Method for detecting trace amount of aflatoxin in wheat

Abstract

The invention relates to the technical field of wheat aflatoxin trace detection, and discloses a wheat aflatoxin trace detection method, which comprises the following steps: carrying out a storage experiment of wheat polluted by inoculated aflatoxin-producing fungus strains, researching the change rule of volatile substance components of wheat under different aflatoxin pollution levels, and determining the content of aflatoxin in wheat. The invention relates to controllable synthesis and color development mechanism research of a magnetic nano color-sensitive material. According to the invention, wheat indicative odorant molecules polluted by aflatoxin can be efficiently captured through the created magnetic nano color-sensitive sensor, and a theoretical basis is provided for developing a special color-sensitive sensor system for rapidly detecting grain mycotoxin on site; the novel controllable synthesis and self-assembly method of the magnetic nano color-sensitive material based on the MMOFs framework can realize the specific color development of the magnetic nano color-sensitive material on the wheat AFB1 indicative odorant molecules, and can ensure the stability of sensor signals.

Description

Technical Field

[0001] The invention relates to the technical field of wheat aflatoxin trace detection, in particular to a wheat aflatoxin trace detection method. Background Art

[0002] Wheat is the third largest grain crop, and China is the world's largest wheat producer [1]. Grains such as wheat go through a long storage and transportation process from harvest to processing. During the storage and transportation process, grains are very likely to become moldy due to environmental factors. Among them, Aspergillus flavus is a common saprophytic fungus that is often found on moldy grains, grain products, and other moldy organic matter. Its toxin contamination is a key issue of concern in the food and agricultural products industry. Wheat and other grain crops are easily contaminated by Aspergillus flavus when stored under hot and humid conditions, producing aflatoxin.

[0003] However, the properties of mycotoxins are relatively stable and are generally not removed during subsequent processing, which poses a serious threat to human and animal health [2-4]. Aflatoxin secreted by Aspergillus flavus consists of 15-20 compounds, among which aflatoxin B1 (abbreviated as AFB1) is the most harmful and toxic. It is listed as the number one carcinogen in humans by the International Agency for Research on Cancer and is one of the main factors inducing malignant tumors such as primary hepatocellular carcinoma [5]. In recent years, due to global warming, the extent and scope of aflatoxin damage to grains such as wheat have been on the rise [6].

[0004] Therefore, monitoring and preventing fungal toxin contamination in grain has become a top priority in grain work. Developing rapid, efficient, non-destructive, and trace on-site detection technology for grain fungal toxin contamination levels in order to achieve early detection and early prevention is of great significance to ensuring food security.

[0005] At present, in daily life, the identification of grain mold contamination mainly relies on human sense of smell, vision and other sensory means. In the laboratory, the common methods for detecting the aflatoxin contamination level of grain include thin layer chromatography (TLC), enzyme-linked immunosorbent assay (ELISA), liquid chromatography-tandem mass spectrometry (LC-MS / MS), high performance liquid chromatography (HPLC), etc. [7-9]. Among the above methods, the identification results of artificial sensory means are usually delayed. If the grain is detected to be moldy by artificial sensory means, the grain has often suffered irreversible damage, thus losing the significance of grain mold detection, and it is too empirical; although the TLC and ELISA determination methods are simple to operate, they are prone to false positives and have poor quantitative effects, and are not suitable for final confirmation; although the LC-MS / MS and HPLC determination methods have high detection accuracy and low detection limit, they require expensive liquid chromatography-tandem mass spectrometers and high performance liquid chromatography, the sample pretreatment process is cumbersome, the detection cost is high, the efficiency is low, and the requirements for test personnel are high, and the application of on-site detection is subject to certain restrictions. Therefore, conventional methods for detecting aflatoxin in grain cannot meet the needs of rapid and non-destructive modern food testing and are not suitable for application scenarios of on-site in-situ testing under grain storage conditions. Summary of the invention

[0006] Technical issues solved

[0007] In view of the shortcomings of the prior art, the present invention provides a method for detecting trace amounts of aflatoxin in wheat, which is mainly used to solve the problem that monitoring and preventing fungal toxin contamination in grain has become a top priority in grain work, and to develop a rapid, efficient, non-destructive, trace on-site detection technology for grain fungal toxin contamination levels to achieve the purpose of early detection and early prevention, which is of great significance for ensuring food security, and the conventional detection methods for aflatoxin in grain cannot meet the needs of rapid and non-destructive modern food testing, and are not suitable for application scenarios of on-site in-situ detection under grain storage conditions.

[0008] Technical Solution

[0009] To achieve the above object, the present invention provides the following technical solutions:

[0010] A method for detecting trace amounts of aflatoxin in wheat comprises the following steps:

[0011] S1: Conduct a storage experiment of wheat inoculated with aflatoxin-producing fungal strains: Wheat inoculated with different doses of aflatoxin-producing fungal strains were packed into different fresh-keeping bags and placed in different intelligent artificial climate incubators for temperature-controlled storage until visible mold spots appeared. The storage experiment was terminated; AFB1 physical and chemical tests were performed on the wheat samples according to the method specified in GB 5009.22, and the data were recorded;

[0012] S2: Study on the changing rules of volatile components of wheat under different aflatoxin contamination levels: For the wheat samples obtained from the S1 storage experiment, the project intends to use HS-SPME technology combined with GC-MS technology to study the changing rules of volatile components of wheat samples under different aflatoxin contamination levels;

[0013] S3: Controllable synthesis and color development mechanism of magnetic nano-color-sensitive materials: It is planned to try to achieve the controllable synthesis of nano-color-sensitive materials based on magnetic metal-organic frameworks (MMOFs) by evaporation solvent method, diffusion method, hydrothermal or solvothermal method, ultrasound and microwave method; use UV-visible spectroscopy, fluorescence and other methods to characterize the spectral changes of the molecular aggregation of magnetic nano-color-sensitive materials, so as to obtain magnetic nano-color-sensitive materials with strong adsorption, porosity, unique spatial configuration, large specific surface area and strong coordination ability; study the coordination reaction between wheat AFB1 indicator odor molecules and magnetic nano-color-sensitive materials under different aflatoxin contamination levels;

[0014] S4: Creation of magnetic nano colorimetric sensor and analysis of specific color development mechanism: The prepared different magnetic nano colorimetric materials were dissolved in DMAC, and the above solutions were respectively transferred with a spotting capillary, and adsorbed on the hydrophobic reaction substrate material polyvinylidene fluoride (PVDF) film in sequence in the fume hood. After the solvent was completely evaporated, the magnetic nano colorimetric sensor array was obtained; the medium substrate and the magnetic nano colorimetric material were characterized before and after the combination using scanning electron microscopy (SEM) and transmission electron microscopy (TEM), and on this basis, the magnetic nano colorimetric material on the sensor array was reconstructed in a targeted manner to improve the color development specificity of the magnetic nano colorimetric material; the morphology, size, and surface properties of the magnetic nano colorimetric material were regulated and optimized by adjusting the emulsifier concentration, type, and emulsification method, and the binding force between the sensor medium substrate and the magnetic nano colorimetric material was enhanced; finally, the sensor response characteristics of wheat indicator odor molecules under different aflatoxin contamination levels were analyzed, the response mechanism of the sensor to the indicator odor molecules was explored, and the non-destructive detection mechanism of the magnetic nano colorimetric sensor for stored wheat AFB1 was clarified;

[0015] S5: Data collection and preprocessing of wheat volatile components by magnetic nano color sensors: The collected wheat samples were placed in a closed sensor chamber with a ring-shaped LED light source inside to ensure uniform illumination on the sensor array surface; the magnetic nano color sensor was placed on the top of the sensor chamber and left to stand for five minutes to obtain the odor information of the wheat volatile components and present it in the form of an image;

[0016] S6: Research on deep feature extraction of magnetic nano color sensor images: Based on the existing convolutional neural network (CNN) model structure, the new CNN structure is redesigned according to their effect of deep feature learning on preprocessed color sensor images. The multi-objective group intelligent optimization algorithm (MOPSO, MOACO, MOABC, etc.) is used to optimize the parameters such as width and depth of the new CNN, and the new CNN training is realized in a supervised learning manner; the optimal model is determined by combining model accuracy and inference real-time performance to realize deep extraction of magnetic nano color sensor image features; finally, the image features are abstracted layer by layer by sharing weight parameters, and the top-level output of the model is used as the extracted image deep features;

[0017] S7: Construction and optimization of wheat AFB1 nonlinear detection model: According to the deep features of the magnetic nano-color sensor image extracted by S6, a stacked extreme learning machine autoencoder network (SELM-AEN) is proposed to construct a nonlinear detection model for stored wheat AFB1; the established model is verified by the national standard test results, and the SELM-AEN nonlinear detection model is further optimized according to the verification feedback results, and finally a nonlinear detection model based on deep learning theory with high detection accuracy and strong fault tolerance is obtained.

[0018] As a further solution of the present invention, the storage temperature in S1 is controlled at 12°C, 20°C, 28°C, 36°C and 44°C respectively; and sampling is performed at intervals of 3-5 days.

[0019] Furthermore, S2 introduces a swarm intelligence optimization algorithm (PSO, ACO, ABC, etc.) to search for indicative odor molecules related to AFB1, establishes a general mathematical model between indicative odor molecules and AFB1 content, clarifies the change mechanism of indicative odor molecules under different aflatoxin contamination levels, and lays a theoretical foundation for on-site non-destructive and trace detection of wheat AFB1 by magnetic nano color sensitive sensors.

[0020] On the basis of the above scheme, S3 intends to use the spectral dynamics method to analyze the law of spectral redshift or blueshift in the coordination reaction between the two, and intends to use density functional theory (DFT) to reveal the change law of electron cloud density and spatial structure in the coordination reaction, so as to clarify the color display mechanism of magnetic nano-color-sensitive materials.

[0021] As a further solution of the present invention, S4 intends to use X-ray electron spectroscopy analysis (XPS) to study the interaction between molecules of magnetic nano-color-sensitive materials, and on this basis, to carry out targeted modification of the surface of the magnetic nano-color-sensitive materials to increase the contact area between the magnetic nano-color-sensitive materials and volatile substances, thereby improving the sensitivity and stability of the magnetic nano-color-sensitive sensors.

[0022] Furthermore, in S5, a high-performance CCD camera is used to obtain the sensor array image data before and after the reaction. The image processing software is used to perform filtering, threshold segmentation and other pre-processing on the images of the sensor array before and after the reaction. Then, a circle with the center of each color-sensitive point on the sensor as the origin and a radius of 15 pixels is constructed as a template, the grayscale value of each template is calculated, and the grayscale values ​​before and after the reaction are subtracted to obtain the difference image before and after the reaction. After normalization, the GRB color image of each sample can be obtained for subsequent analysis.

[0023] On the basis of the above scheme, in the process of constructing the nonlinear detection model of S7, the model structure of SELM-AEN is first designed, and then multi-objective group intelligent optimization algorithms such as MOPSO, MOACO, and MOABC are introduced to optimize the network depth and the number of neurons in each hidden layer of SELM-AEN to determine the optimal network structure and parameters, and establish a high-precision SELM-AEN nonlinear detection model to realize trace detection of wheat AFB1.

[0024] Beneficial Effects

[0025] Compared with the prior art, the present invention provides a method for detecting trace amounts of aflatoxin in wheat, which has the following beneficial effects:

[0026] 1. The present invention proposes a new idea for on-site trace detection of wheat AFB1 based on the depth characteristics of magnetic nano-color sensitive sensor images. The invented magnetic nano-color sensitive sensor can efficiently capture the indicative odor molecules of wheat contaminated with aflatoxin, providing a theoretical basis for the development of a special color sensitive sensor system for on-site rapid detection of grain mycotoxins.

[0027] 2. The present invention proposes a new method for controllable synthesis and self-assembly of magnetic nano-color-sensitive materials based on the MMOFs framework, which can realize the specific color development of magnetic nano-color-sensitive materials for wheat AFB1 indicative odor molecules and ensure the stability of sensor signals, thereby improving the credibility of detection results.

[0028] 3. The present invention introduces some new theories, technologies and methods in the fields of chemical science, nanoscience and information science into the controllable synthesis and self-assembly of special materials for food fungal toxin detection sensors, as well as the deep feature extraction of sensor images and the construction of nonlinear detection models, which enhances the theoretical depth of the research of this project and promotes multidisciplinary cross-development. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a schematic diagram of the process of a method for detecting trace amounts of aflatoxin in wheat proposed by the present invention;. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below through embodiments and in conjunction with the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0031] The serial numbers assigned to the components herein, such as "first", "second", etc., are only used to distinguish the objects described and do not have any order or technical meaning. The "connection" and "coupling" mentioned in the present invention, unless otherwise specified, include direct and indirect connections (couplings). In the description of the present invention, it should be understood that the orientation or position relationship indicated by the terms "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", etc. are based on the orientation or position relationship shown in the accompanying drawings, which is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation to the present invention.

[0032] In the present invention, unless otherwise clearly specified and limited, a first feature being "above" or "below" a second feature may mean that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. Moreover, a first feature being "above", "above" or "above" a second feature may mean that the first feature is directly above or obliquely above the second feature, or simply means that the first feature is higher in level than the second feature. A first feature being "below", "below" or "below" a second feature may mean that the first feature is directly below or obliquely below the second feature, or simply means that the first feature is lower in level than the second feature.

[0033] Embodiment 1:

[0034] Wheat infected with mold will gradually volatilize various specific organic gases. The detection and analysis of volatile substances to determine the moldy condition is an important way to detect wheat mold infection. It is planned to use HS-SPME technology combined with GC-MS technology to study the dynamic changes of wheat volatile substances under different aflatoxin contamination levels, and use swarm intelligence optimization algorithm to search for indicative odor molecules closely related to AFB1; study the controllable synthesis and self-assembly methods of magnetic nano-color-sensitive materials, and use quantum chemistry and spectral dynamics to reveal the coordination reaction between magnetic nano-color-sensitive materials and indicative odor molecules, clarify the specific color development mechanism, and construct a highly sensitive magnetic nano-color-sensitive sensor array; based on this, further clarify the correlation between indicative odor molecules and sensor array image feature changes; based on deep learning theory, establish a wheat AFB1 detection model of magnetic nano-color-sensitive sensor image depth features to achieve on-site rapid trace detection of wheat AFB1, thus making up for the current shortcomings in the on-site rapid detection of grain mycotoxins.

[0035] The GC-MS technique was used to analyze the changes in the composition of volatile substances in wheat under different aflatoxin contamination levels. Preliminary research results found that the four volatile substances that changed more significantly were benzaldehyde, nonanal, 1-octen-3-ol and 2-methylbutanal. Based on the indicative odor molecules obtained by the above GC-MS analysis, the project team attempted to use the evaporation solvent method and diffusion method to preliminarily synthesize magnetic nano-color-sensitive materials through quantum chemical calculations, and prepared corresponding color-sensitive sensors to image and express aldehydes, alcohols and ketones of different contents, all of which achieved good color development effects.

[0036] Reference Figure 1 , a method for detecting trace amounts of aflatoxin in wheat, comprising the following steps:

[0037] S1: Conduct a storage experiment of wheat inoculated with aflatoxin-producing fungal strains: Wheat inoculated with different doses of aflatoxin-producing fungal strains were packed into different fresh-keeping bags and placed in different intelligent artificial climate incubators for temperature-controlled storage, with the temperatures controlled at 12°C, 20°C, 28°C, 36°C and 44°C respectively; samples were taken at intervals of 3-5 days until visible mold spots appeared, and the storage experiment was terminated; AFB1 physical and chemical tests were performed on the sampled wheat samples according to the method specified in GB 5009.22, and the data were recorded;

[0038] S2: Study on the changing rules of volatile components of wheat under different aflatoxin contamination levels: For the wheat samples obtained from the S1 storage experiment, the project intends to use HS-SPME technology combined with GC-MS technology to study the changing rules of volatile components of wheat samples under different aflatoxin contamination levels; introduce swarm intelligence optimization algorithms (PSO, ACO, ABC, etc.) to search for indicative odor molecules related to AFB1, establish a general mathematical model between indicative odor molecules and AFB1 content, clarify the changing mechanism of indicative odor molecules under different aflatoxin contamination levels, and lay a theoretical foundation for the on-site non-destructive and trace detection of wheat AFB1 by magnetic nano color sensitive sensors;

[0039] S3: Controllable synthesis and color development mechanism of magnetic nano-color-sensitive materials: It is planned to try to achieve the controllable synthesis of nano-color-sensitive materials based on magnetic metal-organic frameworks (MMOFs) by means of evaporation solvent method, diffusion method, hydrothermal or solvothermal method, ultrasound and microwave method; use UV-visible spectroscopy, fluorescence and other methods to characterize the spectral changes of the aggregation of magnetic nano-color-sensitive materials molecules, so as to obtain magnetic nano-color-sensitive materials with strong adsorption, porosity, unique spatial configuration, large specific surface area and strong coordination ability; study the coordination reaction between wheat AFB1 indicator odor molecules and magnetic nano-color-sensitive materials under different aflatoxin contamination levels, and use the spectral dynamics method to analyze the spectral red shift or blue shift law during the coordination reaction between the two, and use density functional theory (DFT) to reveal the change law of electron cloud density and spatial structure in the coordination reaction, and clarify the color development mechanism of magnetic nano-color-sensitive materials;

[0040] S4: Creation of magnetic nano-colorimetric sensor and analysis of specific color development mechanism: The prepared different magnetic nano-colorimetric materials were dissolved in DMAC, and the above solutions were respectively transferred with a spotting capillary, and adsorbed on the hydrophobic reaction substrate material polyvinylidene fluoride (PVDF) film in sequence in a fume hood. After the solvent was completely evaporated, the magnetic nano-colorimetric sensor array was obtained; the medium substrate and the magnetic nano-colorimetric material were characterized before and after the combination using a scanning electron microscope (SEM) and a transmission electron microscope (TEM). On this basis, the magnetic nano-colorimetric material on the sensor array was specifically reconstructed to enhance the specificity of the color development of the magnetic nano-colorimetric material; by adjusting the concentration, type and emulsification method of the emulsifier, the magnetic nano-colorimetric material on the sensor array was reconstructed to enhance the specificity of the color development of the magnetic nano-colorimetric material. , regulate and optimize the morphology, size, and surface properties of magnetic nano-color-sensitive materials, and enhance the binding force between the sensor medium substrate and the magnetic nano-color-sensitive materials; X-ray electron spectroscopy analysis (XPS) is proposed to study the interaction between the molecules of magnetic nano-color-sensitive materials, and on this basis, the surface of magnetic nano-color-sensitive materials is modified in a targeted manner to increase the contact area between magnetic nano-color-sensitive materials and volatile substances, thereby improving the sensitivity and stability of magnetic nano-color-sensitive sensors; finally, the sensor response characteristics of wheat indicator odor molecules under different aflatoxin contamination levels are analyzed, the response mechanism of the sensor to the indicator odor molecules is explored, and the non-destructive detection mechanism of magnetic nano-color-sensitive sensors for stored wheat AFB1 is clarified;

[0041] S5: Data acquisition and preprocessing of volatile components of wheat by magnetic nano color sensors: The collected wheat samples were placed in a closed sensor chamber with a ring-shaped LED light source inside to ensure uniform illumination on the surface of the sensor array; the magnetic nano color sensors were placed on the top of the sensor chamber and left to stand for five minutes to obtain the odor information of the volatile components of wheat and present it in the form of images; the image data of the sensor array before and after the reaction were obtained by a high-performance CCD camera, and the images of the sensor array before and after the reaction were filtered and threshold segmented by image processing software, and then a circle with the center of each color-sensitive point on the sensor as the origin and a radius of 15 pixels was constructed as a template, the gray value of each template was calculated, and the gray values ​​before and after the reaction were subtracted to obtain the difference image before and after the reaction, and the GRB color image of each sample was obtained after normalization for subsequent analysis;

[0042] S6: Research on deep feature extraction of magnetic nano color sensor images: Based on the existing convolutional neural network (CNN) model structure, the new CNN structure is redesigned according to their effect of deep feature learning on preprocessed color sensor images. The multi-objective group intelligent optimization algorithm (MOPSO, MOACO, MOABC, etc.) is used to optimize the parameters such as width and depth of the new CNN, and the new CNN training is realized in a supervised learning manner; the optimal model is determined by combining model accuracy and inference real-time performance to realize deep extraction of magnetic nano color sensor image features; finally, the image features are abstracted layer by layer by sharing weight parameters, and the top-level output of the model is used as the extracted image deep features;

[0043] S7: Construction and optimization of wheat AFB1 nonlinear detection model: According to the deep features of the magnetic nano-colorimetric sensor image extracted by S6, a stacked extreme learning machine autoencoder network (SELM-AEN) is proposed to construct a nonlinear detection model for stored wheat AFB1. In the process of model construction, the model structure of SELM-AEN is first designed, and then multi-objective group intelligent optimization algorithms such as MOPSO, MOACO, and MOABC are introduced to optimize the network depth and the number of neurons in each hidden layer of SELM-AEN to determine the optimal network structure and parameters, and establish a high-precision SELM-AEN nonlinear detection model to realize trace detection of wheat AFB1. The established model is verified by the national standard test results, and the SELM-AEN nonlinear detection model is further optimized according to the verification feedback results, and finally a nonlinear detection model based on deep learning theory with high detection accuracy and strong fault tolerance is obtained.

[0044] The trace detection method of wheat aflatoxin based on the deep features of color-sensitive sensor images has the following beneficial effects: a new idea for on-site trace detection of wheat AFB1 based on the deep features of magnetic nano-color-sensitive sensor images is proposed. The created magnetic nano-color-sensitive sensor can efficiently capture the indicator odor molecules of wheat contaminated with aflatoxin, providing a theoretical basis for the development of a special color-sensitive sensor system for rapid on-site detection of grain mycotoxins; a new method for the controllable synthesis and self-assembly of magnetic nano-color-sensitive materials based on the MMOFs framework is proposed, which can achieve the specific color development of magnetic nano-color-sensitive materials for wheat AFB1 indicator odor molecules and ensure the stability of sensor signals, thereby improving the credibility of detection results; some new theories, new technologies and new methods in the fields of chemical science, nanoscience and information science are introduced into the controllable synthesis and self-assembly of special materials for grain mycotoxin detection sensors, as well as the extraction of sensor image deep features and the construction of nonlinear detection models, which enhances the theoretical depth of the research of this project and promotes multidisciplinary cross-development.

[0045] Based on the change law of the volatile substance composition of wheat under different storage time and storage conditions using HS-SPME combined with GC-MS technology, the indicative odor molecule of wheat AFB1 was identified, and a general mathematical model between the indicative odor molecule and the content of wheat AFB1 was established; the controllable synthesis method of new magnetic nano-color-sensitive materials based on MMOFs was studied, and magnetic nano-color-sensitive sensors with excellent specificity and stability were prepared. The change law of the sensor response characteristics of the indicator odor molecule of wheat AFB1 under different aflatoxin contamination levels was analyzed, the response mechanism of the sensor to the indicator odor molecules was explored, and the non-destructive detection mechanism of the magnetic nano-color-sensitive sensor for stored wheat AFB1 was clarified; based on deep learning theory, the network structure was studied, and the optimal parameters of the neural network were determined using a multi-objective swarm intelligent optimization algorithm to achieve efficient extraction of deep features of sensor images and construction of a high-precision nonlinear detection model, ultimately achieving on-site rapid, non-destructive, and trace detection of stored wheat AFB1.

[0046] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0047] The above-described embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the present invention. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the attached claims.

Claims

1. A method for detecting trace amounts of aflatoxin in wheat, characterized in that: It includes the following steps: S1: Conduct a storage experiment of wheat inoculated with aflatoxin-producing fungal strains: Wheat inoculated with different doses of aflatoxin-producing fungal strains were packed into different fresh-keeping bags and placed in different intelligent artificial climate incubators for temperature-controlled storage until visible mold spots appeared. The storage experiment was terminated; AFB1 physical and chemical tests were performed on the wheat samples according to the method specified in GB 5009.22, and the data were recorded; S2: Study on the changing rules of volatile components of wheat under different aflatoxin contamination levels: For the wheat samples obtained from the S1 storage experiment, the project intends to use HS-SPME technology combined with GC-MS technology to study the changing rules of volatile components of wheat samples under different aflatoxin contamination levels; S3: Controllable synthesis and color development mechanism of magnetic nano-color-sensitive materials: It is planned to try to achieve the controllable synthesis of nano-color-sensitive materials based on magnetic metal-organic frameworks (MMOFs) by evaporation solvent method, diffusion method, hydrothermal or solvothermal method, ultrasound and microwave method; use UV-visible spectroscopy, fluorescence and other methods to characterize the spectral changes of the molecular aggregation of magnetic nano-color-sensitive materials, so as to obtain magnetic nano-color-sensitive materials with strong adsorption, porosity, unique spatial configuration, large specific surface area and strong coordination ability; study the coordination reaction between wheat AFB1 indicator odor molecules and magnetic nano-color-sensitive materials under different aflatoxin contamination levels; S4: Creation of magnetic nano colorimetric sensor and analysis of specific color development mechanism: The prepared different magnetic nano colorimetric materials were dissolved in DMAC, and the above solutions were respectively transferred with a spotting capillary, and adsorbed on the hydrophobic reaction substrate material polyvinylidene fluoride (PVDF) film in sequence in the fume hood. After the solvent was completely evaporated, the magnetic nano colorimetric sensor array was obtained; the medium substrate and the magnetic nano colorimetric material were characterized before and after the combination using scanning electron microscopy (SEM) and transmission electron microscopy (TEM), and on this basis, the magnetic nano colorimetric material on the sensor array was reconstructed in a targeted manner to improve the color development specificity of the magnetic nano colorimetric material; the morphology, size, and surface properties of the magnetic nano colorimetric material were regulated and optimized by adjusting the emulsifier concentration, type, and emulsification method, and the binding force between the sensor medium substrate and the magnetic nano colorimetric material was enhanced; finally, the sensor response characteristics of wheat indicator odor molecules under different aflatoxin contamination levels were analyzed, the response mechanism of the sensor to the indicator odor molecules was explored, and the non-destructive detection mechanism of the magnetic nano colorimetric sensor for stored wheat AFB1 was clarified; S5: Data collection and preprocessing of wheat volatile components by magnetic nano color sensors: The collected wheat samples were placed in a closed sensor chamber with a ring-shaped LED light source inside to ensure uniform illumination on the sensor array surface; the magnetic nano color sensor was placed on the top of the sensor chamber and left to stand for five minutes to obtain the odor information of the wheat volatile components and present it in the form of an image; S6: Research on deep feature extraction of magnetic nano color sensor images: Based on the existing convolutional neural network (CNN) model structure, the new CNN structure is redesigned according to their effect of deep feature learning on preprocessed color sensor images. The multi-objective group intelligent optimization algorithm (MOPSO, MOACO, MOABC, etc.) is used to optimize the parameters such as width and depth of the new CNN, and the new CNN training is realized in a supervised learning manner; the optimal model is determined by combining model accuracy and inference real-time performance to realize deep extraction of magnetic nano color sensor image features; finally, the image features are abstracted layer by layer by sharing weight parameters, and the top-level output of the model is used as the extracted image deep features; S7: Construction and optimization of wheat AFB1 nonlinear detection model: According to the deep features of the magnetic nano-color sensor image extracted by S6, a stacked extreme learning machine autoencoder network (SELM-AEN) is proposed to construct a nonlinear detection model for stored wheat AFB1; the established model is verified by the national standard test results, and the SELM-AEN nonlinear detection model is further optimized according to the verification feedback results, and finally a nonlinear detection model based on deep learning theory with high detection accuracy and strong fault tolerance is obtained.

2. A method for detecting trace amounts of aflatoxin in wheat according to claim 1, characterized in that: The storage temperatures in the S1 were controlled at 12° C., 20° C., 28° C., 36° C. and 44° C., respectively; and sampling was performed at intervals of 3-5 days.

3. A method for detecting trace amounts of aflatoxin in wheat according to claim 1, characterized in that: The S2 introduces a swarm intelligence optimization algorithm (PSO, ACO, ABC, etc.) to search for indicative odor molecules related to AFB1, establishes a general mathematical model between indicative odor molecules and AFB1 content, clarifies the change mechanism of indicative odor molecules under different aflatoxin contamination levels, and lays a theoretical foundation for on-site non-destructive and trace detection of wheat AFB1 by magnetic nano color sensitive sensors.

4. A method for detecting trace amounts of aflatoxin in wheat according to claim 1, characterized in that: In the above S3, the spectral dynamics method is used to analyze the law of spectral red shift or blue shift during the coordination reaction between the two, and the density functional theory (DFT) is used to reveal the change law of electron cloud density and spatial structure in the coordination reaction, so as to clarify the color development mechanism of magnetic nano-color-sensitive materials.

5. A method for detecting trace amounts of aflatoxin in wheat according to claim 1, characterized in that: In S4, X-ray electron spectroscopy (XPS) is used to study the interaction between molecules of magnetic nano-color-sensitive materials, and based on this, the surface of the magnetic nano-color-sensitive materials is modified in a targeted manner to increase the contact area between the magnetic nano-color-sensitive materials and volatile substances, thereby improving the sensitivity and stability of the magnetic nano-color-sensitive sensors.

6. A method for detecting trace amounts of aflatoxin in wheat according to claim 1, characterized in that: In the S5, a high-performance CCD camera is used to obtain the sensor array image data before and after the reaction. The image processing software is used to perform filtering, threshold segmentation and other pre-processing on the images of the sensor array before and after the reaction. Then, a circle with the center of each color-sensitive point on the sensor as the origin and a radius of 15 pixels is constructed as a template, the grayscale value of each template is calculated, and the grayscale values ​​before and after the reaction are subtracted to obtain the difference image before and after the reaction. After normalization, the GRB color image of each sample can be obtained for subsequent analysis.

7. A method for detecting trace amounts of aflatoxin in wheat according to claim 1, characterized in that: In the process of constructing the nonlinear detection model of S7, the model structure of SELM-AEN is first designed, and then multi-objective group intelligent optimization algorithms such as MOPSO, MOACO, and MOABC are introduced to optimize the network depth and the number of neurons in each hidden layer of SELM-AEN to determine the optimal network structure and parameters, and establish a high-precision SELM-AEN nonlinear detection model to realize trace detection of wheat AFB1.

Images