Soybean producing area soil micro-ecological risk assessment method and system based on combined pollution

Abstract

The invention discloses a soybean producing area soil combined pollution micro-ecological risk assessment method and system, and belongs to the technical field of agricultural environment monitoring. The method comprises the following steps: acquiring soil pollutant data and microbial community data, calculating a risk entropy, constructing a micro-ecological health index, and respectively obtaining a first result and a second result by utilizing a risk assessment neural network and a comprehensive model combined with principal component analysis, a clustering algorithm and a hunter optimization algorithm; and determining a final risk level and high-risk pollutants after key factor matching and weighted correction mechanism verification. According to the method, the problems of single evaluation dimension, insufficient dynamic adaptability and the like in the prior art are solved, accurate diagnosis and risk early warning of soil combined pollution are realized through multi-source data fusion and multi-method cooperation, the evaluation accuracy and decision-making efficiency are improved, and technical support is provided for soybean producing area soil micro-ecological risk management and control and agricultural sustainable development.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural environmental monitoring, and more particularly to a method and system for assessing soil microecological risk in soybean producing areas based on complex pollution. Background Art

[0002] Within the OneHealth framework, soil health is directly linked to food security and ecosystem stability. However, the highly intensive development of agriculture and human activities have led to increasingly serious problems with the combined residues of heavy metals, pesticides (herbicides, insecticides, etc.), and emerging pollutants (such as phthalate plasticizers), exacerbating the risk of combined soil contamination. The combined residues of these pollutants in the soil can create complex ecological and environmental effects, significantly disrupting the structure and function of microbial communities, reducing the stability and disease resistance of soil ecosystems, and ultimately threatening crop health, food security, and sustainable agricultural development.

[0003] China is heavily reliant on soybean imports. To overcome this situation, expanding soybean planting and increasing yields on a large scale have been prioritized as key agricultural research tasks. Currently, increasing soybean yields on a large scale is often limited by industrial and technological bottlenecks, such as the frequent occurrence of underground diseases such as root rot. The exacerbation of these diseases is closely related to the "sub-health" of the soil caused by complex soil pollution. However, research on complex agricultural soil pollution is still relatively underdeveloped. Existing studies have mostly focused on a single or a few pollutants, or the toxicological effects and degradation mechanisms of the same type of pollutants, lacking a systematic and comprehensive understanding of the impacts of long-term exposure to multiple pollutants.

[0004] Therefore, how to propose a soil microecological risk assessment method and system for soybean producing areas based on complex pollution, and comprehensively analyze multidimensional pollutants to improve the accuracy of risk assessment and decision-making efficiency, is an urgent problem that technical personnel in this field need to solve. Summary of the Invention

[0005] In view of this, the present invention provides a soil microecological risk assessment method and system for soybean producing areas based on composite pollution, which solves the problems of single assessment dimension, insufficient dynamic adaptability and delayed data processing in the existing technology, and realizes accurate diagnosis and risk warning of composite pollution in soybean producing areas.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] In one aspect, the present invention provides a method for assessing soil microecological risk in soybean production areas based on composite pollution, comprising the following steps:

[0008] Obtain pollutant data and microbial community data of soybean-producing soil;

[0009] Calculating the risk entropy of the pollutant data; constructing a microecological health index based on the microbial community data;

[0010] Constructing a risk assessment neural network, training the risk assessment neural network based on the pollutant data and the microbial community data, and obtaining a first result using the trained risk assessment neural network, the first result including a first risk prediction result and a combination of key pollution factors;

[0011] Combining the risk entropy and the microecological health index to obtain a second result, the second result including a second risk prediction result and a priority control factor;

[0012] The soil microecological risk assessment results are verified based on the first result and the second result to obtain the final soil microecological risk level and output high-risk pollutants at the same time.

[0013] Preferably, the calculation formula of the risk entropy RQ is as follows:

[0014]

[0015] Where MEC is the measured concentration of a pollutant and PNEC is the predicted no-effect concentration of a pollutant. The PNEC value is determined by the ratio between the short-term / long-term toxicity test results and the additional safety factor.

[0016] Preferably, constructing a microecological health index based on the microbial community data includes:

[0017] Calculating microbial health indicators based on the microbial community data, wherein the microbial health indicators include but are not limited to microbial diversity indicators, functional gene abundance, and pathogen load;

[0018] normalizing the microbial health indicator;

[0019] The normalized microbial health indicators are weighted and summed to form a microecological health index.

[0020] Preferably, the second result is obtained by combining the risk entropy and the microecological health index, including:

[0021] Based on principal component analysis, the risk entropy and the microecological health index are integrated to define a pollution load index, a functional damage index, and a microbial diversity attenuation index;

[0022] Using the pollution load index, the functional damage index, and the microbial diversity attenuation index as coordinate axes, a clustering algorithm is used to divide the risk levels to obtain the second risk prediction result;

[0023] Priority control factors are determined based on the clustering results combined with key pollution factors.

[0024] Preferably, in the process of dividing the risk levels by using a clustering algorithm, the weights of the pollution load index, the functional damage index and the microbial diversity attenuation index are adaptively adjusted using a predator optimization algorithm.

[0025] Preferably, a clustering algorithm is used to divide the risk levels to obtain the second risk prediction result, including:

[0026] Initialize the population and generate N random weight vectors W = [w1, w2, w3]; w1 is the pollution load index weight, w2 is the functional damage index weight, and w3 is the microbial diversity decay index weight;

[0027] For each weight vector W, the pollution load index, the functional damage index, and the microbial diversity attenuation index are calculated, a clustering operation is performed, and a fitness value Fit(W) is calculated based on the clustering results. The fitness value calculation formula is as follows:

[0028] Fit(W)=Silhouette Score(W)-Davies-Bouldin Index(W);

[0029] Where Silhouette Score (W) represents the silhouette coefficient, Davies-Bouldin Index (W) represents the Davidson-Bouldin index;

[0030] Iteratively update the positions of predators and prey based on the current fitness value;

[0031] After each iteration, the position with the best fitness value is selected as the current optimal solution until the fitness value converges or the maximum number of iterations is reached, and the optimal weight vector W is output. best ;

[0032] Using the optimal weight vector W best Calculate the comprehensive risk index;

[0033] Taking the comprehensive risk index as input, clustering is performed in combination with the pollution load index, the functional damage index and the microbial diversity attenuation index to divide the risk level.

[0034] Preferably, the soil microecological risk assessment result is verified based on the first result and the second result to obtain the final soil microecological risk level, including:

[0035] Intersection matching is performed between the key pollution factor combination in the first result and the priority control factors in the second result to screen out high-risk pollutants;

[0036] When the first risk prediction result and the second risk prediction result are consistent, the final soil microecological risk level is directly determined;

[0037] When there is a level difference between the first risk prediction result and the second risk prediction result, a key factor weighted correction mechanism is triggered;

[0038] According to the difference in weights of the high-risk pollutants in the first result and the second result, a correction coefficient is assigned to the difference dimension, and the final grade is determined by weighted average.

[0039] On the other hand, the present invention also proposes a soybean-producing area soil microecological risk assessment system based on composite pollution, which is used to implement the above-mentioned soybean-producing area soil microecological risk assessment method based on composite pollution, comprising:

[0040] A data acquisition module is used to obtain pollutant data and microbial community data of the soybean-producing soil;

[0041] A parameter calculation module is used to calculate the risk entropy of the pollutant data; and to construct a microecological health index based on the microbial community data;

[0042] a first result prediction module, configured to construct a risk assessment neural network, train the risk assessment neural network based on the pollutant data and the microbial community data, and obtain a first result using the trained risk assessment neural network, the first result including a first risk prediction result and a combination of key pollution factors;

[0043] A second result prediction module, configured to combine the risk entropy and the microecological health index to obtain a second result, wherein the second result includes a second risk prediction result and a priority control factor;

[0044] The risk level and high-risk pollutant module is used to verify the soil microecological risk assessment results based on the first result and the second result, obtain the final soil microecological risk level, and output high-risk pollutants at the same time.

[0045] Through the above technical solutions, it can be seen that compared with the existing technology, the present invention discloses a method and system for assessing soil microecological risk in soybean production areas based on complex pollution. By obtaining soil pollutant data and microbial community data, calculating risk entropy, and constructing a microecological health index, the first and second results are obtained respectively using a risk assessment neural network and a comprehensive model combining principal component analysis, K-means clustering, and predator optimization algorithm. After verification by key factor matching and weighted correction mechanism, the final risk level and high-risk pollutants are determined. The present invention solves the problems of the existing single assessment dimension and insufficient dynamic adaptability. Through multi-source data fusion and multi-method collaboration, it realizes accurate diagnosis and risk warning of soil complex pollution, improves assessment accuracy and decision-making efficiency, and provides technical support for soil microecological risk control and agricultural sustainable development in soybean production areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0047] Figure 1 A schematic flow chart of the method provided by the present invention;

[0048] Figure 2 This is a schematic diagram of the system architecture provided by the present invention. DETAILED DESCRIPTION

[0049] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0050] On the one hand, the embodiment of the present invention proposes a method for assessing soil microecological risk in soybean production areas based on complex pollution, such as Figure 1 As shown, the following steps are included:

[0051] S1. Obtain soil pollutant and microbial community data from soybean-producing areas.

[0052] In this example, pollutant data includes multi-source data, including field-measured concentrations of heavy metals, herbicides, pesticides, and phthalates, as well as spatial coordinates, soil physical and chemical properties, climate factors, and sub-units such as different crop rotation systems and continuous cropping years. This pollutant data is summarized and compiled into a spatially visualized multidimensional database using HTML, representing a correlation matrix between pollutant concentration, spatial coordinates, and physical and chemical properties. Electronic charts (Echarts) are used to visualize and interactively display the various types of pollution at each location, facilitating data overviews and risk assessment and analysis.

[0053] Microbial community data include: obtaining species annotation information of prokaryotes and fungi through 16SrRNA / ITS sequencing (used to calculate α-diversity indicators such as the Shannon index, ACE index, and Pielou index), using metagenomic sequencing to analyze the abundance of functional genes (KEGG annotation metabolic potential, C / N / P / S cycle-related genes), and detecting pathogen loads (such as the average abundance of Fusarium pathogenicity genes FgSLPs, Fmk1, etc.).

[0054] S2. Calculate the risk entropy of pollutant data; construct a microecological health index based on microbial community data.

[0055] The calculation formula of risk entropy RQ is as follows:

[0056]

[0057] Where MEC is the measured concentration of a pollutant, and PNEC is the predicted no-effect concentration (PNEC) of a pollutant. The PNEC value for an individual pollutant compound is determined by the ratio of short-term / long-term toxicity study results (including no-effect concentration (NOEC), lowest effect concentration (LOEC), half-maximal inhibitory concentration (IC50), half-maximal lethal concentration (LC50), or half-maximal effective concentration (EC50) values) to an additional safety factor (AF). According to the recommendations of the European Food Safety Authority and the European Chemicals Agency, the AF value for acute toxicity data on terrestrial organisms (primarily EC50 / LC50) is 1000, and the AF value for long-term or subacute toxicity data (primarily NOEC) is 100. Toxicity data for all target pollutants were retrieved from the ECOTOX knowledgebase, the Pesticide Properties Database (PPDB), and QSARToolbox 4.6 (OECD & ECHA, 2023), a database jointly developed by the Organization for Economic Co-operation and Development and the European Chemicals Agency. In the absence of measured toxicological data, the assessment data were obtained through the US Environmental Protection Agency's prediction software (Ecosar2.2, USEPA).

[0058] The microecological health index is constructed based on microbial community data, including:

[0059] Microbial health indicators are calculated based on microbial community data. Microbial health indicators include but are not limited to microbial diversity indicators (Shannon index (community richness), Pielou index (evenness), ACE index (species coverage)), functional gene abundance (KEGG-annotated metabolic pathway genes (such as carbohydrate metabolism, nitrogen cycle genes), relative abundance of C / N / P / S cycle functional genes), pathogen load (average abundance of target pathogenic genes (such as FgSLPs, Hog1)), etc.

[0060] The possibility of optimizing the indicators will be retained in the subsequent research process of this embodiment. There are also other indicators such as biomarker abundance, which may be increased or decreased after application verification in the future.

[0061] During the actual implementation process, the annotation information of prokaryotes and fungal species was obtained through 16S and ITS sequences to study the impact of different pollution conditions on the characteristics of microbial communities. The Shannon index (Shannonindex), ACE index (Abundance-based coverage estimator metric) representing species richness, and Pielou index (Pielou's sevenness index) representing species distribution uniformity were selected as alpha diversity measurement indicators. The FAPROTAX tool was used to link community structure information with metabolic function to predict the role of microorganisms in the ecosystem, evaluate the functional diversity of sample microorganisms, and combine it with structural equation modeling (SEM) for in-depth analysis. A co-occurrence network analysis was performed on the microbial communities of samples with different pollution conditions, and the degree of influence of microbial interactions was observed through network feature values ​​such as the number of edges, modularity, and positive and negative connections. Based on metagenomic data, the metabolic potential of microorganisms was evaluated with the help of KEGG and VB12, and the nutrient cycling potential was evaluated by calculating the abundance of C / N / P / S genes. The potential pathogenicity was assessed by calculating the average abundance of target pathogenic genes (such as FgSLPs, Fmk1, Hog1, Mpk1 genes, etc.) of pathogens in farmland soil.

[0062] The microbial health indicators were normalized. In this embodiment, the minimum-maximum normalization method was used, in which the pathogen load was an inverse indicator (the higher the value, the higher the health risk, which is inverted after normalization).

[0063] The normalized microbial health indicators are weighted and summed to form the microecological health index MEHI:

[0064]

[0065] w iis the indicator weight, which is determined by the hierarchical analysis method or entropy weight method; X norm,i is the normalized biological health indicator.

[0066] S3. Construct a risk assessment neural network, train the risk assessment neural network based on pollutant data and microbial community data, and use the trained risk assessment neural network to obtain a first result, which includes a first risk prediction result and a combination of key pollution factors.

[0067] The input layer of the risk assessment neural network is the pollutant concentration (RQ value), microbial health indicators (normalized data) and soil physical and chemical parameters (common m-dimensional features). The hidden layer uses the ReLU activation function, and the output layer is the risk level (classification task) and the weight of key pollution factors (regression task).

[0068] The training process includes using a labeled dataset (correspondence between pollution levels and microbial responses), optimizing the loss function (cross entropy + L2 regularization) through stochastic gradient descent, and using feature importance ranking to screen key pollution factor combinations.

[0069] S4. Combining the risk entropy and the microecological health index to obtain a second result, which includes a second risk prediction result and priority control factors, including:

[0070] S41. Principal component analysis (PCA) was used to reduce the dimension and extract features of the component variables of risk entropy (RQ) and microecological health index (MEHI) to obtain the principal components, which were defined as pollution load index (PLI), functional damage index (FDI) and microbial diversity decay index (MDI), respectively.

[0071] Calculate the covariance matrix Σ of the component variables of the standardized risk entropy (RQ) and microecological health index (MEHI), and extract the eigenvalue λ l and the eigenvector e l (l=1,2,…,k, k is the number of principal components, and the cumulative variance contribution rate is usually taken as ≥85%).

[0072] The score of the i-th sample on the l-th principal component is: Among them, p is the total number of indicators (pollution dimension + microbial dimension).

[0073] The Pollution Load Index (PLI) reflects the cumulative intensity of complex pollution and focuses on the pollution dimension. is the principal component score based only on the pollutant RQ value (the first k1 principal components), is the variance contribution rate (weight) of the lth principal component.

[0074] The Functional Damage Index (FDI) quantifies the degree of damage to ecological functions, focusing on the microbial functional dimension. is the principal component score (first k2 principal components) based on the abundance of functional genes and metabolic potential indicators;

[0075] The Microbial Diversity Decline Index (MDI) characterizes the degradation trend of community diversity and focuses on the dimension of microbial structure. is the principal component score based on the diversity index (the first k3 principal components).

[0076] S42. Using the pollution load index, functional damage index, and microbial diversity attenuation index as coordinate axes, a clustering algorithm is used to divide the risk levels and obtain the second risk prediction result.

[0077] In the specific implementation process, the classification of risk levels is not limited to clustering algorithms. Other relevant methods can also be used according to the data situation.

[0078] This embodiment uses K-means clustering as an example. In the process of using K-means clustering to divide risk levels, the predator optimization algorithm is used to adaptively adjust the weights of the pollution load index, functional damage index, and microbial diversity attenuation index. The details are as follows:

[0079] S421. Initialize the population and generate N random weight vectors W = [w1, w2, w3], where w1 is the pollution load index weight, w2 is the functional damage index weight, and w3 is the microbial diversity decay index weight;

[0080] S422. For each weight vector W, calculate the pollution load index, functional damage index, and microbial diversity attenuation index, perform K-means clustering, and calculate the fitness value Fit(W) based on the clustering results. The fitness value calculation formula is as follows:

[0081] Fit(W)=Silhouette Score(W)-Davies-Bouldin Index(W);

[0082] Where Silhouette Score (W) represents the silhouette coefficient, an index of cluster compactness and separation. The closer the value is to 1, the higher the clustering quality. Davies-Bouldin Index (W) represents the Davidson-Bouldin index, an index of inter-class similarity. The smaller the value, the better the clustering effect.

[0083] S423. Iteratively update the positions of the predator and prey according to the fitness value.

[0084] The formula for prey position update (moving towards the optimal predator) is:

[0085] Where, P t is the prey position at the tth iteration (i.e., weight vector); is the current optimal predator position (the weight vector corresponding to the highest fitness in history); r1~U(0,1) are uniformly distributed random numbers; α(t) is the dynamic step size, α max =1,α min =0.1.

[0086] The formula for predator position update (global search + local fine-tuning) is:

[0087]

[0088] r2, r3~U(0,1) are random numbers; β(t) is the predator movement coefficient, β max =0.8,β min =0.2; ΔW is the random perturbation vector; γ is the perturbation intensity, with an initial value of 0.5 and decreasing with iteration.

[0089] S424. After each iteration, the position with the best fitness value is selected as the current optimal solution until the fitness value converges or the maximum number of iterations is reached, and the optimal weight vector W is output. best =[w 1best ,w 2best ,w 3best ];

[0090] S425. Using the optimal weight vector W best The comprehensive risk index I is calculated as follows:

[0091] I=w 1best ·PLI+w 2best FDI+w 3best MDI;

[0092] S426. Taking the comprehensive risk index as input, clustering is performed in combination with the pollution load index, functional damage index and microbial diversity attenuation index to divide the risk level.

[0093] The comprehensive risk index (I) is combined with the three-dimensional indices (PLI, FDI, and MDI) to form a four-dimensional feature vector [PLI, FDI, MDI, I], which is used as the input data for K-means clustering.

[0094] The number of risk levels is preset, and the optimal cluster number K is verified by the elbow rule or silhouette coefficient.

[0095] The K-means++ algorithm is used to optimize the initial centroid distribution.

[0096] Calculate the Euclidean distance of each sample to each centroid and assign it to the cluster with the closest distance;

[0097] Recalculate the centroid of each cluster until the centroid position is stable or the maximum number of iterations is reached;

[0098] The clustering results are labeled to form the second risk prediction result.

[0099] S43. Determine priority control factors based on clustering results and key pollution factor combinations.

[0100] The mean values ​​of PLI, FDI and MDI for each risk level (cluster) were statistically analyzed to analyze the dominant characteristics of pollution load, functional damage and diversity degradation.

[0101] The main pollutants corresponding to each risk level were extracted from the first result (the combination of key pollution factors output by the neural network), and cross-validated with the second result (the pollution load characteristics in the clustering results) to screen out pollutants that were significantly correlated in multiple dimensions (PLI, FDI, MDI).

[0102] Based on the weights of pollutants in the neural network (combination of key pollution factors) and their contributions to PLI, FDI, and MDI, a comprehensive priority score is calculated:

[0103]

[0104] Among them, α and β are adjustment coefficients.

[0105] Set a priority score threshold and screen out pollutants with scores greater than the priority score threshold as pollutants that require priority control.

[0106] S5. Verify the soil microecological risk assessment results based on the first and second results to obtain the final soil microecological risk level, including:

[0107] S51. Intersect and match the key pollution factor combination in the first result with the priority control factors in the second result to screen out high-risk pollutants.

[0108] The “key pollution factor combination” of the first result (neural network) and the “priority control factors” of the second result (cluster analysis) are intersected and matched to screen out “high-risk pollutants” that meet both high data-driven importance and cluster high-risk association, and further filtered through risk entropy thresholds (such as RQ ≥ 1).

[0109] S52. When the first risk prediction result and the second risk prediction result are consistent, the final soil microecological risk level is directly determined.

[0110] S53. When there is a level difference between the first risk prediction result and the second risk prediction result, the key factor weighted correction mechanism is triggered;

[0111] According to the difference in weights of high-risk pollutants in the first and second results, a correction coefficient is assigned to the difference dimension, and the final grade is determined by weighted average.

[0112] The weights of high-risk pollutants in the "key pollution factor combination" (the connection weights of the neural network input layer) are obtained from the neural network model; the contribution weights of high-risk pollutants to the three-dimensional index (PLI / FDI / MDI) are obtained from the results of the predator optimization algorithm.

[0113] Define the weight difference coefficient Δw i =|w 神经网络,i -w 聚类,i |.

[0114] Differentiated correction coefficients are assigned based on the pollutant impact dimension (pollution load / functional damage / diversity attenuation). If pollutant i mainly affects PLI, the correction coefficient q1 = 1-f1Δw i If the main impact is FDI, then the correction factor q2 = 1-f2Δw i ; If the main impact is MDI, then the correction coefficient q3=1-f3Δw i . f1, f2, and f3 are set based on experience.

[0115] Take the average of the correction factors of all high-risk pollutants to obtain the final correction factor M is the number of high-risk pollutants.

[0116] The risk level is digitized (the risk level is mapped to a value: "low risk" = 1, "medium risk" = 2, "high risk" = 3, "very high risk" = 4), and the first result value R1 and the second result value R2 are obtained, and then the weighted average (R final =q·R1+(1-q)·R2) to determine the final level, which is converted into the final risk by rounding or interval division (1.5-2.4 is classified as "medium risk").

[0117] The final output of this step is risk level and high-risk pollutants.

[0118] On the other hand, the embodiment of the present invention also proposes a soybean-producing area soil microecological risk assessment system based on complex pollution, which is used to implement the above-mentioned soybean-producing area soil microecological risk assessment method based on complex pollution, such as Figure 2 As shown, the system includes:

[0119] A data acquisition module is used to obtain pollutant data and microbial community data of the soybean-producing soil;

[0120] Parameter calculation module, used to calculate the risk entropy of pollutant data; construct the microecological health index based on microbial community data;

[0121] A first result prediction module is used to construct a risk assessment neural network, train the risk assessment neural network based on pollutant data and microbial community data, and use the trained risk assessment neural network to obtain a first result, which includes a first risk prediction result and a combination of key pollution factors;

[0122] A second result prediction module is used to combine the risk entropy and the microecological health index to obtain a second result, which includes a second risk prediction result and a priority control factor;

[0123] The risk level and high-risk pollutant output module is used to verify the soil microecological risk assessment results based on the first result and the second result, obtain the final soil microecological risk level, and output high-risk pollutants at the same time.

[0124] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0125] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for assessing soil microecological risk in soybean producing areas based on complex pollution, characterized in that: The following steps are involved: Obtain pollutant data and microbial community data of soybean-producing soil; Calculating the risk entropy of the pollutant data; constructing a microecological health index based on the microbial community data; Constructing a risk assessment neural network, training the risk assessment neural network based on the pollutant data and the microbial community data, and obtaining a first result using the trained risk assessment neural network, the first result including a first risk prediction result and a combination of key pollution factors; Combining the risk entropy and the microecological health index to obtain a second result, the second result including a second risk prediction result and a priority control factor; The soil microecological risk assessment results are verified based on the first result and the second result to obtain the final soil microecological risk level and output high-risk pollutants at the same time.

2. A soybean producing area soil microecological risk assessment method based on complex pollution according to claim 1, characterized in that: The calculation formula of the risk entropy RQ is as follows: Where MEC is the measured concentration of a pollutant and PNEC is the predicted no-effect concentration of a pollutant. The PNEC value is determined by the ratio between the short-term / long-term toxicity test results and the additional safety factor.

3. The method for assessing soil microecological risk in soybean producing areas based on composite pollution according to claim 1, characterized in that: A microecological health index is constructed based on the microbial community data, including: Calculating microbial health indicators based on the microbial community data, wherein the microbial health indicators include but are not limited to microbial diversity indicators, functional gene abundance, and pathogen load; normalizing the microbial health indicator; The normalized microbial health indicators are weighted and summed to form a microecological health index.

4. The method for assessing soil microecological risk in soybean producing areas based on composite pollution according to claim 1, characterized in that: Combining the risk entropy and the microecological health index to obtain a second result includes: Based on principal component analysis, the risk entropy and the microecological health index are integrated to define a pollution load index, a functional damage index, and a microbial diversity attenuation index; Using the pollution load index, the functional damage index, and the microbial diversity attenuation index as coordinate axes, a clustering algorithm is used to divide the risk levels to obtain the second risk prediction result; Priority control factors are determined based on the clustering results combined with key pollution factors.

5. The method for assessing soil microecological risk in soybean producing areas based on composite pollution according to claim 4, characterized in that: In the process of dividing the risk levels by using the clustering algorithm, the weights of the pollution load index, the functional damage index and the microbial diversity attenuation index are adaptively adjusted using the predator optimization algorithm.

6. A soybean producing area soil microecological risk assessment method based on complex pollution according to claim 5, characterized in that: The risk level is divided by a clustering algorithm to obtain the second risk prediction result, including: Initialize the population and generate N random weight vectors W = [w1, w2, w3]; w1 is the pollution load index weight, w2 is the functional damage index weight, and w3 is the microbial diversity decay index weight; For each weight vector W, the pollution load index, the functional damage index, and the microbial diversity attenuation index are calculated, a clustering operation is performed, and a fitness value Fit(W) is calculated based on the clustering results. The fitness value calculation formula is as follows: Fit(W)=Silhouette Score(W)-Davies-Bouldin Index(W); Where Silhouette Score (W) represents the silhouette coefficient, Davies-Bouldin Index (W) represents the Davidson-Bouldin index; Iteratively update the positions of predators and prey based on the current fitness value; After each iteration, the position with the best fitness value is selected as the current optimal solution until the fitness value converges or the maximum number of iterations is reached, and the optimal weight vector W is output. best ; Using the optimal weight vector W best Calculate the comprehensive risk index; Taking the comprehensive risk index as input, clustering is performed in combination with the pollution load index, the functional damage index and the microbial diversity attenuation index to divide the risk level.

7. The method for assessing soil microecological risk in soybean producing areas based on complex pollution according to claim 1, characterized in that: The soil microecological risk assessment results are verified based on the first and second results to obtain the final soil microecological risk level, including: Intersection matching is performed between the key pollution factor combination in the first result and the priority control factors in the second result to screen out high-risk pollutants; When the first risk prediction result and the second risk prediction result are consistent, the final soil microecological risk level is directly determined; When there is a level difference between the first risk prediction result and the second risk prediction result, a key factor weighted correction mechanism is triggered; According to the difference in weights of the high-risk pollutants in the first result and the second result, a correction coefficient is assigned to the difference dimension, and the final grade is determined by weighted average.

8. A soybean producing area soil microecological risk assessment system based on complex pollution, characterized in that: include: A data acquisition module is used to obtain pollutant data and microbial community data of the soybean-producing soil; A parameter calculation module, used to calculate the risk entropy of the pollutant data; constructing a microecological health index based on the microbial community data; a first result prediction module, configured to construct a risk assessment neural network, train the risk assessment neural network based on the pollutant data and the microbial community data, and obtain a first result using the trained risk assessment neural network, the first result including a first risk prediction result and a combination of key pollution factors; A second result prediction module, configured to combine the risk entropy and the microecological health index to obtain a second result, wherein the second result includes a second risk prediction result and a priority control factor; The risk level and high-risk pollutant output module is used to verify the soil microecological risk assessment results based on the first result and the second result, obtain the final soil microecological risk level, and output high-risk pollutants at the same time.

Images