A risk assessment method and system for pesticide residues in traditional Chinese medicine

Abstract

The present invention discloses a risk assessment method and system for pesticide residues in traditional Chinese medicines, which specifically relate to the technical field of traditional Chinese medicine quality control. The method is used to solve the problem of inaccurate assessment caused by ignoring the influence of processing technology on pesticide morphological transformation and the synergistic toxicity of multiple pesticides in the existing technology. The method obtains pesticide residue detection data and processing technology parameters; associates with a pesticide chemical transformation database to determine the transformation morphology and conversion rate; quantifies toxicity weights based on biological network topology data and a graph convolution model; combines chaos theory with a recurrent neural network to fuse environmental disturbances to correct actual residues; dynamically optimizes weights for the synergistic effects of multiple pesticide residues; generates cross-batch risk warnings based on historical data; and significantly improves the accuracy of pesticide residue toxicity assessment through multi-dimensional dynamic coupling analysis, thereby reducing the safety risks of pesticide residues in traditional Chinese medicines.

Description

Technical Field

[0001] The present invention relates to the technical field of traditional Chinese medicine quality control, and more particularly to a risk assessment method and system for pesticide residues in traditional Chinese medicine. Background Art

[0002] During the production process, Chinese medicinal materials often need to be processed to enhance their efficacy or reduce their toxicity, and pesticide residues are one of the key factors affecting their safety. The chemical form of pesticide residues may change during the processing of Chinese medicinal materials. The current risk assessment of pesticide residues in Chinese medicinal materials mainly relies on the detection of total pesticide residues and directly compares the detection data of total pesticide residues with the preset threshold to determine whether the medicinal materials meet the safety standards. However, the potential impact of the processing technology on the toxicity of pesticide residues has not been fully considered, and the impact of the processing technology on the toxicity of residual pesticides has not been included in the evaluation logic, which may lead to deviations between the evaluation results and the actual safety of the medicinal materials.

[0003] In the existing technology, the determination of pesticide residue risk is based solely on the detection results of the total residue amount, without distinguishing the toxicity differences between the original form of the residual pesticide and the transformed form that may be generated after processing. This leads to some Chinese medicinal materials whose actual toxicity is significantly reduced after processing being misjudged as high-risk due to excessive total residue amounts. This not only affects the rational use of medicinal materials, but may also cause inaccurate assessment results and unnecessary waste of resources. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present invention provide a risk assessment method and system for pesticide residues in traditional Chinese medicines to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A risk assessment method for pesticide residues in traditional Chinese medicines, comprising the following steps:

[0007] S1. Obtain pesticide residue test data including pesticide names and total residue amounts in the Chinese medicinal materials to be evaluated; extract processing parameters including processing temperature and duration;

[0008] S2. Associating a pre-stored pesticide chemical transformation database with the pesticide name and preparation process parameters to determine the transformation form and corresponding conversion rate of the pesticide under the preparation process;

[0009] S3. Based on the pre-stored biological network topology data, the node influence of the transformation morphology is calculated through the graph convolution model to generate the toxicity weight value;

[0010] S4. Based on the micro-disturbance data of the preparation environment collected by sensors, the impact of nonlinear disturbances on the conversion rate is quantified through chaos theory, and the actual residual amount is dynamically corrected in combination with a recursive neural network;

[0011] S5. If two or more pesticide residues are present, the toxicity weight value is dynamically adjusted based on the synergistic effect relationship associated with the mechanism of action type;

[0012] S6. Generate risk assessment results based on actual residues and adjusted toxicity weight values, and perform cross-batch risk prediction in combination with historical batch data.

[0013] In a preferred embodiment, S1 includes:

[0014] S1a, using IoT sensors to obtain pesticide residue detection data in Chinese herbal medicine planting areas. The pesticide residue detection data includes the pesticide name and the corresponding total residue amount;

[0015] S1b, matching the pesticide names with the pre-stored pesticide detection data cleaning rule library, and performing outlier removal and standardization on the pesticide residue detection data;

[0016] S1c, extracting the time series data of processing temperature and duration from the control system of the processing equipment to generate processing parameters;

[0017] S1d. According to the pesticide name and processing parameters, associate the production batch code of the Chinese medicinal materials and establish a batch data index.

[0018] In a preferred embodiment, S2 includes:

[0019] S2a, matching the reaction pathway templates in the pesticide chemical transformation database based on the pesticide name and processing temperature, and screening the candidate transformation forms corresponding to the preparation process;

[0020] S2b, calculating the conversion rate of each conversion form using a conditional probability model based on the duration and kinetic parameters of the candidate conversion forms;

[0021] S2c, based on experimental verification data in the pesticide chemical transformation database associated with the preparation stage labels, the conversion rate is corrected with confidence;

[0022] S2d. Store the corrected transformation form and conversion rate into the pesticide chemical transformation database according to the batch data index.

[0023] In a preferred embodiment, S3 includes:

[0024] S3a, construct a metabolism-toxicity association network, where nodes are transformation forms and human metabolites, and edges are toxicity transmission pathways;

[0025] S3b, extracting node feature vectors based on biological network topology data, including chemical descriptors of transformation forms and toxicity values ​​of metabolites;

[0026] S3c, aggregate neighborhood node features through the multi-layer convolutional layers of the graph convolutional model to generate node embedding representation;

[0027] S3d, calculate the topological centrality score based on the node embedding representation, and generate the toxicity weight value by combining the toxicity value;

[0028] S3e. The toxicity weight value is associated and stored in the pesticide chemical transformation database according to the batch data index.

[0029] In a preferred embodiment, S4 includes:

[0030] S4a, collecting time series data of temperature, humidity, and airflow velocity in the processing equipment as micro-disturbance data of the processing environment, and storing them in association with batch data index;

[0031] S4b, based on the Lyapunov exponent, the chaotic characteristics of the perturbation data are analyzed in stages. The first stage quantifies the intensity of the immediate nonlinear perturbation, and the second stage extracts the cumulative effect coefficient of the perturbation of historical batches.

[0032] S4c, fusing the instantaneous nonlinear perturbation intensity and the perturbation cumulative effect coefficient through a recursive neural network to generate a dynamic correction factor, synchronously correlating the conversion rate of step S2 with the toxicity weight value of step S3;

[0033] S4d, performing bidirectional correction on the total residue based on the dynamic correction factor to generate the actual residue;

[0034] S4e. The actual residue amount is stored in the pesticide chemical transformation database according to the batch data index and the correction path type is marked.

[0035] In a preferred embodiment, the bidirectional correction includes: forward correction to amplify the residue of highly toxic forms based on a dynamic factor, and reverse correction to suppress low-confidence conversion rate errors.

[0036] In a preferred embodiment, S5 includes:

[0037] S5a. If two or more pesticide residues are present, determine the type of pesticide action mechanism based on the pesticide name;

[0038] S5b, based on the matching of the mechanism of action type with the pre-existing synergistic effect rule library, to screen the synergistic effect pattern of the pesticide combination;

[0039] S5c, extracting metabolic pathway intersection nodes based on the metabolism-toxicity association network of step S3, and quantifying the competitive inhibition strength of the pesticide combination on the intersection nodes;

[0040] S5d, based on the competitive inhibition strength and the topological centrality score of step S3, construct a dynamic attenuation model of the toxicity transmission path and generate a comprehensive toxicity correction coefficient;

[0041] S5e: Apply the comprehensive toxicity correction coefficient to the toxicity weight value of step S3 to generate an adjusted comprehensive toxicity weight, which is stored in the pesticide chemical transformation database according to the batch data index.

[0042] In a preferred embodiment, the synergistic mode of action includes competitive inhibition of metabolic pathways and superposition of target organ toxicity.

[0043] In a preferred embodiment, S6 includes:

[0044] S6a, constructing a time series data set of risk index based on the actual residue and the comprehensive toxicity weight adjusted in step S5e;

[0045] S6b, combining the time series dataset of historical batch data, extracting cross-batch risk trend characteristics through grey correlation analysis;

[0046] S6c. Generate a dynamic risk warning model based on the node influence of cross-batch risk trend characteristics and transformation forms;

[0047] S6d. The output results of the dynamic risk warning model are stored in association with the batch data index to generate a cross-batch risk prediction report;

[0048] S6e. Update the parameters of the dynamic risk warning model according to the real-time batch data and adjust the risk warning level threshold.

[0049] In another aspect, the present invention provides a risk assessment system for pesticide residues in traditional Chinese medicines, comprising the following modules:

[0050] Data collection module: obtains pesticide residue test data including pesticide names and total residue amounts in the Chinese medicinal materials to be evaluated; extracts processing parameters including processing temperature and duration;

[0051] Conversion analysis module: associates the pre-stored pesticide chemical conversion database with the pesticide name and preparation process parameters to determine the pesticide's conversion form and corresponding conversion rate under the preparation process;

[0052] Weight generation module: Based on the pre-stored biological network topology data, the node influence of the transformation morphology is calculated through the graph convolution model to generate the toxicity weight value;

[0053] Dynamic correction module: Based on the micro-disturbance data of the preparation environment collected by sensors, the impact of nonlinear disturbances on the conversion rate is quantified through chaos theory, and the actual residual amount is dynamically corrected in combination with a recursive neural network;

[0054] Synergistic adjustment module: If there are two or more pesticide residues, the toxicity weight value is dynamically adjusted according to the synergistic effect relationship associated with the mechanism of action type;

[0055] Prediction and assessment module: Generates risk assessment results based on actual residues and adjusted toxicity weight values, and performs cross-batch risk prediction in combination with historical batch data.

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] 1. Through multi-dimensional dynamic coupling analysis, the accuracy and adaptability of pesticide residue risk assessment in traditional Chinese medicines have been significantly improved. By correlating processing parameters with the pesticide chemical transformation database, the impact of pesticide form transformation on toxicity during the processing process is directly quantified, solving the problem of misjudgment of toxicity caused by ignoring the processing process in traditional methods. Through conversion rate calculation and dynamic adjustment of toxicity weights, the actual toxicity risk is accurately identified, avoiding misjudgment caused by relying solely on total residues. At the same time, the metabolic-toxicity network topology analysis is introduced, combined with the graph convolution model to quantify the propagation influence of the transformation form in the biological network, making the toxicity assessment more in line with the actual biological effects.

[0058] 2. By integrating environmental micro-disturbance data through chaos theory and recursive neural networks, pesticide residues and toxicity weights are dynamically corrected, solving the problem that traditional static models cannot adapt to dynamic changes in the preparation environment; through real-time disturbance analysis and dynamic correction, the robustness of residue prediction is significantly improved; for the synergistic effect of multiple pesticide residues, based on the coupling analysis of the type of action mechanism and the cross-interference of metabolic network nodes, adaptive adjustment of toxicity weights is achieved to avoid the error accumulation of traditional linear superposition models, and cross-batch risk prediction is combined with historical data trends and real-time parameter updates to form a risk assessment-early warning-feedback closed loop, providing continuous optimization decision support for medicinal material safety control. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 This is a flow chart of a risk assessment method for pesticide residues in traditional Chinese medicines according to the present invention;

[0060] Figure 2 This is a schematic diagram of the structure of a risk assessment system for pesticide residues in traditional Chinese medicines according to the present invention. DETAILED DESCRIPTION

[0061] The following will provide a clear and complete description of 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. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0062] Example 1: Figure 1 The present invention provides a risk assessment method for pesticide residues in traditional Chinese medicines, comprising the following steps:

[0063] S1. Obtain pesticide residue test data including pesticide names and total residue amounts in the Chinese medicinal materials to be evaluated; extract processing parameters including processing temperature and duration;

[0064] S2. Associating a pre-stored pesticide chemical transformation database with the pesticide name and preparation process parameters to determine the transformation form and corresponding conversion rate of the pesticide under the preparation process;

[0065] S3. Based on the pre-stored biological network topology data, the node influence of the transformation morphology is calculated through the graph convolution model to generate the toxicity weight value;

[0066] S4. Based on the micro-disturbance data of the preparation environment collected by sensors, the impact of nonlinear disturbances on the conversion rate is quantified through chaos theory, and the actual residual amount is dynamically corrected in combination with a recursive neural network;

[0067] S5. If two or more pesticide residues are present, the toxicity weight value is dynamically adjusted based on the synergistic effect relationship associated with the mechanism of action type;

[0068] S6. Generate risk assessment results based on actual residues and adjusted toxicity weight values, and perform cross-batch risk prediction in combination with historical batch data.

[0069] In step S1a, pesticide residue detection data is collected in real time by high-precision Internet of Things sensors deployed at key nodes in the Chinese medicinal materials planting area. The Internet of Things sensors include gas chromatography sensors, mass spectrometry sensors and multispectral imaging sensors. The gas chromatography sensor is used to detect the chemical characteristic peaks corresponding to the pesticide name, the mass spectrometry sensor is used to determine the total pesticide residue, and the multispectral imaging sensor is used to assist in locating the hot spots of pesticide residue distribution. The pesticide residue detection data is uploaded to the central data processing platform through an encrypted transmission protocol. The pesticide residue detection data includes the pesticide name, total residue and corresponding sampling timestamp. The pesticide name is matched to a standardized name through the international pesticide coding library, and the total residue is stored in milligrams per kilogram.

[0070] In step S1b, the pre-stored pesticide detection data cleaning rule library sets differentiated outlier judgment rules according to the pesticide category. For example, the detection value of organophosphorus pesticides is allowed to fluctuate within ±5% when the temperature is higher than 30 degrees Celsius, and the detection value of pyrethroid pesticides needs to be logarithmically transformed when the humidity exceeds 70%; the cleaning rule library is generated through historical data training, and specifically includes the following operations: time series analysis of the total residue in the pesticide residue detection data, identifying abnormal data points that exceed the preset fluctuation threshold, for example, if the total residue in a certain detection suddenly increases by 50% at adjacent time points, it is marked as abnormal; the abnormal data points are corrected using the sliding window mean interpolation method, and the window size is set according to the pesticide degradation half-life, for example, a 3-hour window is used for pesticides with a half-life of less than 7 days, and a 24-hour window is used for pesticides with a half-life of more than 30 days; standardization processing includes unifying the dimensions of data collected by different sensors, for example, converting the ion intensity value output by the mass spectrometer sensor into a concentration unit consistent with the gas chromatography sensor.

[0071] In step S1c, the control system of the processing equipment is an intelligent temperature control device with a data interface, including a PLC control system or a distributed temperature controller based on the Internet of Things; the time series data of the processing temperature and duration are extracted from the equipment control log, and the data format includes CSV, JSON or time series database records; after extraction, the temperature data is smoothed, for example, the exponentially weighted moving average method is used to eliminate the equipment temperature measurement noise, and the duration data is converted into a cumulative value in minutes; the processing temperature and duration are stored in segments according to the processing stage, for example, the temperature data of the frying stage is separated from the temperature data of the steaming stage, and the corresponding process stage labels are associated.

[0072] In step S1d, the production batch code of Chinese medicinal materials consists of the year, month, day, origin code and production line number, for example, 20231001-GZB-003 represents the batch produced by production line No. 3 in Guizhou production area on October 1, 2023; the association operation includes establishing a mapping relationship between the pesticide name, processing temperature and duration according to the batch code, for example, creating a batch data index table in a relational database, and the index fields include batch code, pesticide name, temperature range and duration interval; the index table is linked to the pesticide chemical transformation database through a foreign key, for example, the batch code is used as the primary key to link to the transformation morphology table in the subsequent steps to ensure that the data across steps is traceable; after the index is generated, multiple groups of pesticide residue detection data under the same batch code are aggregated, for example, the arithmetic mean of the total residue data of different sampling points in the same batch is taken to avoid repeated calculations.

[0073] In step S2a, the reaction path template in the pesticide chemical transformation database is constructed by pre-defining the correspondence between the pesticide category and the temperature range. The pesticide category is classified according to the chemical structure of the pesticide name, for example, pesticides containing organophosphorus groups are classified as organophosphorus, and pesticides containing pyrethroid groups are classified as pyrethroids; the temperature range is divided according to the common temperature range of the preparation process, for example, the temperature range of the frying process is 100 degrees Celsius to 150 degrees Celsius, and the temperature range of the steaming process is 80 degrees Celsius to 120 degrees Celsius; the matching operation includes the following steps: Query the pesticide category and filter the corresponding reaction pathway template based on the processing temperature range. The reaction pathway template includes the main reaction pathway and the side reaction pathway. For example, in the hydrolysis reaction pathway template of organophosphorus pesticides in the range of 100 degrees Celsius to 150 degrees Celsius, the main reaction produces phosphate compounds and the side reaction produces thiophosphate compounds. Candidate transformation forms are extracted from the product list of the template. For example, when the pesticide name is "chlorpyrifos" and the processing temperature is 120 degrees Celsius, the candidate transformation forms include phosphate ester A, phosphate ester B and unreacted chlorpyrifos technical.

[0074] In step S2b, the kinetic parameters include the reaction rate constant and the activation energy. The reaction rate constant is determined by a kinetic experiment simulating the preparation process in a laboratory. For example, a constant temperature chamber is set at 120 degrees Celsius, samples are taken every 30 minutes, and the concentration change of chlorpyrifos is detected by mass spectrometry. The hydrolysis rate constant is calculated to be 0.05h -1 The activation energy parameter is obtained by differential scanning calorimetry. For example, a chlorpyrifos sample is placed in a differential scanning calorimeter, and the thermal decomposition curve is tested at different heating rates. The activation energy is fitted to be 50 kJ / mol. The conditional probability model is constructed based on the Arrhenius equation. The specific logic is: the temperature-corrected rate constant is calculated based on the processing temperature and the activation energy. The formula is: the rate constant is equal to the ratio of the pre-exponential factor multiplied by the negative activation energy of the natural exponential function to the product of the gas constant and the absolute temperature; the reaction progress is calculated based on the product of the duration and the corrected rate constant. For example, the reaction progress of a reaction with a duration of 2 hours is 0.05 h. -1 ×2h=0.1; the conversion rate of each candidate transformation form is distributed according to the molar ratio of reaction progress and product. For example, if the molar ratio of phosphate ester A, phosphate ester B and unreacted original drug is 6:3:1, their conversion rates are 60%, 30% and 10%, respectively; finally, normalization is performed to ensure that the sum of all conversion rates is 100%.

[0075] In step S2c, the experimental verification data comes from the test results of the laboratory simulation of the actual preparation process. For example, under the same processing temperature of 120 degrees Celsius and duration of 2 hours, the peak area of ​​phosphate A is separated and detected by high performance liquid chromatography, and the actual conversion rate is calculated to be 58%, which is 2% different from the calculated value of 60% in step S2b; the confidence correction rule is set according to the deviation range, and the deviation range is determined by the absolute difference between the calculated value and the experimental value. For example, when the deviation is less than 5%, it is high confidence, when the deviation is 5% to 10%, it is medium confidence, and when the deviation is greater than 10%, the manual review process is triggered; the correction operation includes taking the weighted average of the calculated conversion rate and the experimental value as the final conversion rate, and the weight is set according to the confidence level. For example, when the confidence is high, the experimental value weight is 0.8, the calculated value weight is 0.2, and when the confidence is medium, the experimental value weight is 0.5, and the calculated value weight is 0.5; the corrected conversion rate is rounded to two decimal places, for example, the corrected conversion rate of phosphate A is 59.2%, and is associated with the experimental verification data and stored in the pesticide chemical conversion database.

[0076] In step S2d, the batch data index is composed of the batch code generated in step S1d and the pesticide name, for example, the batch code 20231001-GZB-003 and the pesticide name "chlorpyrifos" are combined into 20231001-GZB-003-chlorpyrifos; the storage operation includes creating two data tables in the pesticide chemical transformation database: 1) the transformation form table, the fields of which include the batch data index, the candidate transformation form name, the chemical formula and the CAS number, for example, the batch data index 20231001-GZ The chemical formula of phosphate ester A corresponding to B-003-chlorpyrifos is C10H12O4PS; 2) a conversion rate table, whose fields include batch data index, candidate conversion form name, calculated conversion rate, revised conversion rate, confidence level, and update timestamp. For example, the revised conversion rate of phosphate ester A is 59.2%, and the confidence level is high. The data table establishes a foreign key association with the batch data index table of step S1d via the batch data index. The foreign key association is implemented using SQL statements in a relational database, such as executing "ALTER TABLE conversion rate table ADDFOREIGN KEY (batch data index) REFERENCES batch data index table (batch data index)". During storage, multiple sets of conversion rate data under the same batch data index are sorted in ascending order by timestamp to form a time series data set for cross-batch trend analysis in subsequent steps.

[0077] It is worth noting that the pre-existing pesticide chemical transformation database is constructed by integrating laboratory simulation preparation data, literature public data and public toxicology databases (such as PubChem and KEGG). The specific steps include: 1) Determine the transformation forms of pesticides under different temperatures, humidity and durations through controlled environment experiments. For example, organophosphorus pesticides generate phosphate ester transformation products under a 120°C stir-fry process. The conversion rate is quantified by high-performance liquid chromatography (HPLC), and the data is stored as a triplet of pesticide name-process parameters-transformation form; 2) Integrate verified pesticide transformation pathway data in the literature, such as the oxidative metabolites and half-lives of pyrethroid pesticides under an 80°C steaming process; 3) Associate chemical structure information with toxicity data in public databases, such as mapping the CAS number of the transformation product with the ToxCast toxicity value; 4) Remove conflicting data through data cleaning (such as outliers with a conversion rate difference of >20% under the same conditions) and standardize the storage into a relational database table. The table structure includes pesticide name, process parameters (temperature, time), transformation form, conversion rate and data source label.

[0078] In step S3a, the metabolism-toxicity association network is constructed based on pre-stored biological network topology data, which is derived from the integrated data of the KEGG metabolic pathway database and the ToxCast toxicity database; the nodes include transformation forms and human metabolites, for example, the pesticide transformation form "phosphate A" is used as node A, and its metabolite "hydroxyphosphate" is used as node B; the edges represent the toxicity transmission path, for example, the edge between node A and node B is defined as the transformation path of phosphate A to hydroxyphosphate, and the edge between node B and the toxic target organ "liver cell" is defined as the toxic effect path of hydroxyphosphate on the liver; when the network is constructed, the relationship between nodes and edges is stored in the graph database Neo4j. The node attributes include chemical formula and toxicity level, and the edge attributes include reaction type and toxicity intensity. The reaction type is divided into hydrolysis, oxidation, and reduction. The toxicity intensity is divided into high (<10μM), medium (10-100μM), and low (>100μM) according to the IC50 value.

[0079] In step S3b, the extraction of node feature vectors is performed based on the biological network topology data, and the chemical descriptors of the transformed morphology are calculated using the open source chemical informatics toolkit RDKit. The chemical descriptors include molecular weight, lipid-water partition coefficient LogP, and number of hydrogen bond donors. For example, the molecular weight of phosphate A is 246.2 g / mol, the lipid-water partition coefficient LogP is 2.3, and the number of hydrogen bond donors is 2; the toxicity value of the metabolite is derived from the in vitro cytotoxicity experimental data of the ToxCast database, for example, the IC50 value of hydroxyphosphate is 10 μM; the feature vector is generated by splicing the chemical descriptor with the standardized numerical value of the toxicity value. For example, the chemical descriptor vector of phosphate A is [246.2, 2.3, 2], the toxicity value vector is

[10] , and the spliced ​​feature vector is [246.2, 2.3, 2, 10].

[0080] In step S3c, the multi-layer convolutional layers of the graph convolutional model are used to aggregate neighborhood node features. The first convolution operation multiplies the node feature vector by the adjacency matrix to generate a primary embedding representation. The adjacency matrix is ​​defined according to the connection relationship between the nodes. For example, if there is an edge between phosphate A and hydroxyphosphate, the corresponding matrix position value is 1, otherwise it is 0; the second convolution operation performs a nonlinear transformation on the primary embedding representation, and the activation function uses the ReLU function to generate a final node embedding representation with a dimension of 64. For example, the embedding representation of phosphate A is a floating-point vector with a length of 64.

[0081] In step S3d, the topological centrality score is calculated by the PageRank algorithm. The algorithm parameters are set according to the conventional settings in the literature, the damping coefficient is set to 0.85, and the number of iterations is 100. For example, the PageRank score of phosphate A is 0.032, and the PageRank score of hydroxyphosphate is 0.045; the toxicity weight value is generated by multiplying the topological centrality score by the toxicity value, for example, the toxicity weight value of hydroxyphosphate is 0.045×10=0.45; the final toxicity weight value is mapped to the range of 0 to 1 by maximum-minimum normalization, for example, the original weight value of phosphate A of 0.32 is normalized to 0.72.

[0082] In step S3e, the batch data index is composed of the batch code generated in step S1d and the pesticide name, for example, the batch code 20231001-GZB-003 and the pesticide name "chlorpyrifos" are combined to form 20231001-GZB-003-chlorpyrifos; the storage operation includes creating a toxicity weight table in the pesticide chemical transformation database, the fields of the toxicity weight table include the batch data index, the transformation form name, the toxicity weight value and the update timestamp, for example, the phosphate A toxicity weight value corresponding to the batch data index 20231001-GZB-003-chlorpyrifos is 0.72; the toxicity weight table is associated with the conversion rate table of step S2d through the batch data index, and the association operation is implemented by the SQL statement "SELECT * FROM toxicity weight table JOIN conversion rate table ON toxicity weight table.batch data index = conversion rate table.batch data index"; when storing, the toxicity weight values ​​under the same batch data index are sorted in ascending order by timestamp to form a time series data set.

[0083] In step S4a, the time series data of temperature, humidity and air flow velocity in the processing equipment are collected in real time by high-precision Internet of Things sensors deployed inside the equipment. The Internet of Things sensors include thermocouple temperature sensors, capacitive humidity sensors and ultrasonic airflow sensors; the thermocouple temperature sensor is installed in the heat source core area of ​​the processing equipment, for example, it is fixedly installed at the center point of the bottom of the frying pan, with a measurement range of 0 degrees Celsius to 300 degrees Celsius and an accuracy of ±0.5 degrees Celsius; the capacitive humidity sensor is distributed at multiple monitoring points in the medicinal material stacking area, for example, a sensor is set at the upper, middle and lower layers of the medicinal material pile, with a measurement range of 0% to 100% relative humidity and an accuracy of ±2%; the ultrasonic airflow sensor Fixed to the equipment vent, the measurement range is 0m / s to 10m / s, with an accuracy of ±0.1m / s. The collected micro-disturbance data is stored in association with the batch data index. The batch data index is composed of the batch code generated in step S1d and the pesticide name. For example, the batch code 20231001-GZB-003 and the pesticide name "chlorpyrifos" are combined into 20231001-GZB-003-chlorpyrifos. The storage format is a time series database record. Each record contains a timestamp, temperature value (unit: Celsius), humidity value (unit: percentage) and airflow velocity value (unit: meter / second). The data is transmitted to the central database through an encrypted protocol and stored in partitions according to the batch data index.

[0084] In step S4b, the phased analysis operation of the Lyapunov exponent is divided into the first phase of instant disturbance intensity quantification and the second phase of historical cumulative effect extraction; the first phase analysis is for the current batch of micro-perturbation data, and the Lyapunov exponent of the time series is calculated by the sliding window method, with a window size of 10 minutes and a step size of 1 minute. The window size is set according to the typical reaction time of the preparation process. For example, the pesticide conversion reaction time in the frying process is about 10 minutes, so the window size is set to 10 minutes; the Wolf algorithm is used in the calculation to estimate the maximum Lyapunov exponent by tracking the divergence rate of adjacent tracks. For example, in the micro-perturbation data with a temperature fluctuation variance of 2.5, the calculated Lyapunov exponent is 0.35, and the immediate nonlinear disturbance intensity is judged to be high; the second stage analysis extracts the cumulative disturbance effect of the same origin in the historical batch data, and calculates the disturbance intensity attenuation coefficient by the exponentially weighted moving average method. The number of historical batch data is set to the latest 5 batches according to data availability, and the attenuation coefficient is set according to the stability of the pesticide. For example, the attenuation coefficient of the pesticide with a half-life of less than 7 days is set to 0.8, and the pesticide with a half-life of more than 30 days is set to 0.5; the cumulative effect coefficient is calculated by the weighted average of the disturbance intensity of the historical batches multiplied by the attenuation coefficient. For example, the weighted average of the temperature fluctuation variance of the past 5 batches is 2.0, and the attenuation coefficient is 0.8, then the cumulative effect coefficient is 2.0×0.8=1.6.

[0085] In step S4c, the recursive neural network adopts a long short-term memory network (LSTM) structure, and the input layer receives the concatenation vector of the immediate nonlinear perturbation intensity and the cumulative effect coefficient, for example, the immediate perturbation intensity of 0.35 and the cumulative effect coefficient of 1.6 are concatenated as [0.35, 1.6]; two LSTM units are set in the hidden layer, and the time step is aligned with the timestamp of the micro-perturbation data, for example, one time step per minute, and a 2-hour preparation process corresponds to 120 time steps; the output layer generates a dynamic correction factor through the Sigmoid activation function, and the dynamic correction factor is a scalar value between 0 and 1, for example, the output value is 0.72; the synchronous association operation includes multiplying the dynamic correction factor with the conversion rate of step S2 and the toxicity weight value of step S3. For example, when the conversion rate is 60% and the toxicity weight value is 0.72, the correction factor 0.72 acts on the product of the two (60%×0.72=43.2%) to generate a comprehensive correction coefficient.

[0086] In step S4d, the two-way correction includes forward correction and reverse correction. Forward correction is for highly toxic form residues, and the form residues with toxicity weight values ​​higher than the preset threshold (for example, ≥0.5) are amplified by a dynamic correction factor. For example, the phosphate A residue with a toxicity weight value of 0.72 is corrected from 100 mg / kg to 100×0.72=72 mg / kg. Reverse correction is for low-confidence conversion rate errors, and the conversion rate deviation with medium confidence level in step S2 is suppressed by a dynamic correction factor. For example, when the confidence level is medium, the conversion rate deviation of 5% is reduced to 5%×0.72=3.6% by the correction factor of 0.72. The actual residue calculation formula is the total residue multiplied by the corrected conversion rate. For example, a total residue of 200 mg / kg is corrected to 200×43.2%=86.4 mg / kg. During the correction process, a balance check is performed on the amplification and suppression operations. For example, if the forward correction causes the residue to exceed the safety threshold, the manual review process is triggered.

[0087] In step S4e, the actual residue amount is associated and stored in the pesticide chemical conversion database according to the batch data index. The storage table structure includes the fields batch data index, pesticide name, actual residue amount, correction path type and update timestamp; the correction path type is marked according to the bidirectional correction rule, for example, the forward correction is marked as "high toxicity amplification" and the reverse correction is marked as "low confidence suppression"; the association operation connects the actual residue table with the conversion rate table of step S2d and the toxicity weight table of step S3e through an SQL statement, for example, executing "SELECT * FROM actual residue table JOIN conversion rate table ON actual residue table.batch data index = conversion rate table.batch data index JOIN toxicity weight table ON actual residue table.batch data index = toxicity weight table.batch data index"; when storing, a time series data set is generated by sorting by timestamp, for example, the correction data at 14:30:00 on October 1, 2023 and the data at the previous and next time points are arranged in sequence to form a continuous trend curve for cross-batch risk prediction in subsequent steps.

[0088] In step S5a, the system matches the pesticide name with a pre-stored pesticide toxicology database to determine the type of mechanism of action of the pesticide; the pesticide toxicology database is constructed based on the World Health Organization (WHO) pesticide classification standard, and the mechanism of action types are divided into neurotoxicity, metabolic toxicity and endocrine disruption; for example, the pesticide name "chlorpyrifos" corresponds to neurotoxicity, the pesticide name "cypermethrin" corresponds to metabolic toxicity, and the pesticide name "DDT" corresponds to endocrine disruption; the determination operation includes entering the pesticide name through the database query interface and returning the corresponding mechanism of action type label, for example, entering "chlorpyrifos" returns a "neurotoxicity" label, and entering "cypermethrin" returns a "metabolic toxicity" label.

[0089] In step S5b, the system queries the pre-stored synergistic effect rule library based on the combination of mechanism types to screen the synergistic effect patterns of the pesticide combination; the synergistic effect rule library is constructed by integrating known pesticide interaction research literature, and the synergistic effect patterns include competitive inhibition of metabolic pathways and superposition of target organ toxicity; for example, when the mechanism types of action of two pesticides are neurotoxicity and metabolic toxicity respectively, the "competitive inhibition of metabolic pathways" pattern is matched, which is defined as two pesticides competing for the active site of the same metabolic enzyme; when both pesticides are metabolic toxic, the "target organ toxicity superposition" pattern is matched, which is defined as the cumulative toxic effects of the two pesticides on the same target organ (such as the liver); the screening operation includes using the mechanism type of the pesticide combination as a joint key to query the rule library, for example, inputting "neurotoxicity + metabolic toxicity" returns the "competitive inhibition of metabolic pathways" pattern.

[0090] In step S5c, the system extracts metabolic path intersection nodes based on the metabolism-toxicity association network constructed in step S3a; the extraction method includes traversing the shared metabolic pathway nodes in the network, for example, the metabolic path of pesticide A is node A→node B→node C, and the metabolic path of pesticide B is node D→node B→node E, then the intersection node is node B; the competitive inhibition strength is calculated by the enzyme kinetic parameters of the intersection node, and the enzyme kinetic parameters include the inhibition constant Ki value, for example, the Ki value of pesticide A to node B is 10μM, and the Ki value of pesticide B to node B is 20μM; the competitive inhibition strength calculation formula is 1 / Ki_A+1 / Ki_B, that is, 1 / 10+1 / 20=0.15; the calculation result is normalized to the range of 0 to 1 in proportion, for example, the maximum competitive inhibition strength of 0.15 corresponds to the normalized value of 1.0, and the minimum strength 0 corresponds to 0.

[0091] In step S5d, the system constructs a dynamic attenuation model of the toxicity propagation path based on the competitive inhibition strength and the topological centrality score calculated in step S3d; the topological centrality score comes from the PageRank value of step S3d, for example, the PageRank value of the cross node B is 0.045; the attenuation coefficient calculation formula of the dynamic attenuation model is α=1 / (1+PR(v)), where PR(v) is the PageRank value of node v, for example, the attenuation coefficient of node B is 1 / (1+0.045)≈0.957; the comprehensive toxicity correction coefficient is calculated by multiplying the competitive inhibition strength and the attenuation coefficient, for example, if the competitive inhibition strength is 0.8 and the attenuation coefficient is 0.957, then the correction coefficient is 0.8×0.957≈0.766; the correction coefficient is mapped to the range of 0 to 1 by the Sigmoid function, for example, when 0.766 is input, 0.68 is output, and the mapping parameters are determined based on training of historical synergistic toxicity experimental data.

[0092] In step S5e, the system applies the comprehensive toxicity correction coefficient to the toxicity weight value of step S3 to generate an adjusted comprehensive toxicity weight; for example, if the original toxicity weight value is 0.72 and the correction coefficient is 0.68, the adjusted weight is 0.72×0.68≈0.49; the storage operation includes creating a comprehensive toxicity weight table in the pesticide chemical transformation database, the fields of which include batch data index, pesticide combination name, adjusted comprehensive toxicity weight, and update timestamp; for example, the adjusted weight corresponding to the batch data index 20231001-GZB-003-chlorpyrifos-cypermethrin is 0.49; the association operation connects the comprehensive toxicity weight table with the toxicity weight table of step S3e and the actual residue table of step S4e through an SQL statement, for example, executing "SELECT * FROM comprehensive toxicity weight table INNER JOIN toxicity weight table ON comprehensive toxicity weight table.batch data index = toxicity weight table.batch data index INNER JOIN "Actual Residue Table" ON "Comprehensive Toxicity Weight Table.Batch Data Index = Actual Residue Table.Batch Data Index"; when storing, sort the data in ascending order by timestamp to generate a time series dataset for cross-batch trend analysis in subsequent steps.

[0093] The determination of "the presence of two or more pesticide residues" in step S5 is based on the pesticide residue detection data from step S1, which includes the pesticide names and the corresponding total residue amounts. The determination operation includes traversing the number of pesticide names in the detection data, and triggering the synergistic effect correction process if the number is ≥2. For example, when the pesticide names in the detection data are "chlorpyrifos" and "cypermethrin", the number of pesticides is 2, and it is determined that two pesticide residues exist. The pesticide names and total residue amounts are derived from the standardized processing results of step S1a, and the data storage format is a "pesticide name-total residue amount" key-value pair, for example, "chlorpyrifos: 100 mg / kg, cypermethrin: 80 mg / kg".

[0094] Step S5 constructs a dynamic attenuation model to correct the toxicity weight value by combining the type of pesticide mechanism of action, the cross-node of the metabolic pathway and the topological centrality score, so as to solve the problem of evaluation bias caused by the traditional method ignoring the synergistic effect of multiple pesticides. Compared with the existing technology that only relies on a single mechanism of action or simple toxicity superposition, step S5 introduces a two-dimensional analysis of competitive inhibition of metabolic pathways and topological attenuation, which significantly improves the quantification accuracy of synergistic toxicity. For example, the enzyme competition intensity of the cross-node of the metabolic pathway is integrated with the network topological influence to accurately identify the toxic form with high transmission risk and avoid the underestimate of high toxic residues; the dynamic attenuation model is combined with historical data training to adaptively correct the weight distribution and reduce the error of low confidence conversion rate. Step S5 dynamically associates the topological characteristics of the metabolic network with the synergistic effect, realizing the qualitative change of risk assessment from "single linear" to "multi-dimensional coupling", which is in line with the complex toxic interaction characteristics of the processing technology of Chinese medicinal materials.

[0095] In step S6a, the system constructs a time series data set of risk index based on the actual residue amount in step S4e and the adjusted comprehensive toxicity weight in step S5e; the actual residue amount is read from the actual residue amount table of the pesticide chemical transformation database, and the adjusted comprehensive toxicity weight is read from the comprehensive toxicity weight table; the risk index is generated by multiplying the actual residue amount by the adjusted comprehensive toxicity weight; for example, when the actual residue amount is 86.4 mg / kg and the comprehensive toxicity weight is 0.49, the risk index is 86.4×0.49≈42.3; the time series data set fields include batch data index, pesticide combination name, risk index and timestamp, for example, the risk index corresponding to the batch data index 20231001-GZB-003-chlorpyrifos-chlorpermethrin is 42.3; the data set is sorted by timestamp and stored in the time series database, and the timestamp granularity is accurate to minutes, for example, the risk index 42.3 is recorded at 14:30:00 on October 1, 2023.

[0096] In step S6b, the system combines the time series data set of historical batch data and extracts cross-batch risk trend characteristics through grey correlation analysis; the time range of historical batch data is the most recent 12 months, and the data is retrieved from the risk index table of the pesticide chemical transformation database; grey correlation analysis is implemented by calculating the correlation between the current batch and the historical batch risk index sequence, for example, the current batch risk index sequence [42.3, 45.1, 40.8] and the historical batch sequence [38.5, The correlation between the two series is 0.85; the correlation calculation formula is the sum of the reciprocals of the absolute values ​​of the differences between the two series at each time point, normalized to the range of 0 to 1. For example, when the differences are 3.8, 3.9, and 1.1, the sum of the reciprocals is 1 / 3.8+1 / 3.9+1 / 1.1≈1.43, which is 1.43 / 3≈0.48 after normalization. Risk trend characteristics include correlation threshold, trend slope, and fluctuation variance. For example, a correlation ≥ 0.7 is judged to be a high-risk trend, and a positive trend slope indicates an increase in risk.

[0097] In step S6c, the system generates a dynamic risk warning model based on the cross-batch risk trend characteristics and the node topology influence (i.e., PageRank value) of step S3d; the node topology influence is read from the topological centrality score table of the metabolism-toxicity association network, for example, the PageRank value of the cross node B is 0.045; the model input is the concatenation vector of the risk trend characteristics and the node influence, for example, the correlation degree 0.85, the trend slope 0.12, the volatility variance 1.5 and the PageRank value 0.045 are concatenated to [0.85, 0.12, 1.5, 0.045]; the model fits the historical risk data and the current trend characteristics through linear regression, and the weight parameters are optimized according to the least squares method, for example, the weight vector is [0.6, 0.3, 0.1, 0.2]; the output is the risk warning level, which is divided into three levels: low, medium, and high. For example, the output value 0.72 corresponds to high risk.

[0098] In step S6d, the system associates the output results of the dynamic risk warning model with the pesticide chemical conversion database according to the batch data index. The storage table structure includes the batch data index, pesticide combination name, warning level, and update timestamp. For example, the warning level corresponding to the batch data index 20231001-GZB-003-chlorpyrifos-cypermethrin is high risk. The cross-batch risk prediction report is generated through an SQL joint query, for example, executing "SELECT batch data index, warning level FROM risk warning table WHERE warning level = 'high risk'" and associating it with the actual residue table in step S4e. The report format is a structured table, including the batch code, pesticide name, risk index, and warning level. For example, the table records "20231001-GZB-003-chlorpyrifos-cypermethrin, high risk, 42.3".

[0099] In step S6e, the system updates the parameters of the dynamic risk warning model based on the real-time batch data, and the real-time data is collected and updated by the sensors in step S4a; the parameter adjustment adopts the sliding window method, and the window size is the data of the last three months. For example, when the proportion of high-risk batches in the window exceeds 20%, the risk level threshold is lowered from 0.7 to 0.6; the adjustment operation includes recalculating the grey correlation weight and regression coefficient, for example, the original weight vector [0.6, 0.3, 0.1, 0.2] is adjusted to [0.65, 0.25, 0.1, 0.15]; the updated parameters are stored by timestamp version, for example, version number 20231001-1 corresponds to the first adjustment on October 1, 2023.

[0100] Example 2: Figure 2 The present invention provides a schematic structural diagram of a risk assessment system for pesticide residues in traditional Chinese medicines. The risk assessment system for pesticide residues in traditional Chinese medicines includes the following modules:

[0101] Data collection module: obtains pesticide residue test data including pesticide names and total residue amounts in the Chinese medicinal materials to be evaluated; extracts processing parameters including processing temperature and duration;

[0102] Conversion analysis module: associates the pre-stored pesticide chemical conversion database with the pesticide name and preparation process parameters to determine the pesticide's conversion form and corresponding conversion rate under the preparation process;

[0103] Weight generation module: Based on the pre-stored biological network topology data, the node influence of the transformation morphology is calculated through the graph convolution model to generate the toxicity weight value;

[0104] Dynamic correction module: Based on the micro-disturbance data of the preparation environment collected by sensors, the impact of nonlinear disturbances on the conversion rate is quantified through chaos theory, and the actual residual amount is dynamically corrected in combination with a recursive neural network;

[0105] Synergistic adjustment module: If there are two or more pesticide residues, the toxicity weight value is dynamically adjusted according to the synergistic effect relationship associated with the mechanism of action type;

[0106] Prediction and assessment module: Generates risk assessment results based on actual residues and adjusted toxicity weight values, and performs cross-batch risk prediction in combination with historical batch data.

[0107] The calculations involved in the embodiments are all dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to actual conditions.

[0108] It should be noted that the present invention can be deployed on the device itself to implement embedded applications, and can also be run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.

[0109] The above embodiments can be implemented in whole or in part via software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions or computer programs. When loaded or executed on a computer, the processes or functions described in the embodiments of this application are fully or partially performed. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired means (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium accessible by a computer or a data storage device such as a server or data center that contains a collection of one or more available media. The available medium can be magnetic media (e.g., floppy disks, hard disks, tapes), optical media (e.g., DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0110] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0111] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0112] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed across multiple network modules. Some or all of the modules may be selected to achieve the purpose of this embodiment according to actual needs.

[0113] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0114] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0115] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

[0116] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A risk assessment method for pesticide residues in traditional Chinese medicines, characterized in that: The steps include: S1. Obtain pesticide residue test data including pesticide names and total residue amounts in the Chinese medicinal materials to be evaluated; extract processing parameters including processing temperature and duration; S2. Associating a pre-stored pesticide chemical transformation database with the pesticide name and preparation process parameters to determine the transformation form and corresponding conversion rate of the pesticide under the preparation process; S3. Based on the pre-stored biological network topology data, the node influence of the transformation morphology is calculated through the graph convolution model to generate the toxicity weight value, including: S3a, construct a metabolism-toxicity association network, where nodes are transformation forms and human metabolites, and edges are toxicity transmission pathways; S3b, extracting node feature vectors based on biological network topology data, including chemical descriptors of transformation forms and toxicity values ​​of metabolites; S3c, aggregate neighborhood node features through the multi-layer convolutional layers of the graph convolutional model to generate node embedding representation; S3d, calculate the topological centrality score based on the node embedding representation, and generate the toxicity weight value by combining the toxicity value; S3e, storing the toxicity weight value in a pesticide chemical transformation database according to the batch data index; S4. Based on the micro-disturbance data of the preparation environment collected by sensors, the impact of nonlinear disturbances on the conversion rate is quantified through chaos theory, and the actual residual amount is dynamically corrected in combination with a recursive neural network, including: S4a, collecting time series data of temperature, humidity, and airflow velocity in the processing equipment as micro-disturbance data of the processing environment, and storing them in association with batch data index; S4b, based on the Lyapunov exponent, the chaotic characteristics of the perturbation data are analyzed in stages. The first stage quantifies the intensity of the immediate nonlinear perturbation, and the second stage extracts the cumulative effect coefficient of the perturbation of historical batches. S4c, fusing the instantaneous nonlinear perturbation intensity and the perturbation cumulative effect coefficient through a recursive neural network to generate a dynamic correction factor, synchronously correlating the conversion rate of step S2 with the toxicity weight value of step S3; S4d, performing bidirectional correction on the total residue based on the dynamic correction factor to generate the actual residue; S4e, storing the actual residue amount in the pesticide chemical transformation database according to the batch data index, and marking the correction path type; S5. If two or more pesticide residues are present, the toxicity weight value is dynamically adjusted based on the synergistic effect relationship associated with the mechanism of action type; S6. Generate risk assessment results based on actual residues and adjusted toxicity weight values, and perform cross-batch risk prediction in combination with historical batch data.

2. A risk assessment method for pesticide residues in traditional Chinese medicines according to claim 1, characterized in that: S1 includes: S1 a. Obtain pesticide residue detection data in the Chinese medicinal material planting area through IoT sensors. The pesticide residue detection data includes the pesticide name and the corresponding total residue amount; S1 b, matching the pre-stored pesticide detection data cleaning rule library according to the pesticide name, and performing outlier removal and standardization on the pesticide residue detection data; S1 c, extracting the time series data of processing temperature and duration from the processing equipment control system to generate processing parameters; S1 d. According to the pesticide name and processing parameters, associate the production batch code of the Chinese medicinal materials and establish a batch data index.

3. A risk assessment method for pesticide residues in traditional Chinese medicine according to claim 2, characterized in that S2 include: S2a, matching the reaction pathway templates in the pesticide chemical transformation database based on the pesticide name and processing temperature, and screening the candidate transformation forms corresponding to the preparation process; S2b, calculating the conversion rate of each conversion form using a conditional probability model based on the duration and kinetic parameters of the candidate conversion forms; S2c, based on experimental verification data in the pesticide chemical transformation database associated with the preparation stage labels, the conversion rate is corrected with confidence; S2d. Store the corrected transformation form and conversion rate into the pesticide chemical transformation database according to the batch data index.

4. A risk assessment method for pesticide residues in traditional Chinese medicines according to claim 3, characterized in that: Bidirectional correction includes: forward correction to amplify the residue of highly toxic forms based on dynamic factors, and reverse correction to suppress low-confidence conversion rate errors.

5. The risk assessment method for pesticide residues in traditional Chinese medicine according to claim 4, characterized in that S5 include: S5a. If two or more pesticide residues are present, determine the type of pesticide action mechanism based on the pesticide name; S5b, based on the matching of the mechanism of action type with the pre-existing synergistic effect rule library, to screen the synergistic effect pattern of the pesticide combination; S5c, extracting metabolic pathway intersection nodes based on the metabolism-toxicity association network of step S3, and quantifying the competitive inhibition strength of the pesticide combination on the intersection nodes; S5d, based on the competitive inhibition strength and the topological centrality score of step S3, construct a dynamic attenuation model of the toxicity transmission path and generate a comprehensive toxicity correction coefficient; S5e: Apply the comprehensive toxicity correction coefficient to the toxicity weight value of step S3 to generate an adjusted comprehensive toxicity weight, and store it in the pesticide chemical transformation database according to the batch data index.

6. A risk assessment method for pesticide residues in traditional Chinese medicines according to claim 5, characterized in that: The synergistic mode of action includes competitive inhibition of metabolic pathways and additive target organ toxicity.

7. The risk assessment method for pesticide residues in traditional Chinese medicine according to claim 5, wherein S6 include: S6a, constructing a time series data set of risk index based on the actual residue and the comprehensive toxicity weight adjusted in step S5e; S6b, combining the time series dataset of historical batch data, extracting cross-batch risk trend characteristics through grey correlation analysis; S6c. Generate a dynamic risk warning model based on the node influence of cross-batch risk trend characteristics and transformation forms; S6d. The output results of the dynamic risk warning model are stored in association with the batch data index to generate a cross-batch risk prediction report; S6e. Update the parameters of the dynamic risk warning model according to the real-time batch data and adjust the risk warning level threshold.

8. A risk assessment system for pesticide residues in traditional Chinese medicines, used to implement the risk assessment method for pesticide residues in traditional Chinese medicines according to any one of claims 1 to 7, characterized in that: Includes the following modules: Data collection module: obtains pesticide residue test data including pesticide names and total residue amounts in the Chinese medicinal materials to be evaluated; extracts processing parameters including processing temperature and duration; Conversion analysis module: associates the pre-stored pesticide chemical conversion database with the pesticide name and preparation process parameters to determine the pesticide's conversion form and corresponding conversion rate under the preparation process; Weight generation module: Based on the pre-stored biological network topology data, the node influence of the transformation morphology is calculated through the graph convolution model to generate the toxicity weight value; Dynamic correction module: Based on the micro-disturbance data of the preparation environment collected by sensors, the impact of nonlinear disturbances on the conversion rate is quantified through chaos theory, and the actual residual amount is dynamically corrected in combination with a recursive neural network; Synergistic adjustment module: If there are two or more pesticide residues, the toxicity weight value is dynamically adjusted according to the synergistic effect relationship associated with the mechanism of action type; Prediction and assessment module: Generates risk assessment results based on actual residues and adjusted toxicity weight values, and performs cross-batch risk prediction in combination with historical batch data.

Images