etc., made within the spirit and principles of this invention should be included within the scope of protection.

[0079] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the protection scope of the present invention.

Claims

1. A method for detecting pesticide residues in agricultural products with multi-sample labeling, characterized in that, Includes the following steps: Step 1, Extraction State Setting: Take parallel test samples of each agricultural product sample in the target batch, first test the temperature of the sample to obtain the target temperature, and then test the pressure of the parallel test samples in the same batch to obtain the target pressure. Using the target temperature and target pressure as the core parameters, set the pressurized solvent extraction process for the target batch. Step 2, Mixture State Setting: According to the pressurized solvent extraction process of the target batch, extract each agricultural product sample of the target batch to obtain the corresponding sample extract. According to the preset multiple mixing formulas, mix the extract of each agricultural product sample with the matching reagents to obtain the sample mixture corresponding to each mixing formula. Collect and obtain the state data of the sample mixture corresponding to each mixing formula. Analyze the state data of each mixture formulation to obtain the target formulation; Step 3: Pesticide Residue Analysis: The extracts of each agricultural product sample in the target batch are mixed with the target formula to obtain the test mixture corresponding to each sample. Three-dimensional dynamic sequence analysis is performed on each test mixture. The pesticide residue concentration is calculated based on the analysis results. Multiple samples exceeding the standard are labeled, and graded early warning is implemented based on the labeling results.

2. The method for detecting pesticide residues in agricultural products using multi-sample labeling according to claim 1, characterized in that, The temperature test of the sample is performed as follows: Place the parallel test sample on the Peltier temperature control platform, start the Peltier temperature control platform, and perform gradient heating according to the preset heating program: starting from the preset reference temperature, the temperature is increased uniformly at the preset heating rate. Each increase in the preset temperature gradient value is recorded as a temperature change node and the corresponding timestamp is recorded. The heating process is divided according to each timestamp to obtain multiple continuous temperature change segments. During the isothermal holding phase corresponding to each temperature change node, spectral reflectance images in the 400nm–700nm band are acquired using a hyperspectral imager to obtain spectral reflectance images corresponding to each temperature change segment.

3. The method for detecting pesticide residues in agricultural products with multi-sample labeling according to claim 2, characterized in that, The specific process for obtaining the target temperature is as follows: From the spectral reflectance images corresponding to each temperature change segment, extract the reflectance value of each pixel at a wavelength of 550 nm to construct a T×M array. 2 A 3D temperature reflectance matrix, where T is the number of temperature change segments and M is the pixel side length of the spectral reflectance image; The standard deviation of the temperature reflectivity matrix along the T dimension is calculated to obtain M. 2 A spatial standard deviation vector; the positions of elements in the spatial standard deviation vector whose absolute value is greater than a preset threshold are marked as high-response regions; The number of pixels in the high-response region corresponding to each temperature change segment is counted to form a T-dimensional response intensity vector. The response intensity vector is normalized so that the sum of its elements is 1. The temperature change segment number corresponding to the maximum value in the normalized response intensity vector is used as the target temperature identifier, and the end temperature of the temperature change segment corresponding to the target temperature identifier is the target temperature.

4. The method for detecting pesticide residues in agricultural products with multi-sample labeling according to claim 3, characterized in that, The pressure test was performed on parallel test samples from the same batch, and the specific test process is as follows: A fluorescent tracer with the same extraction kinetics as the target pesticide was mixed into the parallel test samples of the same batch to obtain each pressure analysis sample; Each pressure analysis sample is placed on a microporous polytetrafluoroethylene membrane inside a sealed extraction chamber. The vacuum-positive pressure dual-mode pressure generator is activated to set a gradient extraction pressure for each pressure analysis sample. At the target temperature, each pressure analysis sample is subjected to pressure extraction testing. Starting from the moment of pressure extraction, the fluorescence intensity values ​​of the characteristic peaks of the fluorescent tracer in the extract at the corresponding pressure are collected at preset sampling intervals to obtain the time-series change sequence of fluorescence intensity for each pressure analysis sample.

5. The method for detecting pesticide residues in agricultural products with multi-sample labeling according to claim 4, characterized in that, The specific process for obtaining the target pressure is as follows: A standard fluorescence intensity variation curve of the target pesticide under the condition of complete extraction is preset. The sampling time points of the standard variation curve correspond one-to-one with the sampling time in the pressure test. The fluorescence intensity values ​​of each sampling point in the standard variation curve are arranged in time sequence to obtain the basic fluorescence intensity sequence vector. From the time-series changes in fluorescence intensity of each pressure analysis sample, the fluorescence intensity values ​​of each sampling point that are strictly aligned with the sampling time points of the standard change curve are extracted, and the measured fluorescence intensity sequence vector of each pressure analysis sample is constructed. The cosine similarity between the basic fluorescence intensity sequence vector and the measured fluorescence intensity sequence vector of each pressure analysis sample is calculated to obtain the similarity between each pressure analysis sample and the standard variation curve; the extraction pressure of the pressure analysis sample corresponding to the maximum similarity is denoted as the target pressure.

6. The method for detecting pesticide residues in agricultural products with multi-sample labeling according to claim 5, characterized in that, The extracts from each agricultural product sample are mixed with the corresponding reagents. The specific mixing process is as follows: The extract of the target batch of agricultural product samples was injected into the micro-mixing chamber along with acetonitrile, phosphate buffer and sodium dodecyl sulfate solution according to the preset volume ratio of each mixing formula to obtain the sample mixture corresponding to each mixing formula. For each mixed formulation, a micro stirrer was started under a preset pressure. After stirring, the mixture was allowed to stand and separate into layers. Then, a UV-Vis spectrometer was triggered to collect UV absorption spectra in the 200nm–400nm band. Each mixed formulation corresponds to an absorption spectrum curve. The absorbance values ​​of each absorption spectrum curve at each wavelength were recorded as the state data of the agricultural product sample mixture of the corresponding mixed formulation.

7. The method for detecting pesticide residues in agricultural products with multi-sample labeling according to claim 6, characterized in that, The analysis of the state data of each mixture formulation is carried out in the following specific process: Baseline correction was performed on each absorption spectrum curve using Savitzky-Golay smoothing filtering. The absorbance value at 280 nm wavelength was extracted from the corrected spectrum as the characteristic response value of the corresponding mixed formulation. At the same time, the standard deviation of absorbance in the 200 nm-230 nm non-characteristic absorption band was extracted as the baseline noise value. The ratio of the characteristic response value to the baseline noise value is recorded as the signal-to-noise ratio of each mixture formulation. The signal-to-noise ratio is corrected by combining the slope of the UV absorption standard curve of the preset target pesticide: the slope of the UV absorption standard curve of the target pesticide is multiplied by the signal-to-noise ratio of each mixture formulation to obtain the corrected signal-to-noise ratio estimate of each mixture formulation. The mixture formulation corresponding to the maximum corrected signal-to-noise ratio estimate is selected as the target formulation.

8. The method for detecting pesticide residues in agricultural products with multi-sample labeling according to claim 1, characterized in that, The three-dimensional dynamic sequence analysis of each mixture to be tested is performed as follows: The mixed solution to be tested from each agricultural product sample of the target batch is injected into the detection cell of the electrochemical microfluidic chip. The detection cell contains a platinum wire counter electrode, a silver chloride reference electrode, and a carbon nanotube modified glassy carbon working electrode. The electrochemical workstation applies voltage according to a three-stage procedure: the first stage is the enrichment stage, which maintains the preset enrichment time at a preset first constant potential, so that the target pesticide is enriched on the surface of the working electrode. The second stage is the initial screening scanning stage, which uses the linear scanning voltammetry method. Within the preset potential scanning range, a uniform scanning is performed at a preset scanning rate, and current-potential response data is collected simultaneously to obtain the linear scanning voltammetry curve. The third stage is the precise detection stage, which uses the differential pulse voltammetry method. Within the preset potential detection range, a pulse potential is applied according to the preset pulse parameters, and the peak current response data corresponding to each pulse potential is collected simultaneously to obtain the differential pulse voltammetry curve. A three-dimensional dynamic sequence matrix was constructed using the potential, time, and current response dimensions to complete the three-dimensional dynamic sequence analysis of the mixed liquid samples of each agricultural product in the target batch.

9. The method for detecting pesticide residues in agricultural products with multi-sample labeling according to claim 8, characterized in that, The execution of tiered early warnings follows a specific warning process: Baseline correction was performed on the linear scan voltammetric curves and differential pulse voltammetric curves in the three-dimensional dynamic sequence matrix, and the characteristic oxidation peak potential and peak current values ​​corresponding to the target pesticide were extracted as feature detection parameters. The feature detection parameters are fitted with the preset target pesticide standard curve to calculate the residual concentration of the target pesticide in the corresponding sample. The pesticide residue concentration of each sample is compared with the preset maximum residue limit for the corresponding agricultural product; if the residue concentration is greater than or equal to the maximum residue limit, the sample is determined to be an excessive sample. Based on the percentage of samples exceeding the standard and the multiple of the exceeding concentration, a three-level warning system is implemented: a level one warning is triggered when the concentration of a single sample exceeds the preset first maximum residue limit range, or when the percentage of samples exceeding the standard is less than or equal to the first percentage; a level two warning is triggered when the concentration of a single sample exceeds the second maximum residue limit range, or when the percentage of samples exceeding the standard is greater than the first percentage but less than the second percentage; and a level three warning is triggered when the concentration of a single sample exceeds the preset first maximum residue limit range, or when the percentage of samples exceeding the standard is the second percentage.

10. A residue detection system utilizing the multi-sample labeled pesticide residue detection method for agricultural products according to any one of claims 1-9, characterized in that, Includes the following modules: The extraction state setting module is used to take parallel test samples of each agricultural product sample in the target batch, first test the temperature of the sample to obtain the target temperature, and then test the pressure of the parallel test samples in the same batch to obtain the target pressure. With the target temperature and target pressure as the core parameters, the pressurized solvent extraction process of the target batch is set. The mixed liquid state setting module is used to extract each agricultural product sample of the target batch according to the pressurized solvent extraction process of the target batch, to obtain the extract of the corresponding sample, and to mix the extract of each agricultural product sample with the matching reagent according to multiple preset mixing formulas to obtain the sample mixed liquid corresponding to each mixing formula, and to collect and obtain the state data of the sample mixed liquid corresponding to each mixing formula. Analyze the state data of each mixture formulation to obtain the target formulation; The pesticide residue analysis module is used to mix the extracts of each agricultural product sample in the target batch with the target formula to obtain the test mixture corresponding to each sample, perform three-dimensional dynamic sequence analysis on each test mixture, calculate the pesticide residue concentration based on the analysis results, label the samples that exceed the standard, and execute graded early warning based on the labeling results.

Images