Regional accumulative environmental risk assessment system and method

Abstract

The invention relates to the technical field of environmental risk assessment, and discloses a regional cumulative environmental risk assessment system and method, and the method comprises the steps: constructing a collection network, and dynamically fusing monitoring data and generated data, and obtaining restoration monitoring data; analyzing and identifying emerging pollutants; a collaborative migration network of emerging pollutants and traditional pollutants is constructed, and the migration probability is calculated; establishing a dynamic conversion kinetic equation, and predicting a newly-emerging pollutant conversion product; in combination with an intergenerational accumulation model, simulating cross-generation accumulation of emerging pollutants in the biocenosis; nonlinear exposure-response mapping is constructed, short-term monitoring data and long-term repair monitoring data are fused, a sudden change early warning index is designed, and a risk critical point is recognized in advance; performing inversion updating on the emerging pollution migration parameters, and optimizing the emerging pollution migration parameters; fusing multi-source evidences, and judging whether the risk of the emerging pollutants exceeds the standard or not; and constructing a full-life-cycle decision-making model, designing a reinforcement learning dynamic strategy library updating strategy, and shortening the measure response time.

Description

Technical Field

[0001] This invention relates to the field of environmental risk assessment technology, and specifically to a regional cumulative environmental risk assessment system and method. Background Technology

[0002] Environmental risk assessment involves data from meteorological, hydrological, pollution source emission, and ecological monitoring departments. However, the spatiotemporal resolution of different systems varies significantly, and there is a lack of unified coding standards and sharing platforms. Groundwater monitoring data in chemical industrial parks is collected by environmental protection departments, while geological structure data is managed by natural resources departments. It is difficult to directly overlay the two to analyze pollutant migration paths, resulting in blind spots in risk source identification, especially in transboundary watersheds or ecologically sensitive areas. Cumulative risk assessment relies on time series data, but existing technologies lack historical baseline data, making it difficult to quantify the long-term accumulation rate of historical baseline data. For emerging pollutants, the current monitoring system does not cover them, and research on the migration and transformation patterns and ecotoxicity of emerging pollutants in soil and water bodies is still in its early stages. Existing assessment models cannot predict the long-term impact of emerging pollutants on human health.

[0003] In existing technologies, data fusion models are mostly statically weighted and cannot dynamically adjust the contribution ratio of data sources according to environmental changes; existing technology assessment reports can identify high-risk areas, but the implementation of supporting control measures is lagging behind, leading to the expansion of the pollution range; current environmental quality standards are mostly based on short-term exposure limits of single pollutants and do not consider long-term cumulative effects, resulting in ecotoxicity events in some areas even if they meet the standards.

[0004] Therefore, there is a need to provide a regional cumulative environmental risk assessment system and method. Summary of the Invention

[0005] The purpose of this invention is to provide a regional cumulative environmental risk assessment system and method. To solve the above-mentioned problems in the prior art, this invention achieves this through the following technical solution:

[0006] In a first aspect, the present invention provides a method for assessing regional cumulative environmental risks, which specifically includes the following steps:

[0007] Step 1: Construct a data acquisition network, dynamically fuse monitoring data and generated data using the exponential equilibrium formula, and repair long-term data gaps to obtain repaired monitoring data; analyze and identify emerging pollutants using a high-resolution mass spectrometry feature library.

[0008] Step 2: Construct a co-migration network of emerging pollutants and traditional pollutants, calculate the migration probability to quantify the interaction between traditional and emerging pollutants; establish a dynamic transformation kinetic equation, and predict the transformation products of emerging pollutants by combining temperature and precursor parameters.

[0009] Step 3: Combining the intergenerational accumulation model, the exponential function is used to simulate the cross-generational accumulation of emerging pollutants in biological communities; a deep belief network is used to construct a nonlinear exposure-response mapping to reduce the prediction error of emerging pollution risks.

[0010] Step 4: Integrate short-term monitoring data with long-term repair monitoring data, and construct a spatiotemporal continuous risk field using the norm formula; combine the Lyapunov index to design a mutation early warning index, and amplify the signal in the chaotic region through the exponential term to identify the risk critical point in advance;

[0011] Step 5: Invert and update the migration parameters of emerging pollutants, optimize the migration parameters of emerging pollutants by combining the exponential constraint formula; integrate multi-source evidence to determine whether the risk of emerging pollutants exceeds the standard;

[0012] Step Six: Based on the judgment of exceeding the standard, construct a full life cycle decision model, and balance the consumption of source emission reduction, migration blocking and remediation through an exponential coupling formula; design a reinforcement learning dynamic strategy library to update the strategy and shorten the response time of the measures.

[0013] Secondly, the regional cumulative environmental risk assessment system provided by the embodiments of the present invention specifically includes the following modules:

[0014] Acquisition and Restoration Module: Constructs an acquisition network, dynamically fuses monitoring data and generated data using an exponential equilibrium formula, repairs long-term data gaps to obtain restored monitoring data; analyzes and identifies emerging pollutants using a high-resolution mass spectrometry feature library;

[0015] Co-migration module: Constructs a co-migration network between emerging and traditional pollutants, calculates migration probabilities to quantify the interaction between traditional and emerging pollutants; establishes dynamic transformation kinetic equations, and predicts the transformation products of emerging pollutants by combining temperature and precursor parameters.

[0016] Simulation mapping module: Combining intergenerational accumulation model, it simulates the cross-generational accumulation of emerging pollutants in biological communities through exponential function; it uses deep belief network to construct nonlinear exposure-response mapping to reduce the prediction error of emerging pollution risk.

[0017] Identification and early warning module: It integrates short-term monitoring data with long-term repair monitoring data and uses the norm formula to construct a spatiotemporal continuous risk field; it combines the Lyapunov index to design a mutation early warning index, which amplifies the signal in the chaotic region through the exponential term and identifies the risk critical point in advance;

[0018] Inversion Judgment Module: Inverts and updates the migration parameters of emerging pollutants, optimizes the migration parameters of emerging pollutants by combining exponential constraint formulas, and integrates multi-source evidence to determine whether the risk of emerging pollutants exceeds the standard;

[0019] Decision update module: Based on the judgment of exceeding the standard, a full life cycle decision model is constructed, and the consumption of source emission reduction, migration blocking and repair is balanced through an exponential coupling formula; a reinforcement learning dynamic strategy library update strategy is designed to shorten the response time of measures.

[0020] The beneficial effects of this invention are:

[0021] 1. By using a GAN-Transformer hybrid model and a hierarchical repair strategy to solve long-term data gaps, a region-specific mass spectrometry library and multi-algorithm joint detection are constructed to overcome the bottleneck of identifying unknown emerging pollutants; a directed weighted network for the collaborative migration of traditional and emerging pollutants is built, and the interaction effect is quantified by the migration probability formula; and the transformation kinetic equations with multiple precursors and temperature dynamic correction are used to accurately predict the transformation products of emerging pollutants.

[0022] 2. The study incorporates detoxification enzyme activity and incubation cycle into an intergenerational cumulative model, using an exponential function to simulate cross-generational transmission. It addresses the challenge of nonlinear exposure-response fitting for the long-term effects of low-dose emerging pollutants through deep belief networks and high-power terms. It integrates long-term remediation data with short-term monitoring data to generate a spatiotemporally continuous risk field. Lyapunov exponents are introduced to design a mutation early warning index, enabling early identification of risk thresholds. Adaptive MCMC-Bayesian inversion and literature constraints are proposed to address the lack of migration parameters for emerging pollutants. Evidence theory is used to integrate multi-source evidence, enhancing the credibility of risk exceedance judgments. A full life-cycle model of source reduction, migration blocking, and remediation is constructed, using an exponential coupling formula to balance costs and risks. A reinforcement learning dynamic strategy library is designed, using a softmax exponential update strategy to shorten the response time of control measures. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart of the steps of a regional cumulative environmental risk assessment method provided in Embodiment 1 of the present invention;

[0025] Figure 2 This is a schematic diagram of the structure of a regional cumulative environmental risk assessment system provided in Embodiment 2 of the present invention. Detailed Implementation

[0026] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0027] Example 1: As Figure 1 As shown in the figure, the regional cumulative environmental risk assessment method provided by this invention specifically includes the following steps:

[0028] Step 1: Construct a data acquisition network and dynamically fuse measured and generated data using the exponential equilibrium formula to repair long-term data gaps; at the same time, analyze and identify emerging pollutants through a high-resolution mass spectrometry feature library to solve the problems of insufficient long-term data continuity and the identification of emerging pollutants.

[0029] In a specific embodiment, emerging pollutant precursors in the target area are monitored by a low-power slow-release sensor, a preset sampling frequency is set, and micro-meteorological data of the target area are recorded synchronously by deploying meteorological stations. The micro-meteorological data includes temperature, humidity and air pressure.

[0030] Water or soil samples are collected according to a preset capture cycle to capture emerging pollutants. Primary or secondary mass spectra are obtained by ultra-high performance liquid chromatography-high resolution mass spectrometry to construct a region-specific characteristic spectral library.

[0031] It should be noted that the region-specific characteristic spectral library refers to a database formed by systematically integrating the mass spectrometry characteristic information of emerging pollutants in a specific region through ultra-high performance liquid chromatography-high resolution mass spectrometry analysis of water or soil samples in the target region. The mass spectrometry characteristic information includes, but is not limited to: mass-to-charge ratio, retention time and fragment ions.

[0032] Integrate online and historical monitoring data of the target area, and calculate the data missing rate of monitoring data for each time interval;

[0033] To obtain the number of actual monitoring data points for the interval period, subtract the theoretical number of monitoring data points for the interval period from the actual number of monitoring data points to obtain the number of missing monitoring data points.

[0034] The missing data rate is obtained by dividing the number of missing monitoring data by the theoretical number of monitoring data.

[0035] For any given time period, if the missing data rate is less than or equal to a preset missing data rate threshold, a GAN-Transformer hybrid model is used for data repair, using the formula:

[0036]

[0037] Analysis yielded repair monitoring data, among which, Based on actual monitoring data, For time steps, It is a natural constant. As a weighting adjustment factor, The time decay factor, To fill in the data, It is random noise;

[0038] It should be noted that the GAN-Transformer hybrid model represents an architecture that combines the generative adversarial training capability of Generative Adversarial Networks (GANs) with the long sequence modeling and attention mechanism capabilities of Transformers. Its core is to overcome the limitations of a single model in complex data generation or sequence understanding by leveraging the complementary advantages of both.

[0039] If the data missing rate is greater than the preset missing rate threshold, the corresponding time interval will be divided into multiple sub-time intervals on average, and the data missing rate of each sub-time interval will be calculated.

[0040] If the missing data rate of a sub-period is less than or equal to the preset missing data rate threshold, the GAN-Transformer hybrid model is used for repair.

[0041] If the data missing rate of a sub-period is greater than the preset missing rate threshold, then ignore it;

[0042] Once all sub-time periods of the corresponding time interval have been repaired, the data missing rate of the corresponding time interval is recalculated.

[0043] If the data missing rate is less than or equal to the preset missing rate threshold, the GAN-Transformer hybrid model is used for repair.

[0044] If the missing data rate is greater than the preset missing data rate threshold, the sub-periods with missing data rates greater than the preset missing data rate threshold will be cleared one by one in descending order of missing data rate until the missing data rate of the corresponding interval is less than or equal to the preset missing data rate threshold, and the GAN-Transformer hybrid model will be used for repair.

[0045] After the monitoring data from all time periods has been repaired, the sample characteristics of the target pollutant are compared with a known common pollutant spectral library to determine whether it is a common pollutant. If it is not a common pollutant, the characteristics of emerging pollutants are extracted using the formula:

[0046]

[0047] Analysis yielded the pollution feature matching degree ,in, This represents the mass of the ion in the mass spectrometry analysis. For the index of characteristic peaks, To adjust the parameters for the decay rate of the exponential term, This represents the charge number of the ions in the mass spectrometry analysis. Given the known charge numbers of ions in the spectral library representing emerging pollutants. Let be the peak intensity of the k-th characteristic peak in the sample. The maximum peak intensity of all characteristic peaks in the sample;

[0048] If the pollution characteristic matching degree is greater than the preset pollutant matching degree standard, it is marked as a known emerging pollutant;

[0049] If the pollution feature matching degree is less than or equal to the preset pollutant matching degree standard, after baseline correction of the mass spectrometry data, the density clustering algorithm is used to identify the characteristic peak clusters of potential unknowns.

[0050] For example, retention time, mass-to-charge ratio, and peak intensity are used as three-dimensional features, combined with local anomaly factors to detect outliers, and then principal component analysis is used to screen clusters that significantly deviate from the known distribution of pollutants.

[0051] Step 2: Construct a collaborative migration network between emerging and traditional pollutants, quantify the interaction between traditional and emerging pollutants using migration probability formulas; establish dynamic transformation kinetic equations, combine temperature and precursor parameters to predict the transformation products of emerging pollutants, and solve the problem of unknown migration and transformation laws.

[0052] Construct a directed weighted network of pollution sources, media, and receptors. Nodes represent pollution sources and environmental media. Pollution sources include both traditional and emerging pollution sources. Edge weights represent migration probabilities, expressed by the formula:

[0053]

[0054] Analysis yields migration probabilities ,in, Let be the migration rate constant, describing the migration rate characteristics along the path from pollution source i to target medium j. Let i be the concentration of pollutants in pollution source i. Let j be the surface area of ​​the target medium. It is a natural constant. Let be the migration distance from source i to target j. For characteristic length, It is the minimum regularization term. The influence coefficient of the medium composition. This is a media composition correction factor; when the media composition correction factor reflects components that are unfavorable to migration, This will decrease, thereby reducing the overall migration probability;

[0055] For any emerging pollutant, the dynamic balance between its consumption and generation is described using the law of conservation of mass. This involves combining the consumption resulting from the pollutant's own transformation with the generation resulting from the transformation of its precursors. The summation of all precursors that can generate the emerging pollutant reflects the comprehensive contribution of multiple precursors. The instantaneous rate of change in the concentration of the emerging pollutant is then analyzed and calculated using the formula:

[0056]

[0057] The instantaneous change rate of concentration of emerging pollutants was obtained through analysis. ,in, Let t be the concentration of the emerging pollutant. For time steps, Let be the rate constant for the self-transformation of emerging pollutants, describing the intrinsic transformation rate of emerging pollutants. For temperature correction, It is a natural constant. This is the temperature influence coefficient. For real-time temperature, Index of precursors to emerging pollutants. This is the rate constant for the conversion of precursors into emerging pollutants. The concentration of the precursor at time t;

[0058] The analysis shows the rising or falling trend of emerging pollutant concentrations when real-time temperature increases (i.e., the temperature correction term increases, and the self-conversion accelerates), as well as the rising or falling trend of emerging pollutant concentrations when precursor concentrations increase (i.e., the formation term increases), thus enabling dynamic prediction.

[0059] Step 3: Combining the intergenerational accumulation model, the exponential function is used to simulate the cross-generational accumulation of emerging pollutants in biological communities; a deep belief network is used to construct a nonlinear exposure-response mapping to reduce the prediction error of emerging pollution risks and fill the gap in long-term ecological impact assessment.

[0060] This study simulates the intergenerational accumulation of emerging pollutants in biological communities. Specifically, it addresses the lack of long-term ecological impact assessment through exponential function simulation and combines intergenerational accumulation models to calculate the cumulative concentration of emerging pollutants in offspring organisms using the following formula:

[0061]

[0062] Analysis yielded the cumulative concentration of emerging pollutants in offspring organisms. ,in, It is a natural constant. The concentration of pollution in the parent generation, This is the intergenerational transmission coefficient, preset to a value of 0.05. It controls the inherent efficiency of the transmission of emerging pollutants from parent to offspring. The higher the value, the higher the proportion of emerging pollutants transmitted from parent to offspring. Bioaccumulation rate describes the rate at which emerging pollutants actively accumulate in an organism, i.e., the rate at which the concentration of emerging pollutants in an organism increases per unit time, reflecting the organism's ability to accumulate emerging pollutants. The incubation period refers to the length of time it takes for parental emerging pollutants to develop into offspring emerging pollutants. A longer incubation period strengthens the intergenerational cumulative effect. The influence coefficient of the detoxification enzyme is preset to -0.1. The activity level of detoxification enzymes in an organism refers to the level of activity of these enzymes. Detoxification enzymes reduce the accumulation of pollutants in the body by breaking them down and excreting them.

[0063] A nonlinear exposure-response mapping is constructed using deep belief networks to establish an emerging pollution exposure-response model. High powers are introduced to capture the nonlinear dose-effect relationship, as shown by the formula:

[0064]

[0065] Analysis and calculation yielded the probability of exposure to emerging pollutants. The closer the exposure risk probability is to 1, the higher the probability of adverse reactions after exposure to emerging pollutants. For the Sigmoid function, This represents the number of neurons in a single layer of a deep belief network. For neuron indexing, These are the weight parameters of the upper-layer neurons in a deep belief network. These are the bias parameters for neurons in the intermediate layers of a deep belief network. The highest power of the concentration power. For concentration power index, This represents the connection weights from the concentration power term to intermediate layer neurons in a deep belief network. The term represents the k-th power of the concentration of emerging pollutants;

[0066] For example, PFAS exhibits a pattern where low-dose long-term exposure leads to significant effects, and linear relationships cannot fit this pattern. However, higher power terms have the ability to characterize complex dose-response patterns.

[0067] Step 4: Integrate short-term monitoring data and long-term repair data to construct a spatiotemporal continuous risk field using the norm formula; combine the Lyapunov index to design a mutation early warning index, amplify the signal in the chaotic region through the index term, identify the risk critical point in advance, and solve the problem of early warning lag.

[0068] By integrating short-term monitoring data and long-term restoration data, a multi-scale risk field is coupled to output a spatiotemporally continuous risk field, using the multi-scale risk field coupling formula:

[0069]

[0070] Analysis yields a spatiotemporally continuous risk field ,in, To address the accumulated risks associated with long-term data repair, For short-term risks based on real-time monitoring, The enhancement factor is set to a default value of 0.6. Contributing factors to emerging pollutants It is a natural constant;

[0071] Based on a continuous spatiotemporal risk field, and combined with Lyapunov index analysis, a sudden change warning index for emerging pollutants is calculated to identify risk thresholds in advance and address warning lag. This is achieved through the formula:

[0072]

[0073] Analysis yielded a mutation early warning index. ,in, The total number of grid cells. For grid indexing, For the change in risk, For time intervals, For chaos amplification parameters, Let Lyapunov be the index of the i-th grid. It is a natural constant;

[0074] Based on the obtained mutation early warning index, it is compared with the preset mutation early warning threshold to analyze and determine the early warning level;

[0075] If the mutation warning index is greater than or equal to the preset mutation warning threshold, it is determined to be a Level 1 warning;

[0076] If the mutation warning index is less than the preset mutation warning threshold, it is judged as a level two warning;

[0077] Step 5: Invert and update the migration parameters of emerging pollutants, and optimize the migration parameters of emerging pollutants by combining the exponential constraint formula; integrate multi-source evidence to determine whether the risk of emerging pollutants exceeds the standard, and solve the problem of missing emerging pollution parameters;

[0078] Combining adaptive MCMC-Bayesian inversion, this approach addresses the issue of missing parameters for emerging pollutants by dynamically optimizing model parameters and updating the emerging pollutant migration parameter k through Bayesian inversion.

[0079]

[0080] in, Let be the core distribution, representing the probability distribution of the emerging pollution migration parameter k given the measured data D. This is a data-driven term describing the probability of observing data D when the transfer parameter is k. Let be the prior knowledge term, representing the prior knowledge about the probability distribution of the transfer parameter k before observing the data D. The penalty coefficient is... For migration parameters, As a literature constraint, the reasonable typical value of the emerging pollution migration parameter k represents the existing understanding of the reasonable range of k. It is used to constrain the inversion results and avoid the parameter deviating from the generally accepted reasonable range. It is a natural constant;

[0081] Based on the completion of the Bayesian inversion update, an evidence fusion rule is established to integrate multi-source evidence, improving the credibility of uncertainty quantification. Proposition A is set as excessive comprehensive risk, Proposition B as excessive monitoring data risk, and Proposition C as excessive model prediction risk. This is achieved through the following formula:

[0082]

[0083] The analysis yields the overall confidence level of proposition A. ,in, To monitor the basic level of confidence in proposition B based on the data, The model predicts the basic confidence level for proposition C;

[0084] Based on the overall level of trust obtained, a comprehensive judgment is made as to whether the risk of emerging pollutants exceeds the standard.

[0085] If the standard is exceeded, a decision is made based on the full life cycle decision model; if the standard is not exceeded, monitoring continues.

[0086] Step Six: Based on the judgment of exceeding the standard, construct a full life cycle decision model, and balance the costs of source emission reduction, migration blocking and remediation through an exponential coupling formula; design a reinforcement learning dynamic policy library, use softmax exponential to update the policy, shorten the response time of measures and reduce the lag of solutions;

[0087] Design a full life-cycle prevention and control plan for emerging pollutants, construct a full life-cycle decision-making model to output life-cycle-risk coupled decisions, through the formula:

[0088]

[0089] The analysis yields the total decision-making cost, among which, To reduce emissions at the source, Regarding emerging pollution risks Emerging pollution risk enhancement items To mitigate the impact of migration, Regarding the distance of pollution diffusion The pollution diffusion distance enhancement item, To repair the consumption, For the total cumulative concentration Total cumulative concentration enhancement term, It is a natural constant;

[0090] Based on the total decision-making cost, a reinforcement learning dynamic policy library is designed. The policy is updated using the softmax exponent, shortening the response time and reducing solution lag, as shown by the formula:

[0091]

[0092] The analysis yields the update probability of choosing action a in state s, i.e., the update decision. ,in, Let be the probability of choosing action a in state s before the update, i.e., the original decision. For action, For state, To smooth the adjustment coefficient and control the aggressiveness of the strategy update, the preset value is 0.2. For the value of action, It is a natural constant;

[0093] Lifecycle decision-making reduces the costs of controlling emerging pollution; reinforcement learning strategy libraries shorten response time and address the problem of lagging solutions.

[0094] Example 2: As Figure 2 As shown in the figure, the regional cumulative environmental risk assessment system provided by this invention specifically includes the following modules:

[0095] Acquisition and Restoration Module: Constructs an acquisition network, dynamically fuses monitoring data and generated data using an exponential equilibrium formula, repairs long-term data gaps to obtain restored monitoring data; analyzes and identifies emerging pollutants using a high-resolution mass spectrometry feature library;

[0096] Co-migration module: Constructs a co-migration network between emerging and traditional pollutants, calculates migration probabilities to quantify the interaction between traditional and emerging pollutants; establishes dynamic transformation kinetic equations, and predicts the transformation products of emerging pollutants by combining temperature and precursor parameters.

[0097] Simulation mapping module: Combining intergenerational accumulation model, it simulates the cross-generational accumulation of emerging pollutants in biological communities through exponential function; it uses deep belief network to construct nonlinear exposure-response mapping to reduce the prediction error of emerging pollution risk.

[0098] Identification and early warning module: It integrates short-term monitoring data with long-term repair monitoring data and uses the norm formula to construct a spatiotemporal continuous risk field; it combines the Lyapunov index to design a mutation early warning index, which amplifies the signal in the chaotic region through the exponential term and identifies the risk critical point in advance;

[0099] Inversion Judgment Module: Inverts and updates the migration parameters of emerging pollutants, optimizes the migration parameters of emerging pollutants by combining exponential constraint formulas, and integrates multi-source evidence to determine whether the risk of emerging pollutants exceeds the standard;

[0100] Decision update module: Based on the judgment of exceeding the standard, a full life cycle decision model is constructed, and the consumption of source emission reduction, migration blocking and repair is balanced through an exponential coupling formula; a reinforcement learning dynamic strategy library update strategy is designed to shorten the response time of measures.

[0101] The above provides a detailed description of one embodiment of the present invention, but the content described is only a preferred embodiment of the present invention and should not be considered as limiting the scope of the present invention. The above formulas are all dimensionless numerical calculations, and the formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world situation. The preset parameters in the formulas are set by those skilled in the art based on actual conditions and historical experience, and can be adjusted according to actual conditions. The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. All equivalent changes and improvements made in accordance with the scope of the present invention should still fall within the patent coverage of the present invention.

Claims

1. A method for assessing regional cumulative environmental risks, characterized in that, Includes the following steps: A data acquisition network was constructed, and monitoring data and generated data were dynamically fused using an exponential equilibrium formula to repair long-term data gaps and obtain repaired monitoring data. Emerging pollutants were analyzed and identified using a high-resolution mass spectrometry feature library. Construct a collaborative migration network between emerging and traditional pollutants, calculate migration probabilities to quantify the interaction between traditional and emerging pollutants, establish dynamic transformation kinetic equations, and predict the transformation products of emerging pollutants by combining temperature and precursor parameters. Combining intergenerational accumulation models, an exponential function is used to simulate the cross-generational accumulation of emerging pollutants in biological communities; a deep belief network is used to construct a nonlinear exposure-response mapping to reduce the prediction error of emerging pollution risks. By integrating short-term monitoring data with long-term repair monitoring data, a spatiotemporal continuous risk field is constructed using the norm formula; a mutation early warning index is designed by combining the Lyapunov index, which amplifies the signal in the chaotic region through the exponential term and identifies the risk critical point in advance. The migration parameters of emerging pollution are updated by inversion, and the migration parameters of emerging pollution are optimized by combining the exponential constraint formula. By integrating evidence from multiple sources, we can determine whether emerging pollutants pose a risk exceeding the standard. Based on the judgment of exceeding the standard, a full life cycle decision model is constructed, and the source reduction, migration blocking and remediation consumption are balanced through an exponential coupling formula. Design a dynamic policy library for reinforcement learning to update policies and shorten response time.

2. The method for regional cumulative environmental risk assessment according to claim 1, characterized in that, The method for repairing long-term data gaps is as follows: Integrate online and historical monitoring data of the target area, and calculate the data missing rate of monitoring data for each time interval; The missing data rate is obtained. For any given time period, if the missing data rate is less than or equal to a preset missing rate threshold, a GAN-Transformer hybrid model is used for data repair, using the formula: Analysis yielded repair monitoring data, among which, Based on actual monitoring data, For time steps, It is a natural constant. As a weighting adjustment factor, The time decay factor, To fill in the data, It is random noise; If the data missing rate is greater than the preset missing rate threshold, the corresponding time interval will be divided into multiple sub-time intervals on average, and the data missing rate of each sub-time interval will be calculated. If the missing data rate of a sub-period is less than or equal to the preset missing data rate threshold, the GAN-Transformer hybrid model is used for repair; if the missing data rate of a sub-period is greater than the preset missing data rate threshold, it is ignored. Once all sub-periods of the corresponding time interval have been repaired, the data missing rate of the corresponding time interval is recalculated; if the data missing rate is less than or equal to the preset missing rate threshold, the GAN-Transformer hybrid model is used for repair. If the missing data rate is greater than the preset missing data rate threshold, the missing data rates of the sub-periods with missing data rates greater than the preset missing data rate threshold are cleared one by one in descending order until the missing data rate of the corresponding interval is less than or equal to the preset missing data rate threshold, and the GAN-Transformer hybrid model is used for repair.

3. The method for regional cumulative environmental risk assessment according to claim 1, characterized in that, The method for identifying emerging pollutants is as follows: The sample characteristics of the target pollutant are compared with a known common pollutant spectral library to determine whether it is a common pollutant. If it is not a common pollutant, the characteristics of emerging pollutants are extracted using the formula: Analysis yielded the pollution feature matching degree ,in, This represents the mass of the ion in the mass spectrometry analysis. For the index of characteristic peaks, To adjust the parameters for the decay rate of the exponential term, This represents the charge number of the ions in the mass spectrometry analysis. Given the known charge numbers of ions in the spectral library representing emerging pollutants. Let be the peak intensity of the k-th characteristic peak in the sample. The maximum peak intensity of all characteristic peaks in the sample; If the pollution feature matching degree is greater than the preset pollutant matching degree standard, it is marked as a known emerging pollutant; if the pollution feature matching degree is less than or equal to the preset pollutant matching degree standard, after baseline correction of the mass spectrometry data, the density clustering algorithm is used to identify the characteristic peak clusters of potential unknowns.

4. The method for regional cumulative environmental risk assessment according to claim 1, characterized in that, The method for calculating the migration probability is as follows: Construct a directed weighted network of pollution sources, media, and receptors, where nodes represent pollution sources and environmental media, and edge weights represent migration probabilities, using the formula: Analysis yields migration probabilities ; in, The migration rate constant is Let i be the concentration of pollutants in pollution source i. Let j be the surface area of ​​the target medium. It is a natural constant. Let be the migration distance from source i to target j. For characteristic length, It is the minimum regularization term. The influence coefficient of the medium composition. This is a correction factor for the composition of the medium.

5. The method for regional cumulative environmental risk assessment according to claim 1, characterized in that, The method for predicting the transformation products of emerging pollutants is as follows: The instantaneous rate of change of concentration of emerging pollutants was analyzed and calculated using the formula: The instantaneous change rate of concentration of emerging pollutants was obtained through analysis. ,in, Let t be the concentration of the emerging pollutant. For time steps, The rate constant for the self-transformation of emerging pollutants. For temperature correction, It is a natural constant. This is the temperature influence coefficient. For real-time temperature, Index of precursors to emerging pollutants. This is the rate constant for the conversion of precursors into emerging pollutants. The concentration of the precursor at time t; The system analyzes the rising or falling trends of emerging pollutant concentrations in real time as temperature increases, and the rising or falling trends of emerging pollutant concentrations as precursor concentrations increase, enabling dynamic prediction.

6. The method for regional cumulative environmental risk assessment according to claim 1, characterized in that, The method for constructing the nonlinear exposure-response mapping is as follows: The cross-generational accumulation of emerging pollutants in biological communities was simulated, and the cumulative concentration of emerging pollutants in offspring organisms was obtained through formula analysis. A nonlinear exposure-response mapping is constructed using deep belief networks to establish an emerging pollution exposure-response model. High powers are introduced to capture the nonlinear dose-effect relationship, as shown by the formula: Analysis and calculation yielded the probability of exposure to emerging pollutants. ,in, For the Sigmoid function, This represents the number of neurons in a single layer of a deep belief network. For neuron indexing, These are the weight parameters of the upper-layer neurons in a deep belief network. These are the bias parameters for neurons in the intermediate layers of a deep belief network. The highest power of the concentration power. For concentration power index, This represents the connection weights from the concentration power term to intermediate layer neurons in a deep belief network. This is the k-th power of the concentration of emerging pollutants.

7. The method for regional cumulative environmental risk assessment according to claim 1, characterized in that, The method for identifying risk thresholds in advance is as follows: By integrating short-term monitoring data and long-term restoration data, a multi-scale risk field is coupled to output a spatiotemporally continuous risk field, using the multi-scale risk field coupling formula: Analysis yields a spatiotemporally continuous risk field ,in, To address the accumulated risks associated with long-term data repair, For short-term risks based on real-time monitoring, As a reinforcing factor, Contributing factors to emerging pollutants It is a natural constant; Based on a spatiotemporally continuous risk field, the abrupt change early warning index of emerging pollutants is analyzed and calculated using the formula: Analysis yielded a mutation early warning index. ; in, The total number of grid cells. For grid indexing, For the change in risk, For time intervals, For chaos amplification parameters, Let Lyapunov be the index of the i-th grid. It is a natural constant; Based on the obtained mutation early warning index, it is compared with the preset mutation early warning threshold to analyze and determine the early warning level.

8. The method for regional cumulative environmental risk assessment according to claim 1, characterized in that, The method for determining whether emerging pollutants exceed the risk standard is as follows: Combining adaptive MCMC-Bayesian inversion, this approach addresses the issue of missing parameters for emerging pollutants by dynamically optimizing model parameters and updating the emerging pollutant migration parameter k through Bayesian inversion. in, As the core distribution, For data-driven items, As a priori cognitive item, The penalty coefficient is... For migration parameters, For document constraints, It is a natural constant; Based on the completion of the Bayesian inversion update, an evidence fusion rule is established to integrate multi-source evidence, improving the credibility of uncertainty quantification. Proposition A is set as excessive comprehensive risk, Proposition B as excessive monitoring data risk, and Proposition C as excessive model prediction risk. This is achieved through the following formula: The analysis yields the overall confidence level of proposition A. ; in, To monitor the basic level of confidence in proposition B based on the data, The model predicts the basic confidence level for proposition C; The overall trust level obtained is used to make a comprehensive judgment on whether the risk of emerging pollutants exceeds the standard.

9. The method for regional cumulative environmental risk assessment according to claim 1, characterized in that, The method for shortening the response time of the measures is as follows: A full life-cycle prevention and control plan is designed for emerging pollutants, a full life-cycle decision-making model is constructed to output life-cycle-risk coupled decisions, and the total decision cost is obtained through formula analysis. Based on the total decision-making cost, a reinforcement learning dynamic policy library is designed. The policy is updated using the softmax exponent, shortening the response time. This is achieved through the formula: Analysis leads to updated decisions ,in, For the original decision, For action, For state, For smoothing adjustment coefficient, For the value of action, It is a natural constant.

10. A regional cumulative environmental risk assessment system, the system being used to perform the assessment method according to any one of claims 1-9, characterized in that, include: Data Acquisition and Repair Module: Constructs an acquisition network, dynamically merges monitoring data and generated data using the exponential balance formula, and repairs long-term data gaps to obtain repaired monitoring data; Emerging pollutants were identified by analyzing a high-resolution mass spectrometry feature library. Co-migration module: Constructs a co-migration network between emerging and traditional pollutants, calculates migration probabilities to quantify the interaction between traditional and emerging pollutants; establishes dynamic transformation kinetic equations, and predicts the transformation products of emerging pollutants by combining temperature and precursor parameters. Simulation mapping module: Combining the intergenerational accumulation model, it simulates the cross-generational accumulation of emerging pollutants in biological communities through an exponential function; A nonlinear exposure-response mapping is constructed using deep belief networks to reduce the prediction error of emerging pollution risks. Identification and early warning module: It integrates short-term monitoring data with long-term repair monitoring data and uses the norm formula to construct a spatiotemporal continuous risk field; it combines the Lyapunov index to design a mutation early warning index, which amplifies the signal in the chaotic region through the exponential term and identifies the risk critical point in advance; Inversion Judgment Module: Inverts and updates the migration parameters of emerging pollution, and optimizes the migration parameters of emerging pollution by combining the exponential constraint formula; By integrating evidence from multiple sources, we can determine whether emerging pollutants pose a risk exceeding the standard. Decision update module: Based on the judgment of exceeding the standard, a full life cycle decision model is constructed, and the consumption of source emission reduction, migration blocking and repair is balanced through an exponential coupling formula; Design a dynamic policy library for reinforcement learning to update policies and shorten response time.

Images