Veterinary drug residual risk data analysis method and system

Abstract

The invention relates to the technical field of veterinary drug residue analysis, and discloses a veterinary drug residue risk data analysis method and system, and the method comprises the steps: carrying out the initial multi-dimensional detection of an animal, and obtaining multi-dimensional data; the current state of the animal is analyzed according to the multi-dimensional data, and veterinary drug residues in the current state are calculated; and updating the metabolic velocity by using the optimized iterative algorithm to evaluate the veterinary drug residue risk. The updated multi-dimensional data is obtained by carrying out subsequent multi-dimensional detection; and adjusting the iterative algorithm by using the updated multi-dimensional data, and evaluating the veterinary drug residue risk according to the adjusted iterative algorithm. And the metabolic rate can be dynamically adjusted and optimized, the veterinary drug residue condition can be tracked in real time, a scientific basis is provided for veterinary drug safety supervision, and the risk analysis capability in the animal breeding process is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of veterinary drug residue analysis, and specifically to a method and system for data analysis of veterinary drug residue risks. Background Art

[0002] The problem of veterinary drug residues has always been an important issue in the field of animal food safety. With the modernization of agricultural production, the widespread use of veterinary drugs has not only improved animal health and breeding efficiency but also brought potential threats to food safety and human health due to veterinary drug residues. Veterinary drug residues not only affect animal health but may also enter the human body through the food chain, endangering human health. Therefore, how to accurately evaluate and predict the residues of veterinary drugs in animals has become the key to ensuring food safety.

[0003] Currently, the detection of veterinary drug residues mainly relies on the detection of samples such as blood, urine, or milk at a single time point, but this method often fails to comprehensively reflect the metabolic process and residue dynamics of veterinary drugs in animals. Existing technologies cannot evaluate the metabolic rate of veterinary drugs and their residue contents in different organs and tissues in real time and accurately. Therefore, there is an urgent need for an analytical method that can evaluate veterinary drug residue risks in a multi-dimensional and dynamic manner. Summary of the Invention

[0004] In view of the above problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is the optimization problem that the existing data analysis method for veterinary drug residue risks cannot dynamically adjust the metabolic rate and lacks accurate and real-time evaluation of veterinary drug residues.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: A method for data analysis of veterinary drug residue risks, comprising:

[0007] Obtaining multi-dimensional data by performing an initial multi-dimensional detection on an animal;

[0008] Analyzing the current state of the animal based on the multi-dimensional data and calculating the veterinary drug residues in the current state;

[0009] Updating the metabolic rate by using an optimized iterative algorithm to evaluate the veterinary drug residue risks;

[0010] Obtaining updated multi-dimensional data by performing subsequent multi-dimensional detections;

[0011] Adjusting the iterative algorithm by using the updated multi-dimensional data and evaluating the veterinary drug residue risks according to the adjusted iterative algorithm;

[0012] The process of adjusting the iterative algorithm includes using the metabolic rate after the last update to predict the multi-dimensional data at the time of this detection, obtaining the predicted value of the multi-dimensional data for this time; comparing and analyzing the predicted value of the multi-dimensional data for this time with the updated multi-dimensional data, and obtaining the metabolic rate updated for this time through the iterative algorithm.

[0013] As a preferred embodiment of the veterinary drug residue risk data analysis method of the present invention, wherein: the multi-dimensional detection includes blood sample detection, urine sample detection, milk or egg liquid detection.

[0014] As a preferred embodiment of the veterinary drug residue risk data analysis method of the present invention, wherein: the veterinary drug residues in the current state include obtaining the veterinary drug concentration XY0 in the blood, the veterinary drug concentration NY0 in the urine, and the veterinary drug concentration RY0 in the milk or egg liquid according to the results of the blood sample detection, the urine sample detection, and the milk or egg liquid detection, forming a detection result set {XY0, NY0, RY0};

[0015] Obtain the target object category to be analyzed, and through the category comparison table, obtain the average metabolic rate of each organ and tissue of the target object category at each age, and the average growth rate of each age.

[0016] Use the trained Bayesian algorithm 1, input: the weight of the target object, the average metabolic rate of each organ and tissue at the current age, the set {XY0, NY0, RY0}; output: the residual content of veterinary drugs in each organ or tissue.

[0017] As a preferred embodiment of the veterinary drug residue risk data analysis method of the present invention, wherein: the optimized iterative algorithm includes modeling the process of veterinary drug residues, and starting the risk assessment process after the second multi-dimensional detection at different times;

[0018] And according to the latest detection results, iterate the metabolic rate of each organ and tissue to obtain the optimized metabolic rate; use the optimized metabolic rate to simulate the metabolic process, and finally generate the evaluation result of veterinary drug residue risk.

[0019] As a preferred embodiment of the veterinary drug residue risk data analysis method of the present invention, wherein: the modeling of the process of veterinary drug residues includes building a digital twin model with the time axis as the core, and predicting the weight of the target object at different moments on the time axis by setting the daily feed supply.

[0020] Suppose that at time t, the weight measurement and multi-dimensional detection are carried out. While obtaining the current actual weight m, predict the reference value of the current weight: m' = Δm t+m0; where, Δm t represents the change in body weight from the previous measurement to the current time t; m0 represents the body weight at the previous measurement;

[0021] According to the feeding time of the veterinary drug, the veterinary drug is input on the time axis, and the content of the veterinary drug in each organ at time t is predicted based on the input amount: obtained by subtracting the metabolism amount from the cumulative amount of the veterinary drug; among them, when calculating the metabolism amount, it is completed by analyzing the content and half-life of the veterinary drug.

[0022] As a preferred solution of the method for analyzing veterinary drug residue risk data according to the present invention, wherein: the iteration of the metabolism rate of each organ and tissue includes, according to the predicted content of the veterinary drug in each organ at time t and the average metabolism rate of each organ and tissue, using the trained Bayesian algorithm 2 to predict the multi-dimensional detection results at time t to obtain the predicted veterinary drug concentration XY' in the blood u 、the predicted veterinary drug concentration NY' in the urine u 、the predicted veterinary drug concentration RY' in the milk or egg liquid u , constituting the predicted result set {XY' u 、NY' u 、RY' u};

[0023] Through multi-dimensional detection at time t, the actual veterinary drug concentration XY in the blood is obtained u 、the actual veterinary drug concentration NY in the urine u 、the actual veterinary drug concentration RY in the milk or egg liquid u , constituting the detection result set {XY u 、NY u 、RY u};

[0024] Compare the sequences XY' u with XY u 、NY' u with NY u 、RY' u with RY u one by one. When the deviation of any corresponding element exceeds the corresponding preset value, the metabolism rate k of the veterinary drug in each age group on the organ or tissue e dxcs，e is updated:

[0025] The step size of the change ratio of the preset metabolism rate of the veterinary drug on e is σ e ;

[0026] For each organ or tissue of the target object, update in any growth direction, and the updated metabolism rate is k dxcs，e ;

[0027] Using the metabolic rates of each updated organ or tissue, re-predict the content of veterinary drugs in each organ at time t, and based on the prediction results and the updated metabolic rates of each organ and tissue, use the trained Bayesian algorithm 2 to predict the multi-dimensional detection results at time t, obtaining the predicted veterinary drug concentration XY' in the blood u1 and the predicted veterinary drug concentration NY' in the urine u1 and the predicted veterinary drug concentration RY' in the milk or egg liquid u1 , forming the predicted result set of the detection {XY' u1 , NY' u1 , RY' u1};

[0028] According to the step size of the change ratio of the metabolic rate of each organ or tissue being σ e , and the influence direction of each organ or tissue on the body weight, and the proportion of each organ or tissue in the target object, update the body weight reference value m' after the metabolic rate is updated, obtaining the updated body weight m'0;

[0029] For the sequences XY' u1 and XY u , NY' u1 and NY u , RY' u1 and RY u , and the updated m'0 and m, conduct pairwise comparisons one by one. When the deviation of any corresponding element decreases compared to before the update, confirm the current metabolic rate of each organ or tissue. If the deviation of any corresponding element is less than the corresponding preset value, end the iteration and output the updated metabolic rate of each organ or tissue; if the deviation of any corresponding element is not less than the corresponding preset value, on the basis of the update, repeat the process of updating the metabolic rate.

[0030] As a preferred solution of the veterinary drug residue risk data analysis method described in the present invention, wherein: the assessment of the veterinary drug residue risk includes, for the prediction time P, using the metabolic rates of each organ or tissue at the end of the most recent iteration process to predict the content of veterinary drugs in each organ, and taking the prediction result as the final analysis result of the veterinary drug residue risk.

[0031] A veterinary drug residue risk data analysis system adopting the method described in the present invention, wherein:

[0032] A collection unit, by performing an initial multi-dimensional detection on an animal, obtains multi-dimensional data; by performing subsequent multi-dimensional detections, obtains updated multi-dimensional data;

[0033] An analysis unit, analyzes the current state of the animal according to the multi-dimensional data, and calculates the veterinary drug residue in the current state;

[0034] An update unit uses an optimized iterative algorithm to update the metabolic rate, implement the assessment of veterinary drug residue risk; uses the updated multi-dimensional data to adjust the iterative algorithm, and based on the adjusted iterative algorithm, assess the veterinary drug residue risk.

[0035] A computer device includes: a memory and a processor; the memory stores a computer program, wherein: when the processor executes the computer program, the steps of the method described in any one of the present inventions are implemented.

[0036] A computer-readable storage medium stores a computer program thereon, wherein: when the computer program is executed by a processor, the steps of the method described in any one of the present inventions are implemented.

[0037] Advantages of the present invention: Firstly, the method for analyzing veterinary drug residue risk data provided by the present invention improves the prediction accuracy of veterinary drug residues through comprehensive analysis of multi-dimensional data such as blood, urine, milk, or egg liquid. Secondly, by using a dynamically adjusted metabolic rate model, the risk assessment can adapt to the veterinary drug metabolism characteristics in animals of different ages and growth stages, realizing personalized assessment. Thirdly, continuous iterative optimization can provide the most suitable assessment results at different time points, reducing the deviation caused by fluctuations in detection data. In addition, the present invention can dynamically adjust and optimize the metabolic rate, which helps to track the situation of veterinary drug residues in real time, provides a scientific basis for veterinary drug safety supervision, and improves the risk analysis ability in the animal breeding process. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for description in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0039] Figure 1 It is the overall flowchart of a method for analyzing veterinary drug residue risk data provided by the first embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] To make the above objects, features, and advantages of the present invention more obvious and understandable, the specific embodiments of the present invention will be described in detail below with reference to the drawings of the specification. Obviously, the described embodiments are some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0041] Example 1, refer to Figure 1, which is an embodiment of the present invention, provides a method for analyzing veterinary drug residue risks, including:

[0042] S1: Obtain multi-dimensional data by performing initial multi-dimensional detection on animals;

[0043] Furthermore, the multi-dimensional detection includes blood sample detection, urine sample detection, milk or egg liquid detection.

[0044] Obtain the basic physiological data of animals through initial multi-dimensional detection, including the veterinary drug residue concentration information in blood, urine, milk or egg liquid, and provide basic data for subsequent veterinary drug residue risk assessment. These data can reflect the drug metabolism conditions of different organs or tissues in the current state of animals, help build a more accurate metabolic model, and thus improve the accuracy and real-time performance of risk assessment.

[0045] Collect blood samples of animals, and use high-sensitivity analysis methods such as liquid chromatography-tandem mass spectrometry (LC-MS / MS) and gas chromatography-tandem mass spectrometry (GC-MS / MS) to determine the veterinary drug residues and their concentrations in the blood. These methods can quantitatively analyze the contents of different veterinary drugs and their metabolites, so as to evaluate the residue level of veterinary drugs in the blood.

[0046] Collect urine samples of animals, and determine the veterinary drug residue components in the urine through chemical analysis methods similar to blood detection, such as liquid chromatography-tandem mass spectrometry (LC-MS / MS). Urine is one of the main excretion pathways of drug metabolism. Through urine detection, the drug metabolism rate and the excretion situation of metabolites can be reflected.

[0047] Collect milk or egg liquid samples, and use the same high-sensitivity detection methods, such as liquid chromatography-tandem mass spectrometry (LC-MS / MS), to analyze the veterinary drug components therein.

[0048] S2: Analyze the current state of the animal according to the multi-dimensional data, and calculate the veterinary drug residue in the current state.

[0049] According to the results of the blood sample detection, the urine sample detection, and the milk or egg liquid detection, obtain the veterinary drug concentration XY0 in the blood, the veterinary drug concentration NY0 in the urine, and the veterinary drug concentration RY0 in the milk or egg liquid, and form a detection result set {XY0, NY0, RY0}.

[0050] Obtain the target object category to be analyzed, and through the category comparison table, obtain the average metabolic rate of each organ and tissue of the target object category at each age, and the average growth rate at each age.

[0051] Using the trained Bayesian algorithm 1, the Bayesian algorithm can infer the probability distribution of the target variable (such as the residual content of veterinary drugs in various organs or tissues) through known prior information and observed data. Input: the weight of the target object, the average metabolic rate of each organ and tissue at the current age, the set {XY0, NY0, RY0}; Output: the residual content of veterinary drugs in various organs or tissues.

[0052] Among them, the "category comparison table" refers to a reference table or database that contains different animal breeds, species (such as cattle, sheep, chickens, pigs, etc.) and the metabolic characteristics of each category at different ages. This table not only contains the metabolic rate data of each animal category (such as the drug metabolism rate of organs and tissues), but may also include physiological data related to the growth and development of this category (such as the weight gain rate, etc.). The role of the category comparison table is to obtain the metabolism and growth rate of a specific category at a specific age according to the target animal category, providing an accurate physiological reference basis for the subsequent evaluation of veterinary drug residues.

[0053] S3: Using the optimized iterative algorithm, update the metabolic rate to evaluate the risk of veterinary drug residues.

[0054] Furthermore, the optimized iterative algorithm includes modeling the process of veterinary drug residues. After the second multi-dimensional detection at different times, the risk assessment process starts. And according to the latest detection results, iterate the metabolic rate of each organ and tissue to obtain the optimized metabolic rate; use the optimized metabolic rate to simulate the metabolic process, and finally generate the evaluation result of the risk of veterinary drug residues.

[0055] It should be noted that the metabolic rate of animals will change over time and due to various factors (such as age, weight, health status, etc.). Therefore, through the iteration of multi-dimensional detection and risk assessment, the metabolic rate can be adjusted in real time according to new detection data. The iterative algorithm can ensure that each update takes into account the latest detection results, thus accurately reflecting the drug metabolism process in animals. The residual level of veterinary drugs in animals is affected by multiple factors, and the metabolic rate is one of the key factors. By continuously adjusting and optimizing the metabolic rate, the drug metabolism process can be accurately simulated, and finally help to generate a more accurate evaluation result of the risk of veterinary drug residues. This dynamic and iterative evaluation method can help identify possible risks and potential problems, and then take timely measures to avoid unnecessary risks of veterinary drug residues.

[0056] After the first and second multi-dimensional detections, by continuously iterating and optimizing the metabolic rate, the metabolic process of veterinary drugs in animals at different time points and under different conditions can be better simulated, further improving the prediction ability. Through simulation, more accurate risk analysis can be provided for relevant decisions, predicting the residue levels of veterinary drugs in various organs and tissues, thereby achieving more effective management and control. The iterative algorithm can also capture various physiological and metabolic effects that change over time. As animals grow, diseases change, or the external environment varies, the metabolic rate and the risk of drug residues will all be different. Therefore, continuous iteration can make the evaluation results more timely and dynamically adjust the prediction results according to the physiological development of the animals.

[0057] Furthermore, the modeling of the veterinary drug residue process includes building a digital twin model with the time axis as the core, predicting the weight of the target object at different moments on the time axis by setting the daily feed supply. Considering that the feeding time of veterinary drugs will affect their distribution and residue levels in the body, it is necessary to combine the veterinary drug feeding time with the animal's metabolic process for simulation. By predicting the input amount on the time axis and combining the relationship between the cumulative amount and the metabolic amount, the content of veterinary drugs in different organs at each moment can be evaluated more accurately. Let's assume that at time t, body weight measurement and multi-dimensional detection are carried out. While obtaining the current actual body weight m, the reference value of the current body weight is predicted: m' = Δm t + m0.

[0058] where, Δm t represents the change in body weight from the last measurement to the current moment t; m0 represents the body weight at the last measurement.

[0059]

[0060] where, Δm t represents the change in body weight from the last measurement to the current moment t; di represents the integration with respect to time; t0 represents the time of the last body weight measurement and multi-dimensional detection; t represents the current moment; S i0 represents the daily standard feed supply at the age corresponding to time i; ρ i represents the growth rate of the target object at the age corresponding to time i; S i represents the feed supply on the current day corresponding to time i.

[0061] According to the feeding time of veterinary drugs, the veterinary drugs are input on the time axis, and the content of veterinary drugs in each organ at time t is predicted according to the input amount: obtained by subtracting the metabolism amount from the cumulative amount of veterinary drugs (the metabolism amount is calculated through the half-life of veterinary drugs, considering the degradation process of veterinary drugs in the body, which can more accurately simulate the metabolism process of veterinary drugs, and adjust the drug residues in different organs over time); among them, when calculating the metabolism amount, it is completed by analyzing the content and half-life of veterinary drugs.

[0062]

[0063] Among them, e represents the index of the organ or tissue; HL e，t represents the predicted content of veterinary drugs on e at time t; j represents the index of the number of times of veterinary drug input after the last detection; J represents the total number of times of veterinary drug input between the last detection and time t; β represents the absorption rate of veterinary drugs on e; TR j represents the amount of veterinary drugs input for the jth time; α i represents the control parameter corresponding to time i, taking 0 or 1; D e，i represents the metabolic rate of the organ or tissue e in the target object corresponding to age group at time i; TR e，0 represents the content of veterinary drugs on e at the last detection; y g represents the index value of the first time of veterinary drug input after the last detection of the veterinary drug content; y represents the index of veterinary drug input; C y，0 represents the content of veterinary drugs when the veterinary drugs are input for the yth time; k dxcs，e represents the metabolic rate of veterinary drugs on e (which is a variable. On the time axis, according to different age groups, k dxcs is dynamically changed accordingly; when entering the next age group, it is automatically adjusted to the corresponding value); q represents the duration of metabolism between the yth time of veterinary drug input and the (y - 1)th time of veterinary drug input.

[0064] It should be noted that "dynamically changing k dxcs accordingly" takes into account that the metabolic rate is related to the physiological state and age group of the animal. The younger the age, the faster the metabolism may be, and as the age increases, the metabolic rate of the organ may change. To achieve this, the "dynamic change" mentioned in the design is achieved by continuously updating the metabolic rate (such as according to the average metabolic rate of different age groups). At each age stage, the metabolic rate will be adjusted according to the physiological condition and growth rate of the target object. By setting the metabolic rate as a variable and automatically adjusting the value of the metabolic rate according to the age group of the animal, it can ensure that the metabolic process in different physiological stages is reasonably simulated. Whenever the animal enters a new age stage, the system will automatically adjust the metabolic rate to ensure that it reflects the actual physiological characteristics of the animal. In fact, k dxcsIt changes based on the time axis and is actually an accumulation manifestation of time.

[0065] According to the predicted content of veterinary drugs in each organ at time t and the average metabolic rates of each organ and tissue, using the trained Bayesian algorithm 2, predict the multi-dimensional detection results at time t to obtain the predicted veterinary drug concentration XY' in the blood u , the predicted veterinary drug concentration NY' in the urine u , and the predicted veterinary drug concentration RY' in the milk or egg liquid u , which constitute the predicted result set of the detection {XY' u , NY' u , RY' u}.

[0066] It should be noted that the current Bayesian algorithm focuses on the veterinary drug concentration in external samples (such as blood, urine, milk or egg liquid, etc.), and these data are usually used for detection and monitoring, especially for subsequent risk assessment and compliance inspection. Using Bayesian algorithm 2 to obtain {XY' u , NY' u , RY' u}, rather than using Bayesian algorithm 1 to obtain the residue situation, is because for the same steps, the quantity obtained by Bayesian algorithm 2 is 3. That is to say, in the iterative process, only these three quantities need to be compared, which can optimize the calculation amount. If the veterinary drug content in each organ or tissue is analyzed, then the quantity to be compared is too large, and due to too many comparison parameters, the iterative process will enter an infinite loop. By transferring the reference object, the iterative process can be effectively simplified.

[0067] Through multi-dimensional detection at time t, obtain the actual veterinary drug concentration XY in the blood u , the actual veterinary drug concentration NY in the urine u , and the actual veterinary drug concentration RY in the milk or egg liquid u , which constitute the detection result set {XY u , NY u , RY u}.

[0068] Compare the sequences XY' u with XY u , NY' u with NY u , and RY' u with RY u one by one. When the deviation of any corresponding element exceeds the corresponding preset value, update the metabolic rate k dxcs，e of the veterinary drug in the organ or tissue e at each age:

[0069] The preset veterinary drug is at e, and the step size of the change ratio of the metabolism rate is σ e .

[0070] For each organ or tissue of the target object, update it in any growth direction, and the updated metabolism rate is k dxcs，e (1 + σ e ).

[0071] Using the metabolism rates of each updated organ or tissue, re-predict the content of the veterinary drug in each organ at time t, and based on the prediction results and the updated metabolism rates of each organ and tissue, use the trained Bayesian algorithm 2 to predict the multi-dimensional detection results at time t to obtain the predicted veterinary drug concentration XY' in the blood u1 , the predicted veterinary drug concentration NY' in the urine u1 , the predicted veterinary drug concentration RY' in the milk or egg liquid u1 , which constitute the predicted result set of the detection {XY' u1 , NY' u1 , RY' u1}.

[0072] According to the step size of the change ratio of the metabolism rate of each organ or tissue being σ e , as well as the influence direction of each organ or tissue on the body weight and the proportion of each organ or tissue in the target object, update the body weight reference value m' after the metabolism rate is updated to obtain the updated body weight m'0

[0073]

[0074] Among them, E represents the number of organs or tissues; θ e represents the influence direction of organ or tissue e on the body weight, taking 1 or 0 or -1; δ e represents the proportion of organ or tissue e in the target object

[0075] Compare the sequence XY' u1 with XY u , NY' u1 with NY u , RY' u1 with RY u , and the updated m'0 with m one by one. When the deviation of any corresponding element is reduced compared with that before the update, confirm the current metabolism rate of each organ or tissue. If the deviation of any corresponding element is less than the corresponding preset value, end the iteration and output the updated metabolism rate of each organ or tissue; if the deviation of any corresponding element is not less than the corresponding preset value, on the basis of the update, repeat the update process of the metabolism rate

[0076] It should be noted that the metabolic rate of the organs or tissues of the target object (such as an animal) is not fixed, but will change with factors such as time, age, and weight. Through this design, the metabolic rate of each organ or tissue can be dynamically adjusted according to the actual detection results and comparison data, ensuring that the metabolic rate is more in line with the actual situation. For example, the metabolic rate of an animal may change significantly as it ages or its weight changes. This part of the design introduces a mechanism for gradual optimization, enabling the metabolic rate to adapt to these physiological changes and enhancing the accuracy of prediction.

[0077] Through pairwise comparison between sequences, when the deviation exceeds the preset value, the metabolic rate is updated. This design is a feedback control mechanism that ensures the model is always adjusted in a more accurate direction. By continuously updating the metabolic rate, the model can better handle uncertainties and data deviations, improving the accuracy of the model. The process of continuous iteration can refine the error of each update until the error drops within the preset range. The purpose of this is to improve the stability and robustness of the overall model through multiple corrections, reduce the uncertainty in the prediction process, and ensure that the final prediction result is more accurate and reliable.

[0078] Weight is one of the important factors affecting the metabolic rate. A mechanism for updating the weight reference value is added to the design to ensure that the change in the metabolic rate can be accurately reflected under different weights. By adjusting the metabolic rate and considering the influence of weight, a more accurate prediction result of veterinary drug residues can be finally obtained. By calculating the weight reference value after updating the metabolic rate, the influence of weight can be incorporated into the calculation of veterinary drug metabolism. This design helps to obtain more personalized predictions of veterinary drug residues for animals of different weights.

[0079] By setting a preset deviation value, it is ensured that the process of updating the metabolic rate will not fluctuate excessively. If the error between the updated metabolic rate and the detection result has reached the preset tolerance range, the iteration will end to avoid over-optimization, which may lead to a too-high model complexity or unstable prediction. When the metabolic rates of each organ or tissue have tended to be stable after updating and the error is less than the preset value, the design of terminating the iteration helps to avoid over-debugging and ensures that the model remains simple and operates efficiently.

[0080] S4: Obtain the updated multi-dimensional data through subsequent multi-dimensional detection.

[0081] S5: Adjust the iterative algorithm using the updated multi-dimensional data, and evaluate the veterinary drug residue risk according to the adjusted iterative algorithm.

[0082] Further, the process of adjusting the iterative algorithm includes predicting the multi-dimensional data at the current detection using the metabolic rate updated last time to obtain the predicted value of the current multi-dimensional data; comparing and analyzing the predicted value of the current multi-dimensional data with the updated multi-dimensional data, and obtaining the updated metabolic rate at the current time through the iterative algorithm.

[0083] For the prediction time point P, using the metabolic rates of each organ or tissue at the end of the most recent iteration process, predict the content of veterinary drugs in each organ (the prediction method is the above method), and use the prediction result as the final analysis result of the veterinary drug residue risk.

[0084] It should be noted that by iteratively updating the data after each detection, it can ensure that the risk assessment model is always optimized based on the latest data and status, rather than being fixed on an initial model. By continuously adjusting the metabolic rate according to the latest detection results, the model can adapt to the changes of the target object (such as an animal) at different stages or environmental conditions. Each iteration will adjust the prediction result in real time based on the metabolic rate updated in the previous update. In this way, the model can not only reflect the current situation of veterinary drug residues, but also flexibly respond to new detection data, thereby improving the timeliness and accuracy of prediction.

[0085] By iteratively adjusting the metabolic rate and combining the latest detection data, a precise assessment of the veterinary drug residues in the target object is finally obtained. This enables the risk assessment of veterinary drugs to timely detect potential over-standard risks and provide a more scientific basis for decision-making.

[0086] On the other hand, this embodiment also provides a veterinary drug residue risk data analysis system, which includes:

[0087] A collection unit, which obtains multi-dimensional data by performing an initial multi-dimensional detection on an animal; and obtains updated multi-dimensional data by performing subsequent multi-dimensional detections.

[0088] An analysis unit, which analyzes the current state of the animal according to the multi-dimensional data and calculates the veterinary drug residues in the current state.

[0089] An update unit, which updates the metabolic rate using the optimized iterative algorithm to evaluate the veterinary drug residue risk; adjusts the iterative algorithm using the updated multi-dimensional data, and evaluates the veterinary drug residue risk according to the adjusted iterative algorithm.

[0090] If the above functions are implemented in the form of software functional units 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 invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs.

[0091] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch instructions from the instruction execution system, apparatus, or device and execute the instructions), or in combination with these instruction execution systems, apparatuses, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in combination with an instruction execution system, apparatus, or device.

[0092] More specific examples (non-exhaustive list) of computer-readable media include the following: electrical connection parts with one or more wirings (electronic devices), portable computer disk cartridges (magnetic devices), random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memories), fiber optic devices, and portable compact disc read-only memories (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, then editing, interpreting, or otherwise processing it as appropriate, and then storing it in a computer memory.

[0093] It should be understood that various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits with logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits with suitable combinational logic gate circuits, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0094] Example 2, an embodiment of the present invention, provides a method for analyzing veterinary drug residue risk data. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0095] Animal selection and experimental design: In this embodiment, 3 animal breeds (A, B, C) are selected, namely adult small dogs, large dogs, and rabbits. The body sizes and metabolic characteristics of animals A, B, and C are quite different, aiming to test the accuracy of veterinary drug residue prediction and the applicability of the present invention under different body sizes and metabolic rates. Before the experiment, data such as the basic body weight, organ volume, and metabolic rate of each animal are recorded, and these data provide initial conditions for subsequent experiments.

[0096] Veterinary drug feeding and metabolic model: A common veterinary drug is selected, and its initial feeding amount and feeding time are set. For each animal, according to its body weight and metabolic characteristics, a predetermined veterinary drug feeding schedule is adopted. After each feeding, the content of veterinary drug residue is tracked through regular detection. The metabolic model of each animal adopts the method of dynamically adjusting the metabolic rate described in the present invention, and the metabolic rate is adjusted in real time according to factors such as animal body weight, age, and organ proportion.

[0097] Multi-dimensional data collection: At different time points after each veterinary drug feeding, veterinary drug residue data in each organ (such as liver, kidney, muscle, etc.) of the animal are collected respectively. When updating the data after each sampling, the Bayesian algorithm is used to predict the veterinary drug content in each organ, and the metabolic rate is optimized based on real-time data. After each optimization, the veterinary drug residue amount at the next moment is predicted, and the feeding strategy is adjusted.

[0098] Partial data in the experiment:

[0099] Animal A (small dog): Body weight 5 kg.

[0100] Liver veterinary drug content at initial detection: 2.5 ng / g.

[0101] Kidney veterinary drug content at initial detection: 3.1 ng / g.

[0102] Muscle veterinary drug content at initial detection: 1.2 ng / g.

[0103] Animal B (large dog): Body weight 15 kg.

[0104] Liver veterinary drug content at initial detection: 5.0 ng / g.

[0105] Kidney veterinary drug content at initial detection: 6.5 ng / g.

[0106] Muscle veterinary drug content at initial detection: 2.0 ng / g.

[0107] Animal C (rabbit): Body weight 2 kg.

[0108] Liver veterinary drug content at initial detection: 1.5 ng / g.

[0109] Kidney veterinary drug content at initial detection: 1.8 ng / g.

[0110] Muscle veterinary drug content at initial detection: 0.8 ng / g.

[0111] These data are predicted and optimized through multiple detections by combining the dynamic metabolic model and Bayesian algorithm of the present invention to obtain veterinary drug residue data at different time points.

[0112] At specific time points of the experiment (such as after 24 hours, 48 hours, 72 hours), actual samples are taken from each animal and the veterinary drug residue amount is measured. The following are the prediction time (T) and actual sampling data:

[0113] Animal A (small dog):

[0114] Prediction time T (after 48 hours):

[0115] Predicted liver veterinary drug content: 3.2 ng / g.

[0116] Predicted kidney veterinary drug content: 4.0 ng / g.

[0117] Predicted muscle veterinary drug content: 1.5 ng / g.

[0118] Actual sampling data (after 48 hours):

[0119] Actual liver veterinary drug content: 3.3 ng / g.

[0120] Actual kidney veterinary drug content: 4.1 ng / g.

[0121] Actual muscle veterinary drug content: 1.4 ng / g.

[0122] Animal B (large dog):

[0123] Prediction time T (after 48 hours):

[0124] Predicted veterinary drug content in liver: 5.5 ng / g.

[0125] Predicted veterinary drug content in kidney: 7.2 ng / g.

[0126] Predicted veterinary drug content in muscle: 2.5 ng / g.

[0127] Actual sampling data (after 48 hours):

[0128] Actual veterinary drug content in liver: 5.6 ng / g.

[0129] Actual veterinary drug content in kidney: 7.3 ng / g.

[0130] Actual veterinary drug content in muscle: 2.4 ng / g.

[0131] Animal C (rabbit):

[0132] Prediction time T (after 48 hours):

[0133] Predicted veterinary drug content in liver: 1.8 ng / g.

[0134] Predicted veterinary drug content in kidney: 2.0 ng / g.

[0135] Predicted veterinary drug content in muscle: 1.0 ng / g.

[0136] Actual sampling data (after 48 hours):

[0137] Actual veterinary drug content in liver: 1.9 ng / g.

[0138] Actual veterinary drug content in kidney: 2.1 ng / g.

[0139] Actual veterinary drug content in muscle: 1.1 ng / g.

[0140] By comparing the predicted values with the actual sampling data, the advantages of the present invention in veterinary drug residue prediction can be seen:

[0141] From the above data, it can be seen that the dynamic adjustment mechanism of metabolic rate adopted in the present invention makes the difference between the predicted value and the actual value very small. For example, in the experiment of Animal A, the predicted value of veterinary drug content in liver is 3.2 ng / g, while the actual value is 3.3 ng / g, and the difference between the two is only 0.1 ng / g, indicating a very high prediction accuracy.

[0142] Similarly, the gap between the predicted values and the actual values of Animals B and C is also very small, showing the high precision of the dynamic adjustment of metabolic rate and Bayesian algorithm of the present invention in veterinary drug residue prediction.

[0143] After each sampling, the metabolic rate is continuously optimized based on the actual detection data. For example, the predicted veterinary drug content in the kidney of animal B is 7.2 ng / g, while the actual value is 7.3 ng / g. Although the difference is only 0.1 ng / g, this precisely demonstrates that the present invention can provide increasingly accurate predictions by continuously optimizing the metabolic rate.

[0144] The metabolic model of the present invention can adjust the metabolic rate according to the characteristics of each animal, such as body weight, organ proportion, and age. This makes the prediction of veterinary drug residues in each animal more in line with its physiological characteristics. In the experiment on animal C (rabbit), the difference between the predicted value and the actual value of the veterinary drug residues in the liver and kidney is very small, demonstrating the effectiveness of the personalized metabolic rate model in small animals.

[0145] Compared with traditional static metabolic models or single-detection methods, the present invention not only improves the prediction accuracy by optimizing the metabolic rate in real time but also can better adapt to the metabolic characteristics of different animals. Existing technologies often rely only on single detection or static metabolic rates and cannot perform real-time optimization and adjustment during the experiment, resulting in low prediction accuracy. The dynamic adjustment mechanism of the present invention effectively compensates for this deficiency.

[0146] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A method for analyzing veterinary drug residue risk data, characterized in that: include: Obtain multi-dimensional data by conducting initial multi-dimensional tests on animals; Analyzing the current state of the animal according to the multi-dimensional data, and calculating the veterinary drug residues in the current state; The optimized iterative algorithm is used to update the metabolic rate and evaluate the risk of veterinary drug residues; By performing subsequent multi-dimensional detection, updated multi-dimensional data is obtained; Using the updated multi-dimensional data to adjust the iterative algorithm, and evaluating the veterinary drug residue risk based on the adjusted iterative algorithm; The process of adjusting the iterative algorithm includes predicting the multi-dimensional data of the current test using the metabolic rate after the last update to obtain the predicted value of the multi-dimensional data of the current test; The predicted value of the multi-dimensional data is compared and analyzed with the updated multi-dimensional data, and the updated metabolic rate is obtained through the iterative algorithm.

2. The veterinary drug residue risk data analysis method according to claim 1, characterized in that: The multi-dimensional testing includes blood sample testing, urine sample testing, and milk or egg liquid testing.

3. The veterinary drug residue risk data analysis method according to claim 2, characterized in that: The veterinary drug residues in the current state include, based on the results of the blood sample test, the urine sample test, and the milk or egg liquid test, obtaining the veterinary drug concentration XY0 in the blood, the veterinary drug concentration NY0 in the urine, and the veterinary drug concentration RY0 in the milk or egg liquid, forming a test result set {XY0, NY0, RY0}; Obtain the target object category to be analyzed, and obtain the average metabolic rate of each organ and tissue of the target object category in each age group, as well as the average growth rate of each age group through the category comparison table; Using the trained Bayesian algorithm 1, input: the target object's weight, the average metabolic rate of each organ and tissue in the current age group, and the set {XY0, NY0, RY0}; Output: Residue content of veterinary drugs in various organs or tissues.

4. The veterinary drug residue risk data analysis method according to claim 3, characterized in that: The optimized iterative algorithm includes modeling the process of veterinary drug residues, starting the risk assessment process after a second multi-dimensional test at different times; And according to the latest test results, the metabolic rate of each organ and tissue is iterated to obtain the optimized metabolic rate; The metabolic process is simulated using the optimized metabolic rate, ultimately generating an assessment result of the veterinary drug residue risk.

5. The veterinary drug residue risk data analysis method according to claim 4, characterized in that: The modeling of the process of veterinary drug residues includes building a digital twin model with a timeline as the core, and predicting the weight of the target object at different times on the timeline by setting the daily feed supply; Assume that weight measurement and multi-dimensional detection are performed at time t. While obtaining the current actual weight m, the reference value of the current weight is predicted: m ‘ =Δm t +m0; where Δm t represents the weight change from the current time t to the last measurement; m0 represents the weight at the last measurement; According to the feeding time of veterinary drugs, veterinary drugs are added on the time axis, and the content of veterinary drugs in various organs at time t is predicted based on the amount added: it is obtained by subtracting the metabolic amount from the cumulative amount of veterinary drugs; among them, when calculating the metabolic amount, it is completed by analyzing the content and half-life of veterinary drugs.

6. The veterinary drug residue risk data analysis method according to claim 5, characterized in that: The iterating of the metabolic rate of each organ and tissue includes predicting the multi-dimensional detection results at time t using the trained Bayesian algorithm 2 according to the predicted content of the veterinary drug in each organ at time t and the average metabolic rate of each organ and tissue, and obtaining the predicted veterinary drug concentration XY in the blood ， u , Predicted veterinary drug concentrations in urine NY ， u Predicted veterinary drug concentration in milk or egg ， u , forming the prediction result set {XY ， u ,NY ， u RY ， u }; Through multi-dimensional detection at time t, the actual veterinary drug concentration XY in the blood is obtained u , Actual veterinary drug concentration in urine NY u , actual veterinary drug concentration in milk or egg liquid u , forming a detection result set {XY u ,NY u RY u }; For the sequence XY ， u With XY u ,NY ， u With NY u RY ， u With RY u Compare one by one. When the deviation of any corresponding element exceeds the corresponding preset value, the metabolic rate k of the veterinary drug in the organ or tissue e at each age group is dxcs，e To update: The veterinary drug is preset on e, and the step length of the metabolic rate change ratio is σ e ; For each organ or tissue of the target object, update it in any growth direction, and the updated metabolic rate is k dxcs，e ; Using the updated metabolic rate of each organ or tissue, the content of veterinary drugs in each organ at time t is re-predicted, and based on the prediction results and the updated metabolic rate of each organ and tissue, the trained Bayesian algorithm 2 is used to predict the multi-dimensional detection results at time t to obtain the predicted veterinary drug concentration XY in the blood. ， u1 , Predicted veterinary drug concentrations in urine NY ， u1 Predicted veterinary drug concentration in milk or egg ， u1 , forming the prediction result set {XY ， u1 ,NY ， u1 RY ， u1 }; The step length is σ according to the change ratio of the metabolic rate of each organ or tissue e , as well as the direction of influence of each organ or tissue on body weight, the proportion of each organ or tissue in the target object, and the updated weight reference value m of the metabolic rate ‘ Update and get the updated weight m ’ 0; For the sequence XY ， u1 With XY u ,NY ， u1 With NY u RY ， u1 With RY u 、Updated m ’ 0 and m are compared one by one. When the deviation of any corresponding element decreases compared with that before the update, the metabolic rate of each organ or tissue is confirmed. If the deviation of any corresponding element is less than the corresponding preset value, the iteration is terminated and the metabolic rate of each organ or tissue after the update is output; if the deviation of any corresponding element is not less than the corresponding preset value, the metabolic rate update process is repeated on the basis of the updated value.

7. The veterinary drug residue risk data analysis method according to claim 6, characterized in that: The assessment of the veterinary drug residue risk includes predicting the content of veterinary drugs in each organ using the metabolic rate of each organ or tissue when the most recent iteration process is terminated at the prediction time P, and using the prediction result as the final analysis result of the veterinary drug residue risk.

8. A veterinary drug residue risk data analysis system using the method according to any one of claims 1 to 7, characterized in that: The acquisition unit obtains multi-dimensional data by performing an initial multi-dimensional detection on the animal; and obtains updated multi-dimensional data by performing subsequent multi-dimensional detection; An analysis unit, which analyzes the current state of the animal according to the multi-dimensional data and calculates the veterinary drug residue in the current state; The updating unit uses the optimized iterative algorithm to update the metabolic rate and evaluate the risk of veterinary drug residues; The iterative algorithm is adjusted using the updated multi-dimensional data, and the veterinary drug residue risk is evaluated based on the adjusted iterative algorithm.

9. A computer device comprising: A memory and a processor; the memory stores a computer program, wherein the processor implements the steps of any method as claimed in claim 1 when executing the computer program.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images