A pretreatment method for rapid extraction of lead and cadmium in grain samples

Abstract

The present invention discloses a pretreatment method for rapidly extracting lead and cadmium from grain samples, specifically relating to the technical field of sample processing, including Step 1: preliminarily dividing N pulverized grain samples to obtain R sample types; Step 2: extracting the pulverized grain samples corresponding to the r-th sample type to obtain the r-th supernatant; and predicting the evaluation index of the r-th supernatant in a future time period; r = 1, 2, ……, R; The preliminary classification step of samples based on the homogeneous evaluation coefficient in the present invention ensures the scientificity of sample type division, thereby improving the pertinence and consistency of extraction. Secondly, by predicting the evaluation index of the supernatant and combining with a dynamically adjusted judgment mechanism, samples that do not meet the chelation index can be identified at an early stage, thereby avoiding subsequent processing of invalid samples and saving time and resources.

Description

Technical Field

[0001] The present invention relates to the technical field of sample processing, and more specifically, to a pretreatment method for rapidly extracting lead and cadmium from grain samples. Background Art

[0002] Pretreatment refers to a series of processing steps performed on a sample before analysis to extract target substances (such as lead, cadmium, etc.) and remove interfering substances, thereby improving the accuracy and sensitivity of the analysis. The conventional methods for heavy metal detection mainly include graphite furnace atomization method, flame atomization method, colorimetry, and atomic fluorescence method. These methods have a long pretreatment time and cannot improve the utilization efficiency of equipment. Therefore, how to shorten the pretreatment time while ensuring the extraction of heavy metals from grains as much as possible is a key research issue at present.

[0003] In existing methods, for example, the Chinese patent application with publication number CN113218727A discloses a pretreatment method for rapidly extracting lead and cadmium from grain samples, including: (1) adding an extraction solution to the grain sample for oscillating and mixing, and then centrifuging to collect the supernatant; the extraction solution is a hydrochloric acid solution; (2) passing the supernatant extracted in step (1) through a solid-phase extraction column, and then eluting the column with an acidic solution to collect the eluate; the preparation method of the extraction packing used in the solid-phase extraction column is: dissolving A336 in chloroform, then adding KS-C18 packing for mixing, and then rotary evaporating the chloroform in a fume hood to form C18 packing powder loaded with trioctylmethylammonium chloride; (3) neutralizing the eluate with an alkaline solution to pH 6-8, adding EDTA-2Na solution to form a metal chelate of EDTA, which is directly used for immunological detection. However, through research and application of the above method and the prior art, it is found that the above method and the prior art have at least the following partial defects:

[0004] An additional solid-phase extraction column treatment step is required, and the packing of the extraction column also needs to be prepared by oneself, which greatly increases the operation complexity and the workload during the experiment.

[0005] Therefore, the present invention provides a pretreatment method for rapidly extracting lead and cadmium from grain samples. Summary of the Invention

[0006] In order to overcome the above-mentioned defects of the prior art, the present invention provides a pretreatment method for rapidly extracting lead and cadmium from grain samples to solve the problems raised in the above background art.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] In a first aspect, the present invention provides a pretreatment method for rapidly extracting lead and cadmium from grain samples, including:

[0009] Step 1: Conduct a preliminary division on N crushed grain samples to obtain R sample types;

[0010] Step 2: Extract the crushed grain samples corresponding to the r-th sample type to obtain the r-th supernatant; and predict the r-th supernatant evaluation index for a future time period; r = 1, 2, ……, R;

[0011] Step 3: Based on the r-th supernatant evaluation index, determine whether the supernatant meets the chelation index; if it meets, generate a chelation instruction and jump to Step 4; if it does not meet, set r = r + 1 and return to Step 2;

[0012] Step 4: Receive the chelation instruction, perform a neutralization and chelation reaction on the r-th supernatant to obtain the r-th chelation sample, and obtain the reaction characteristic data of the r-th chelation sample;

[0013] Step 5: Input the reaction characteristic data into a pre-constructed chelation analysis model to obtain the chelation evaluation coefficient of the r-th chelation sample for a future time period;

[0014] Step 6: Based on the chelation evaluation coefficient output by the chelation analysis model, determine whether the r-th chelation sample meets the requirements; if it meets, store the chelation sample; if it does not meet the requirements, set r = r + 1 and return to Step 2.

[0015] Further, the method for obtaining the R sample types includes:

[0016] Step a1: Obtain the homogenization evaluation coefficient of the n-th crushed grain sample;

[0017] Step a2: Set N homogenization evaluation coefficient intervals, and set N sample types corresponding to the N homogenization evaluation coefficient intervals; each of the homogenization evaluation coefficient intervals is associated and bound with only one sample type;

[0018] Step a3: Compare the homogenization evaluation coefficient of the n-th crushed grain sample with each homogenization evaluation coefficient interval to obtain the homogenization evaluation coefficient interval into which the homogenization evaluation coefficient of the n-th crushed grain sample falls;

[0019] Step a4: According to the homogenization evaluation coefficient interval into which the homogenization evaluation coefficient of the n-th crushed grain sample falls, divide the n-th crushed grain sample into the corresponding sample type, set n = n + 1, and jump back to Step a1;

[0020] Step a5: Repeat the above Steps a1 - a4 until the loop ends when n = N, so that each of the crushed grain samples is divided into R sample types in turn.

[0021] Further, the method for obtaining the homogenization evaluation coefficient of the nth pulverized grain sample includes:

[0022] Step b1: Use a pulverizer to pulverize the grain sample to the target particle size to obtain a pulverized grain sample; the pulverizer includes a universal mill or a ball mill;

[0023] Step b2: Obtain the basic grain data of the pulverized grain sample, where the basic grain data includes grain hardness, water content, and particle size;

[0024] Step b3: Perform formula-based calculations on the basic grain data to obtain the homogenization evaluation coefficient.

[0025] Further, the method for obtaining the rth supernatant includes:

[0026] Step c1: Weigh the weight of the pulverized grain sample corresponding to the rth sample type;

[0027] Step c2: Obtain the amount of extraction solution according to the mixing ratio formula and add it to the pulverized grain sample to obtain a first mixture;

[0028] Among them, the calculation method of the mixing ratio formula is ;

[0029] In the formula, represents the amount of extraction solution, represents the weight of the pulverized grain sample, represents the solid ratio;

[0030] Step c3: Transfer the first mixture to an oscillator with preset oscillation parameters for oscillation to obtain a second mixture;

[0031] Step c4: Place the second mixture in a centrifuge with preset centrifugation parameters for separation;

[0032] Step c5: Extract the transparent liquid of the second mixture after separation by a pipette to obtain the rth supernatant.

[0033] Further, the method for predicting the evaluation index of the rth supernatant in a future time period includes:

[0034] Step d1: Obtain the supernatant characterization data of the rth supernatant in the current time period, where the supernatant characterization data includes the pH value, volume, preliminary metal ion concentration, reaction rate constant, and diffusion rate of the supernatant;

[0035] Step d2: Input the supernatant characterization data of the rth supernatant in the current time period into a pre-constructed coefficient prediction model to obtain the evaluation index of the rth supernatant in a future time period.

[0036] Further, the training method of the coefficient prediction model includes:

[0037] Obtain historical coefficient training data, and divide the historical coefficient training data into a coefficient training set and a coefficient test set; the historical coefficient training data includes supernatant characterization data and its corresponding supernatant evaluation index;

[0038] Among them, the method for obtaining the supernatant evaluation index in the historical coefficient training data includes:

[0039] Step d01: Obtain the supernatant characterization data of the r-th supernatant within a set time span;

[0040] Step d02: Perform formula calculation on the supernatant characterization data to obtain the supernatant evaluation index, and its calculation formula is as follows:

[0041]

[0042] In the formula, represents the supernatant evaluation index, represents the pH value, represents the volume, represents the preliminary metal ion concentration, represents the reaction rate constant, represents the diffusion rate; , , , and are weight factors, , represents the logarithmic function with the natural constant e as the base;

[0043] Construct a regression network model, use the supernatant characterization data in the coefficient training set as the input data of the regression network model, use the supernatant evaluation index in the coefficient training set as the output data of the regression network model, train the regression network model, and obtain an initial coefficient prediction regression network;

[0044] Use the coefficient test set to verify the initial coefficient prediction regression network, and output the initial coefficient prediction regression network with a prediction error less than or equal to the preset error threshold as the coefficient prediction model; the regression network model is specifically an RNN recurrent neural network model.

[0045] Further, the method for judging whether the supernatant meets the chelation index based on the r-th supernatant evaluation index includes:

[0046] Step e1: Extract the time series data and the corresponding supernatant evaluation index as

[0047] ;

[0048] Step e2: Smooth the supernatant evaluation index corresponding to the time series data to remove noise and outliers. The calculation formula is as follows: ; In the formula, represents the smoothed value at the current time t, represents the actual observed value at the current time t, represents the smoothed value at the previous time t−1, represents the smoothing coefficient, and the value range is 0.2 - 0.5;

[0049] Step e3: Calculate the change amplitude of the smoothed sequence :

[0050]

[0051] Step e4: Compare the change amplitude with the preset change amplitude threshold ;

[0052] If , it indicates that the r-th supernatant evaluation index tends to be stable;

[0053] If , it indicates that the r-th supernatant evaluation index is unstable;

[0054] Step e5: Compare the r-th supernatant evaluation index that tends to be stable with the preset evaluation index threshold; the evaluation index threshold includes and , where > ; Compare the supernatant evaluation index with the preset evaluation index threshold;

[0055] If > or , no chelation instruction is generated;

[0056] If , a chelation instruction is generated.

[0057] Furthermore, the method for obtaining the r-th chelation sample includes:

[0058] Step f1: Gradually add an alkaline solution to the r-th supernatant to adjust the pH value of the solution to the target value. The alkaline solution is NaOH or Na 2 CO 3 , and the target value is 6 - 8;

[0059] Step f2: Dissolve EDTA in distilled water and slowly add it to the neutralized supernatant for chelation reaction. The EDTA is EDTA-2Na, and the volume of the EDTA is 1 / 5 to 1 / 2 of the volume of the supernatant.

[0060] Step f3: Set the chelation parameters and environmental parameters. The chelation parameters include the stirring speed (200 - 500 rmp) and the stirring time (10 - 15 min), and the environmental parameters include temperature control and light intensity (no light).

[0061] Step f4: The finally obtained solution is the r-th chelation sample.

[0062] Furthermore, the reaction characteristic data includes the formation rate of metal chelates, the difference in optical density, and the difference in conductivity.

[0063] The training method of the chelation analysis model includes:

[0064] Obtain historical chelation training data, which includes reaction characteristic data within multiple time spans and corresponding chelation evaluation coefficients.

[0065] Among them, the acquisition logic of the chelation evaluation coefficient in the historical chelation training data is as follows:

[0066] Extract the reaction characteristic data from the historical chelation training data, perform formulaic calculations on the reaction characteristic data, and obtain the chelation evaluation coefficient. The calculation formula is:

[0067] ;

[0068] In the formula, represents the chelation evaluation coefficient of the r-th chelation sample in the future time period, represents the formation rate of metal chelates, represents the difference in optical density, represents the difference in conductivity, , , are all weighting factors, .

[0069] Divide the historical chelation training data into a chelation training set and a chelation test set, construct a regression network model, use the reaction feature data in the chelation training set as the input of the regression network model, use the chelation evaluation coefficient in the chelation training set as the output of the regression network model, train the regression network model to obtain an initial regression network, with minimizing the sum of prediction accuracies as the training objective, use the chelation test set to evaluate the initial regression network model, and use the initial regression network model when the sum of prediction accuracies reaches convergence as the pre-constructed chelation analysis model; the regression network model is an RNN model, a support vector machine regression network model, a linear regression network model or a random forest regression network model.

[0070] Further, the method for judging whether the r-th chelation sample meets the requirements according to the chelation evaluation coefficient includes:

[0071] Preset a chelation coefficient threshold, and compare the chelation evaluation coefficient with the preset chelation coefficient threshold for comparison;

[0072] If , it is judged that the chelation sample does not meet the requirements;

[0073] If , it is judged that the chelation sample meets the requirements.

[0074] In a second aspect, the present invention provides a pretreatment system for quickly extracting lead and cadmium from grain samples, including:

[0075] A partitioning module for preliminarily partitioning N pulverized grain samples to obtain R sample types;

[0076] A first prediction module for extracting the pulverized grain samples corresponding to the r-th sample type to obtain the r-th supernatant; and predicting the r-th supernatant evaluation index in a future time period; r = 1, 2,..., R;

[0077] A first judgment module for judging whether the supernatant meets the chelation index based on the r-th supernatant evaluation index; if it meets, generate a chelation instruction and jump to the chelation module, if it does not meet, set r = r + 1 and return to the first prediction module;

[0078] A chelation module for receiving the chelation instruction, performing a neutralization and chelation reaction on the r-th supernatant to obtain the r-th chelation sample, and obtaining the reaction feature data of the r-th chelation sample;

[0079] An analysis module for inputting the reaction feature data into the pre-constructed chelation analysis model to obtain the chelation evaluation coefficient of the r-th chelation sample in a future time period;

[0080] A second judgment module, configured to judge whether the r-th chelation sample meets the requirements according to the chelation evaluation coefficient output by the chelation analysis model. If it meets the requirements, the chelation sample is stored; if it does not meet the requirements, set r = r + 1 and return to the first prediction module.

[0081] The technical effects and advantages of the present invention:

[0082] 1. Based on the preliminary classification step of the sample according to the homogeneous evaluation coefficient, the present invention ensures the scientificity of the sample type division, thereby improving the pertinence and consistency of extraction. Secondly, by predicting the evaluation index of the supernatant and combining with the dynamically adjusted judgment mechanism, samples that do not meet the chelation index can be identified at an early stage, thus avoiding subsequent processing of invalid samples and saving time and resources. At the same time, by training the chelation analysis model based on the reaction characteristic data and combining with the data-driven judgment process, samples that meet the quality standards can be accurately screened. In summary, the system can not only quickly adapt to different sample types, but also ensure the accuracy and repeatability of the extraction results, greatly improving the resource utilization efficiency and the degree of intelligence of the processing process during extraction.

[0083] 2. Through the application of the regression network model, the system can learn complex non-linear relationships from historical data and accurately predict the evaluation indexes of samples in future time periods, providing reliable data support for the extraction process. In addition, the system is provided with multiple judgment modules to ensure that the neutralization and chelation reactions of the samples are only carried out on qualified samples, greatly reducing unnecessary experimental operations. Finally, the method significantly reduces the need for manual intervention and realizes the high efficiency, accuracy and automation of the pretreatment work of food samples through intelligent data analysis and real-time adjustment of process parameters. Description of the Drawings

[0084] Figure 1 It is a flowchart of a pretreatment method for quickly extracting lead and cadmium in food samples in Example 1;

[0085] Figure 2 It is a flowchart of the method for obtaining R sample types in Example 1;

[0086] Figure 3 It is a flowchart of the prediction method for the evaluation index of the r-th supernatant in the future time period in Example 1;

[0087] Figure 4 It is a schematic structural diagram of a pretreatment system for quickly extracting lead and cadmium in food samples in Example 2. Detailed Embodiments

[0088] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0089] In addition, the accompanying drawings are only schematic diagrams of the present invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus repeated descriptions thereof will be omitted. Some of the block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities may be implemented in software, or in one or more hardware modules or integrated circuits, or in different networks and / or processor methods and / or microcontroller methods.

[0090] It should be understood that although terms such as "first" and "second" may be used herein to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly, the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the associated listed items.

[0091] The supernatant contains a large amount of interfering substances, resulting in inaccurate analysis results, so it cannot be used directly. Currently, it is processed in a large, thorough and costly manner; since the supernatant of the pulverized grain sample will gradually precipitate the interfering substances insoluble in water over a period of time, that is, the components in the supernatant change dynamically with time. By virtue of this characteristic, in the embodiments of the present application, with the support of an algorithm, the supernatant is evaluated at different time points, and in a predictive manner, it is evaluated that the supernatant meets the subsequent chelation reaction conditions at a certain future time point, and a supernatant meeting the requirements can be obtained without processing in a large, thorough and costly manner. The following is a detailed description of the present application.

[0092] Embodiment 1

[0093] Please refer to Figure 1 As shown, this embodiment discloses and provides a pretreatment method for quickly extracting lead and cadmium from grain samples, and the method includes:

[0094] Step 1: Initially divide N pulverized grain samples to obtain R sample types;

[0095] It should be noted that: the crushed grain sample refers to a sample obtained through a process of physically crushing and grinding grains (such as wheat, rice, or corn) into fine particles or powder state. The purpose of crushing is to improve the uniformity of the sample and the accuracy of analysis and detection, making it suitable for subsequent chemical analysis or biological detection. In this embodiment, wheat, rice, or corn is taken as an example for illustration. The sample types include coarse particle samples and fine particle samples, etc.

[0096] Please refer to Figure 2 As shown, in the implementation, the method for obtaining R sample types includes:

[0097] Step a1: Obtain the homogenization evaluation coefficient of the nth crushed grain sample;

[0098] Step a2: Set N homogenization evaluation coefficient intervals, and set N sample types corresponding to the N homogenization evaluation coefficient intervals; each of the homogenization evaluation coefficient intervals is associated and bound with exactly one sample type;

[0099] Step a3: Compare the homogenization evaluation coefficient of the nth crushed grain sample with each homogenization evaluation coefficient interval to obtain the homogenization evaluation coefficient interval into which the homogenization evaluation coefficient of the nth crushed grain sample falls;

[0100] Step a4: According to the homogenization evaluation coefficient interval into which the homogenization evaluation coefficient of the nth crushed grain sample falls, classify the nth crushed grain sample into the corresponding sample type, and let n = n + 1, then jump back to step a1;

[0101] Step a5: Repeat the above steps a1 - a4 until the loop ends when n = N, so that each of the crushed grain samples is classified into R sample types in turn;

[0102] Among them, the method for obtaining the homogenization evaluation coefficient of the nth crushed grain sample includes:

[0103] Step b1: Use a crusher to crush the grain sample to the target particle size to obtain a crushed grain sample; the crusher includes a universal mill or a ball mill; the target particle size is set by those skilled in the art, and the reference value is 50 - 100 μm.

[0104] Step b2: Obtain the basic grain data of the crushed grain sample, and the basic grain data includes grain hardness, moisture content, and particle size;

[0105] It should be noted that: the grain hardness is the ability of the grain kernel to resist the extrusion of external forces, which is collected by a grain hardness tester; the moisture content is the percentage of the mass of water contained in the grain in the total mass of the grain, which is measured by a capacitive moisture meter; the particle size is the average particle diameter of the grain, which is collected by a laser particle size analyzer.

[0106] Step b3: Perform a formulaic calculation on the basic grain data to obtain a homogenization evaluation coefficient, and its calculation formula is:

[0107]

[0108] In the formula, represents the homogenization evaluation coefficient, represents the grain hardness, represents the moisture content, represents the granularity; , and are weight factors greater than zero, and , and the weight factors are all set by those skilled in the art according to experience.

[0109] In large-scale grain research, if random sampling is not scientifically classified, it is very easy to mix in individuals with large differences, resulting in the sample not being able to represent the overall population. This step reduces the data dispersion and shrinks the deviation during subsequent analysis by scientifically classifying the sample types, making the obtained results more in line with the actual grain population situation.

[0110] Step 2: Extract the crushed grain sample corresponding to the r-th sample type to obtain the r-th supernatant; and predict the r-th supernatant evaluation index within the future time period; r = 1, 2, ……, R;

[0111] In implementation, the method for obtaining the r-th supernatant includes:

[0112] Step c1: Weigh the weight of the crushed grain sample corresponding to the r-th sample type;

[0113] Step c2: Obtain the dosage of the extraction solution according to the mixing ratio formula and add it to the crushed grain sample to obtain a first mixture, and the extraction solution includes dilute hydrochloric acid solution, nitric acid solution or buffer solution;

[0114] Among them, the calculation method of the mixing ratio formula is ;

[0115] In the formula, represents the dosage of the extraction solution, represents the weight of the crushed grain sample, represents the solid ratio; it should be noted that the solid ratio is set by those skilled in the art according to the actual situation and is not specifically limited here.

[0116] Step c3: Transfer the first mixture to an oscillator with preset oscillation parameters for oscillation to obtain a second mixture;

[0117] It should be noted that: the preset oscillation parameters such as the oscillation amplitude are 100 - 200 rmp, and the oscillation time is 3 - 5 min. The specific parameters are set by those skilled in the art according to experience and will not be elaborated here.

[0118] Step c4: Place the second mixture in a centrifuge with preset centrifugation parameters for separation;

[0119] It should be noted that: the preset centrifugation parameters such as the rotation speed are 3000 - 5000 rmp, and the centrifugation time is 1 - 3 min. The specific parameters are set by those skilled in the art according to experience and will not be elaborated here.

[0120] Step c5: Extract the transparent liquid of the separated second mixture with a pipette to obtain the r-th supernatant.

[0121] Please refer to Figure 3 As shown, in the implementation, the prediction method for the r-th supernatant evaluation index in the future time period includes:

[0122] Step d1: Obtain the supernatant characterization data of the r-th supernatant in the current time period. The supernatant characterization data includes the pH value, volume, preliminary metal ion concentration, reaction rate constant, and diffusion rate of the supernatant.

[0123] It should be noted that: the pH value of the supernatant is obtained by measurement with a pH meter; the volume is directly obtained by measurement with a measuring cylinder, and the preliminary metal ion concentration is obtained by determination with an inductively coupled plasma mass spectrometer (ICP-MS) or an atomic absorption spectrometer (AAS); the metal ion is lead ion or cadmium ion. The reaction rate constant is calculated by monitoring the change in the metal ion concentration of the chelation reaction using the chemical kinetics formula; its calculation formula is:

[0124] ;

[0125] In the formula, represents the metal ion concentration at a certain time point after the reaction; represents the initial metal ion concentration of the reaction, and T represents the reaction time.

[0126] The diffusion rate is determined by a diffusion cell experiment or a membrane diffusion experiment, and its calculation formula is: ; In the formula, represents the diffusion rate, represents the diffusion flux, represents the concentration difference;

[0127] It should be noted that: the diffusion flux represents the amount of substance passing through per unit area per unit time during the diffusion process, and its calculation formula is: ;

[0128] In the formula, represents the diffusion flux, represents the diffusion area, and K represents the diffusion time, represents the concentration difference at time K, represents the integral of the concentration difference with respect to time from time 0 to The diffusion area represents the effective area of contact between the two liquids in the diffusion cell.

[0129] It should be added that the diffusion rate can also be obtained by calculation through a membrane diffusion experiment, which is not the focus of the invention, so the present invention will not elaborate too much on this; any technology that can achieve the diffusion rate in the prior art can be used as the application object of the present invention.

[0130] Step d2: Input the supernatant characterization data of the r-th supernatant in the current time period into the pre-constructed coefficient prediction model to obtain the evaluation index of the r-th supernatant in the future time period.

[0131] Among them, the training method of the coefficient prediction model includes:

[0132] Obtain historical coefficient training data, and divide the historical coefficient training data into a coefficient training set and a coefficient test set; the historical coefficient training data includes supernatant characterization data and its corresponding supernatant evaluation index;

[0133] Among them, the method for obtaining the supernatant evaluation index in the historical coefficient training data includes:

[0134] Step d01: Obtain the supernatant characterization data of the r-th supernatant within a set time span;

[0135] Step d02: Perform formula calculation on the supernatant characterization data to obtain the supernatant evaluation index, and its calculation formula is as follows:

[0136]

[0137] In the formula, represents the supernatant evaluation index, represents the pH value, represents the volume, represents the preliminary metal ion concentration, represents the reaction rate constant, represents the diffusion rate; , , , and are weight factors, The weight factors are all set by those skilled in the art according to experience, represents the logarithmic function with the natural constant e as the base.

[0138] Construct a regression network model. Use the supernatant characterization data in the coefficient training set as the input data of the regression network model, and use the supernatant evaluation index in the coefficient training set as the output data of the regression network model. Train the regression network model to obtain an initial coefficient prediction regression network;

[0139] Use the coefficient test set to verify the model of the initial coefficient prediction regression network, and output the initial coefficient prediction regression network with a prediction error less than or equal to the preset error threshold as the coefficient prediction model; The regression network model is specifically an RNN recurrent neural network model.

[0140] Obtaining the r-th supernatant evaluation index within a future time period can optimize the reagent dosage or adjust the experimental steps, reduce the number of experiments, and help judge the sample stability and extraction efficiency. This step ensures that the extraction efficiency and the chemical properties of the sample matrix meet the requirements of subsequent detections.

[0141] Step 3: Judge whether the supernatant meets the chelation index based on the r-th supernatant evaluation index; if it meets, generate a chelation instruction and jump to Step 4; if it does not meet, set r = r + 1 and return to Step 2;

[0142] In implementation, the method for judging whether the supernatant meets the chelation index based on the r-th supernatant evaluation index includes:

[0143] Step e1: Extract the time series data and the corresponding supernatant evaluation index as

[0144] ;

[0145] Step e2: Smooth the supernatant evaluation index corresponding to the time series data to remove noise and outliers. The calculation formula is:

[0146] ;

[0147] In the formula, represents the smoothed value at the current time t, represents the actual observed value at the current time t, represents the smoothed value at the previous time t - 1, represents the smoothing coefficient, and its value range is 0.2 - 0.5.

[0148] Step e3: Calculate the change amplitude of the smoothed sequence :

[0149]

[0150] Step e4: Compare the change amplitude with the preset change amplitude threshold for comparison;

[0151] If , it indicates that the r-th supernatant evaluation index tends to be stable;

[0152] If , it indicates that the r-th supernatant evaluation index is unstable;

[0153] Step e5: Compare the r-th supernatant evaluation index that tends to be stable with a preset evaluation index threshold; the evaluation index threshold includes and , where > ; Compare the supernatant evaluation index with the preset evaluation index threshold;

[0154] If > or , no chelation instruction is generated;

[0155] If , a chelation instruction is generated.

[0156] By performing a stability analysis on the evaluation indexes in multiple future time periods, this step can comprehensively grasp the state of the sample extraction process, ensure the accuracy and reliability of the final judgment result, and provide a scientific basis for optimizing the experimental conditions.

[0157] Step 4: Receive the chelation instruction, perform a neutralization and chelation reaction on the r-th supernatant to obtain the r-th chelated sample, and obtain the reaction characteristic data of the r-th chelated sample;

[0158] It should be noted that: The collected supernatant is neutralized with an alkaline solution to an appropriate pH range (6 - 8), and then an ITCBE-EDTA solution is added to form stable EDTA metal chelates with lead and cadmium in the sample for immunological detection.

[0159] In implementation, the method for obtaining the r-th chelated sample includes:

[0160] Step f1: Gradually add an alkaline solution to the r-th supernatant to adjust the pH value of the solution to the target value, the alkaline solution is NaOH or Na 2 CO 3 , and the target value is 6 - 8;

[0161] Step f2: Dissolve EDTA in distilled water and slowly add it to the neutralized supernatant for chelation reaction, the EDTA is EDTA-2Na, and the volume of the EDTA is 1 / 5 to 1 / 2 of the volume of the supernatant;

[0162] Step f3: Set the chelation parameters and environmental parameters. The chelation parameters include the stirring speed (200 - 500 rmp) and the stirring time (10 - 15 min), and the environmental parameters include temperature control and light intensity (no light); specifically, they can be set according to the experience in this field.

[0163] Step f4: The finally obtained solution is the r-th chelation sample, in which the metal ions have been chelated by EDTA.

[0164] The reaction characteristic data include the formation rate of the metal chelate, the optical density difference, and the conductivity difference.

[0165] It should be noted that: the formation rate refers to the proportion of the target metal ions completely combined with the chelating agent in the chelation reaction; the change in optical density reflects the change in the transparency or light absorption characteristics of the solution, indirectly characterizing the change in the chemical components in the solution during the chelation reaction, and the conductivity difference reflects the change in the ion concentration in the solution, indirectly characterizing the process of the chelating agent binding with the metal ions.

[0166] In implementation, the method for obtaining the formation rate in the reaction characteristic data includes:

[0167] Use atomic absorption spectrometry (AAS) or inductively coupled plasma mass spectrometry (ICP-MS) to measure the initial concentration of metal ions in the supernatant; the metal ions are lead ions or cadmium ions.

[0168] After the chelation reaction is completed, use the same instrument to measure the concentration of unchelated metal ions in the solution; calculate the formation rate ;

[0169] In the formula, represents the formation rate of the metal chelate, represents the initial concentration, represents the concentration of unchelated metal ions;

[0170] It should be noted that: the formation rate CCL is a positive value, and the more complete the chelation reaction, the greater the formation rate CCL.

[0171] In implementation, the method for obtaining the optical density difference in the reaction characteristic data includes:

[0172] According to the characteristics of the metal chelate, select a specific wavelength (such as 260 nm or 280 nm) to measure the optical density, and use a UV-visible spectrophotometer to detect the optical density of the supernatant without adding the chelating agent to obtain the initial optical density;

[0173] After the chelation reaction is completed, measure the optical density of the chelation sample, marked as the final optical density;

[0174] Calculate the optical density difference .

[0175] In the formula, represents the optical density difference value, represents the final optical density, represents the initial optical density.

[0176] It should be noted that: after the chelation reaction, heavy metal ions and the chelating agent form a chelate. When the chelate absorbs a specific wavelength selected (such as 260 nm or 280 nm), the transmittance of the solution becomes smaller and the optical density becomes larger. Then, the optical density difference value MDC is positive, and the more sufficient the chelation reaction is, the larger the optical density difference value MDC is.

[0177] It is worth noting that: ensure the uniformity of the chelation sample, without particle precipitation or impurity interference;

[0178] In implementation, the method for obtaining the conductivity difference value in the reaction characteristic data includes:

[0179] Use a high-precision conductivity meter to measure the total conductivity of metal ions and other ions in the supernatant and mark it as the initial conductivity;

[0180] During the progress of the neutralization and chelation reactions, collect the conductivity at preset intervals (such as every 30 seconds), record the conductivity change curve until the conductivity tends to be stable;

[0181] After the chelation reaction ends, record the final conductivity of the chelation sample;

[0182] Perform a difference calculation on the initial conductivity and the final conductivity to obtain the conductivity difference value, and its calculation formula is: .

[0183] In the formula, represents the conductivity difference value, represents the initial conductivity, represents the final conductivity.

[0184] It should be noted that: after the chelation reaction, heavy metal ions and the chelating agent form a chelate. The fewer the ion concentration in the solution, the more the conductivity decreases. Then, the conductivity difference value DDC is positive, indicating that the chelation reaction is more sufficient and the conductivity difference value DDC is larger.

[0185] Step 5: Input the reaction characteristic data into the pre-constructed chelation analysis model to obtain the chelation evaluation coefficient of the r-th chelation sample in the future time period;

[0186] In implementation, the training method of the chelation analysis model includes:

[0187] Obtain historical chelation training data, where the historical chelation training data includes reaction characteristic data within multiple time spans and corresponding chelation evaluation coefficients;

[0188] It should be noted that: the specific duration of each time span is determined according to the time span preset by those skilled in the art.

[0189] Among them, the acquisition logic of the chelation evaluation coefficient in the historical chelation training data is as follows:

[0190] Extract the reaction characteristic data in the historical chelation training data, perform formulaic calculations on the reaction characteristic data to obtain the chelation evaluation coefficient; its calculation formula is:

[0191] ;

[0192] In the formula, represents the chelation evaluation coefficient of the rth chelation sample in the future time period, represents the formation rate of the metal chelate, represents the optical density difference, represents the conductivity difference, , , are all weighting factors, .

[0193] Divide the historical chelation training data into a chelation training set and a chelation test set, construct a regression network model, use the reaction characteristic data in the chelation training set as the input of the regression network model, use the chelation evaluation coefficient in the chelation training set as the output of the regression network model, train the regression network model to obtain an initial regression network, with minimizing the sum of prediction accuracies as the training objective, use the chelation test set to evaluate the initial regression network model, and use the initial regression network when the sum of prediction accuracies reaches convergence as the pre-constructed chelation analysis model; the regression network model is an RNN model, a support vector machine regression network model, a linear regression network model, or a random forest regression network model.

[0194] Step 6: Judge whether the rth chelation sample meets the requirements according to the chelation evaluation coefficient output by the chelation analysis model. If it meets the requirements, store the chelation sample. If it does not meet the requirements, let r = r + 1, and return to Step 2.

[0195] In implementation, the method for judging whether the rth chelation sample meets the requirements according to the chelation evaluation coefficient includes:

[0196] Preset a chelation coefficient threshold, and compare the chelation evaluation coefficient with the preset chelation coefficient threshold for comparison;

[0197] If , it indicates that a stable chelate has not been formed, resulting in an abnormally high detected value of the metal ion concentration in the chelated sample, and it is determined that the chelated sample does not meet the requirements;

[0198] If , it is determined that the chelated sample meets the requirements, and then the chelated sample is stored or subjected to subsequent heavy metal detection and analysis.

[0199] The metal ion is a lead ion or a cadmium ion.

[0200] Based on the preliminary classification steps of the samples according to the homogenization evaluation coefficient in this embodiment, the scientific nature of the sample type division is ensured, thereby improving the pertinence and consistency of extraction. Secondly, by predicting the evaluation index of the supernatant and combining the dynamically adjusted judgment mechanism, samples that do not meet the chelation index can be identified at an early stage, thus avoiding subsequent processing of invalid samples and saving time and resources. At the same time, training the chelation analysis model based on the reaction characteristic data and combining the data-driven judgment process can accurately screen samples that meet the quality standards. In summary, this system can not only quickly adapt to different sample types, but also ensure the accuracy and repeatability of the extraction results, greatly improving the resource utilization efficiency and the degree of intelligence of the processing flow during the extraction process;

[0201] Through the application of the regression network model in this embodiment, the system can learn complex non-linear relationships from historical data and perform high-precision prediction on the evaluation indexes of samples in future time periods, providing reliable data support for the extraction process. In addition, the system is provided with a multi-layer judgment module to ensure that the neutralization and chelation reactions of the samples are only carried out on qualified samples, greatly reducing unnecessary experimental operations. Finally, this method significantly reduces the need for manual intervention and realizes the high efficiency, precision and automation of the pretreatment work of grain samples through intelligent data analysis and real-time adjustment of process parameters.

[0202] Embodiment 2

[0203] Please refer to Figure 4 As shown, this embodiment provides a pretreatment system for quickly extracting lead and cadmium from grain samples, including a division module, a first prediction module, a first judgment module, a chelation module, an analysis module and a second judgment module; each module is connected by wired and / or wireless means to realize data transmission between modules;

[0204] The division module is used for preliminarily dividing N pulverized grain samples to obtain R sample types;

[0205] The first prediction module is used for extracting the pulverized grain samples corresponding to the rth sample type to obtain the rth supernatant; and predicting the evaluation index of the rth supernatant in the future time period; r = 1, 2,..., R;

[0206] The first judgment module determines whether the supernatant meets the chelation index based on the r-th supernatant evaluation index; if it meets the requirement, it generates a chelation instruction and jumps to the chelation module; if it does not meet the requirement, it sets r = r + 1 and returns to the first prediction module.

[0207] The chelation module is used to receive the chelation instruction, perform a neutralization and chelation reaction on the r-th supernatant to obtain the r-th chelation sample, and obtain the reaction characteristic data of the r-th chelation sample.

[0208] The analysis module is used to input the reaction characteristic data into a pre-constructed chelation analysis model to obtain the chelation evaluation coefficient of the r-th chelation sample in a future time period.

[0209] The second judgment module is used to determine whether the r-th chelation sample meets the requirements according to the chelation evaluation coefficient output by the chelation analysis model. If it meets the requirement, it stores the chelation sample; if it does not meet the requirement, it sets r = r + 1 and returns to the first prediction module.

[0210] Embodiment 3

[0211] This embodiment provides an electronic device, including: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps in the above method embodiment, such as Figure 1 the flowchart shown. Alternatively, when the processor executes the computer program, it implements the functions of each module / unit in the above system embodiments, such as Figure 4 the system structure diagram shown.

[0212] Exemplarily, the computer program can be divided into one or more modules / units. The one or more modules / units are stored in the memory and executed by the processor to complete this application. The one or more modules / units can be a series of computer program instruction segments capable of performing specific functions, and these instruction segments are used to describe the execution process of the computer program in the terminal device.

[0213] The memory is an internal storage unit of the terminal device, such as the hard disk or memory of the terminal device. The memory can also be an external storage device of the terminal device, such as a plug-in hard disk equipped on the terminal device, a Smart Media Card (SMC), a Secure Digital (SD) card, a FlashCard, etc. Further, the memory can also include both the internal storage unit of the terminal device and the external storage device. The memory is used to store the computer program and other programs and data required by the terminal device. The memory can also be used to temporarily store the data that has been output or will be output.

[0214] Embodiment 4

[0215] This embodiment provides a computer-readable storage medium storing instructions, which when run on a computer, cause the computer to execute the pretreatment method for rapidly extracting lead and cadmium in a grain sample in Embodiment 1.

[0216] The formulas involved above are all calculated by removing the dimension and taking their numerical values. It is a formula obtained by software simulation of a large amount of collected data to be closest to the actual situation. The weight factors in the formula and each preset threshold in the analysis process are set by those skilled in the art according to the actual situation or obtained by simulation of a large amount of data; the size of the weight factor is a specific numerical value obtained by quantifying each parameter for subsequent comparison. Regarding the size of the weight factor, it depends on the amount of sample data and the processing coefficients initially set by those skilled in the art for each group of sample data; as long as the proportional relationship between the parameters and the quantified values is not affected.

[0217] As mentioned above, only the specific embodiments of the present application are described, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present application, and all should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0218] Finally: The above are only the preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A pretreatment method for rapidly extracting lead and cadmium from food samples, characterized in that: include: Step 1: Perform a preliminary classification of N crushed grain samples to obtain R sample types; Step 2: extract the crushed grain sample corresponding to the r-th sample type to obtain the r-th supernatant; and predict the evaluation index of the r-th supernatant in the future time period; r=1，2，……，R； The prediction method of the rth supernatant evaluation index in the future time period includes: Step d1: obtaining supernatant characterization data of the rth supernatant in the current time period, wherein the supernatant characterization data includes pH value, volume, initial metal ion concentration, reaction rate constant and diffusion rate of the supernatant; Step d2: inputting the supernatant characterization data of the rth supernatant in the current time period into the pre-built coefficient prediction model to obtain the rth supernatant evaluation index in the future time period; Step 3: Based on the rth supernatant evaluation index, determine whether the supernatant meets the chelation index; if it does, generate a chelation instruction and jump to step 4; if it does not, set r=r+1 and return to step 2; The method for judging whether the supernatant meets the chelation index based on the rth supernatant evaluation index includes: Step e1: Extract time series data and the corresponding supernatant evaluation index: ; Step e2: Smoothing the supernatant evaluation index corresponding to the time series data to remove noise and outliers; Step e3: Calculate the change amplitude of the smoothed sequence , compare the change amplitude with the preset change amplitude threshold Comparative analysis is performed to generate compliance with the Chelation Indicator Directive, Step 4: accepting the chelation instruction, performing neutralization and chelation reaction on the rth supernatant, obtaining the rth chelated sample, and obtaining the reaction characteristic data of the rth chelated sample; The reaction characteristic data include the formation rate of metal chelates, optical density difference and conductivity difference; Step 5: Input the reaction characteristic data into the pre-built chelation analysis model to obtain the chelation evaluation coefficient of the rth chelation sample in the future time period; The chelation analysis model is an RNN model, a support vector machine regression network model, a linear regression network model or a random forest regression network model; Step 6: Determine whether the rth chelation sample meets the requirements according to the chelation evaluation coefficient output by the chelation analysis model. If it does, store the chelation sample. If it does not meet the requirements, set r=r+1 and return to step 2.

2. A pretreatment method for rapid extraction of lead and cadmium from food samples according to claim 1, characterized in that: The methods for obtaining R sample types include: Step a1: obtaining the homogeneity evaluation coefficient of the nth crushed grain sample; Step a2: setting N homogeneity evaluation coefficient intervals, and setting N sample types corresponding to the N homogeneity evaluation coefficient intervals; each of the homogeneity evaluation coefficient intervals is bound if and only if there is one sample type associated with it; Step a3: comparing the homogeneity evaluation coefficient of the nth crushed grain sample with each homogeneity evaluation coefficient interval, and obtaining the homogeneity evaluation coefficient interval into which the homogeneity evaluation coefficient of the nth crushed grain sample falls; Step a4: according to the homogeneity evaluation coefficient interval into which the homogeneity evaluation coefficient of the nth crushed grain sample falls, classify the nth crushed grain sample into the corresponding sample type, set n=n+1, and jump back to step a1; Step a5: repeat the above steps a1 to a4 until the cycle ends when n=N, so that each of the crushed grain samples is divided into R sample types in turn.

3. A pretreatment method for rapid extraction of lead and cadmium from food samples according to claim 2, characterized in that: The method for obtaining the homogeneity evaluation coefficient of the nth crushed grain sample includes: Step b1: using a grinder to grind the grain sample to a target particle size to obtain a grinded grain sample; Step b2: obtaining basic grain data of the crushed grain sample, wherein the basic grain data includes grain hardness, water content and particle size; Step b3: Formulate and calculate the basic data of grains to obtain the homogeneity assessment coefficient.

4. A pretreatment method for rapid extraction of lead and cadmium from food samples according to claim 3, characterized in that: The method of obtaining the rth supernatant comprises: Step c1: weighing the crushed grain sample corresponding to the rth sample type; Step c2: obtaining the amount of the extract according to the ratio formula, and adding the extract to the crushed grain sample to obtain a first mixture; Among them, the calculation method of the ratio formula is: ; In the formula, Indicates the amount of extract used. Indicates the weight of the crushed grain sample, represents the solid ratio; Step c3: transferring the first mixture to an oscillator with preset oscillation parameters for oscillation to obtain a second mixture; Step c4: placing the second mixture in a centrifuge with preset centrifugal parameters for separation; Step c5: extract the transparent liquid of the separated second mixture through a pipette to obtain the rth supernatant.

5. The pretreatment method for rapidly extracting lead and cadmium from food samples according to claim 1, characterized in that: The training method of the coefficient prediction model includes: Acquire historical coefficient training data, and divide the historical coefficient training data into a coefficient training set and a coefficient test set; the historical coefficient training data includes supernatant characterization data and its corresponding supernatant evaluation index; The method for obtaining the supernatant evaluation index in the historical coefficient training data includes: Step d01: obtaining supernatant characterization data of the rth supernatant within a set time span; Step d02: Formulate and calculate the supernatant characterization data to obtain the supernatant evaluation index, and the calculation formula is as follows: ； In the formula, represents the supernatant evaluation index, Indicates pH value, Indicates volume, Indicates the initial metal ion concentration, is the reaction rate constant, represents the diffusion rate; , , , and is the weight factor, , It represents the logarithmic function with the natural constant e as the base; Constructing a regression network model, taking the supernatant characterization data in the coefficient training set as input data of the regression network model, taking the supernatant evaluation index in the coefficient training set as output data of the regression network model, training the regression network model, and obtaining an initial coefficient prediction regression network; The initial coefficient prediction regression network is model verified using the coefficient test set, and the initial coefficient prediction regression network with a prediction error less than or equal to a preset error threshold is output as the coefficient prediction model; the regression network model is specifically an RNN recurrent neural network model.

6. A pretreatment method for rapid extraction of lead and cadmium from food samples according to claim 5, characterized in that: Compare the change magnitude with the preset change magnitude threshold Methods for conducting comparative analysis to generate compliance indicators for the Chelation Directive include: Compare the change magnitude with the preset change magnitude threshold Make a comparison; like , it means that the evaluation index of the rth supernatant tends to be stable; like , it means that the evaluation index of the rth supernatant is unstable; Step e5: Compare the evaluation index of the rth supernatant that tends to be stable with a preset evaluation index threshold; the evaluation index threshold includes and ,in > ; Compare the supernatant evaluation index with the preset evaluation index threshold; like > or , then no chelation instruction is generated; like , then a chelation instruction is generated.

7. A pretreatment method for rapidly extracting lead and cadmium from food samples according to claim 6, characterized in that: The method for obtaining the rth chelated sample includes: Step f1: gradually adding an alkaline solution to the rth supernatant to adjust the pH value of the solution to a target value, wherein the alkaline solution is NaOH or Na2CO3, and the target value is 6-8; Step f2: dissolving EDTA in distilled water, and slowly adding it to the neutralized supernatant to carry out a chelating reaction, wherein the EDTA is EDTA-2Na, and the volume of the EDTA is 1 / 5 to 1 / 2 of the volume of the supernatant; Step f3: setting chelation parameters and environmental parameters, wherein the chelation parameters include stirring speed and stirring time, and the environmental parameters include temperature control and light intensity; Step f4: The solution finally obtained is the rth chelated sample.

8. A pretreatment method for rapidly extracting lead and cadmium from food samples according to claim 7, characterized in that: The training method of the chelation analysis model comprises: Acquiring historical chelation training data, wherein the historical chelation training data includes multiple groups of reaction characteristic data within a time span and corresponding chelation evaluation coefficients; The logic for obtaining the chelation evaluation coefficient in the historical chelation training data is as follows: The reaction characteristic data in the historical chelation training data are extracted, and the reaction characteristic data are calculated by formula to obtain the chelation evaluation coefficient; the calculation formula is: ； In the formula, represents the chelation evaluation coefficient of the rth chelation sample in the future time period, represents the formation rate of metal chelates, Represents the optical density difference, represents the conductivity difference, , , are weight factors, ; The historical chelation training data is divided into a chelation training set and a chelation test set, and a regression network model is constructed. The reaction characteristic data in the chelation training set is used as the input of the regression network model, and the chelation evaluation coefficient in the chelation training set is used as the output of the regression network model. The regression network model is trained to obtain an initial regression network. The training goal is to minimize the sum of prediction accuracies. The initial regression network is evaluated using the chelation test set, and the initial regression network when the sum of prediction accuracies reaches convergence is used as the pre-constructed chelation analysis model.

9. A pretreatment method for rapid extraction of lead and cadmium from food samples according to claim 8, characterized in that: Methods for judging whether the rth chelation sample meets the requirements according to the chelation evaluation coefficient include: Preset chelation coefficient threshold, compare the chelation evaluation coefficient with the preset chelation coefficient threshold For comparison; if , then the chelated sample is judged to not meet the requirements; like , then the chelated sample is judged to meet the requirements.

10. A pretreatment system for rapidly extracting lead and cadmium from food samples, used for implementing a pretreatment method for rapidly extracting lead and cadmium from food samples according to any one of claims 1 to 9, characterized in that: include: A classification module is used to perform preliminary classification on N crushed grain samples to obtain R sample types; The first prediction module is used to extract the crushed grain sample corresponding to the r-th sample type to obtain the r-th supernatant; and predict the r-th supernatant evaluation index in the future time period; r=1, 2, ..., R; The first judgment module judges whether the supernatant meets the chelation index based on the rth supernatant evaluation index; if it meets the index, a chelation instruction is generated and the chelation module is jumped; if it does not meet the index, r=r+1 is set and the first prediction module is returned; A chelation module, used for receiving a chelation instruction, performing a neutralization and chelation reaction on the rth supernatant, obtaining an rth chelated sample, and acquiring reaction characteristic data of the rth chelated sample; an analysis module for inputting the reaction characteristic data into a pre-built chelation analysis model to obtain a chelation evaluation coefficient for the rth chelation sample in a future time period; The second judgment module is used to judge whether the rth chelation sample meets the requirements according to the chelation evaluation coefficient output by the chelation analysis model. If yes, the chelation sample is stored; if not, r=r+1 is set and the result is returned to the first prediction module.

Images