Multi-source data fusion and early warning system for cross-border diarrhea pathogenic bacteria

Abstract

The invention discloses a cross-border diarrhea pathogenic bacterium multi-source data fusion and early warning system, relates to the technical field of epidemic prevention, and aims to solve the problem that in the prior art, it is difficult to realize prospective early warning and accurate evaluation of pathogenic bacterium outbreak risks during cross-border transportation of living animals. The system comprises a multi-source data acquisition module, a data preprocessing module, a pre-stress score calculation module, a pathogenic bacterium outbreak association module and a risk comprehensive assessment module. The multi-source data acquisition module is used for acquiring full-process multi-dimensional data of cross-border living animal transportation, including physiological stress indexes, behavior stress indexes, transportation environment indexes and pathogenic bacteria indexes; the data preprocessing module is used for carrying out standardization processing on the collected full-process multi-dimensional data; and the pre-stress score calculation module obtains the comprehensive pre-stress score of the living animal by establishing a pre-stress scoring model. The method has the advantage of realizing prospective early warning and accurate quantitative evaluation of risks.

Description

Technical Field

[0001] This invention relates to the field of epidemic prevention technology, and more specifically, to a multi-source data fusion and early warning system for cross-border diarrheal pathogens. Background Technology

[0002] With the increasing frequency of global livestock farming and pet trade, the scale of cross-border live animal transportation continues to expand, and the impact of stress reactions during transportation on the animal's physical condition and the risk of carrying and spreading pathogens is becoming increasingly prominent.

[0003] Live animals face multiple stressors during cross-border transportation, including transportation duration, density, temperature and humidity fluctuations, and sudden environmental changes. This can lead to neuroendocrine system disorders and significantly elevated levels of stress hormones such as cortisol and norepinephrine. Animals transported for more than 12 hours have higher levels of stress hormones than animals transported for short distances. Cattle, pigs, and other livestock and poultry experience increased heart rate and elevated cortisol levels under stress.

[0004] This stress response not only reduces animal immunity but also directly regulates the virulence expression and transmission ability of diarrhea pathogens. Furthermore, stress hormones can enhance the intestinal adhesion and invasion ability of pathogens such as Salmonella and Escherichia coli, promote the proliferation and excretion of pathogens in the animal's body, and significantly increase the risk of cross-border transmission.

[0005] However, most existing cross-border pathogenic bacteria early warning systems focus only on the pathogenic bacteria detection data itself, neglecting the inducing effect of transportation stress on pathogenic bacteria outbreaks. They also lack effective models to quantify the correlation between stress intensity and pathogenic bacteria outbreaks, making it difficult to achieve proactive early warning and accurate risk assessment. Therefore, we propose a cross-border diarrheal pathogen multi-source data fusion and early warning system. Summary of the Invention

[0006] The purpose of this invention is to provide a multi-source data fusion and early warning system for cross-border diarrheal pathogens, aiming to solve the problem that existing technologies are unable to achieve forward-looking early warning and accurate assessment of the risk of pathogen outbreaks during cross-border transportation of live animals.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a multi-source data fusion and early warning system for transboundary diarrhea pathogens, the system comprising a multi-source data acquisition module, a data preprocessing module, a pre-stress score calculation module, a pathogen outbreak association module, and a risk comprehensive assessment module; The multi-source data acquisition module is used to collect multi-dimensional data of the entire process of cross-border live animal transportation, including physiological stress indicators, behavioral stress indicators, transportation environment indicators, and pathogen indicators. The data preprocessing module is used to standardize the collected multi-dimensional data throughout the entire process. The pre-stress scoring calculation module obtains the comprehensive pre-stress score of live animals by establishing a pre-stress scoring model; The pathogen outbreak association module obtains the outbreak risk score of diarrhea pathogens based on the pre-stress comprehensive score; The comprehensive risk assessment module calculates a comprehensive risk score based on the outbreak risk score using a risk fusion algorithm.

[0008] Preferably, the physiological stress indicators include cortisol level, heart rate, respiratory rate, and body temperature; the behavioral stress indicators include food intake, activity frequency, and sleep duration; the transportation environment indicators include transportation duration, compartment density, temperature and humidity fluctuation range, bump intensity, ventilation conditions, and compartment sealing; and the pathogenic bacteria indicators include pathogenic bacteria types and detection concentrations.

[0009] Preferably, the standardization process includes using box plots or Z-scores to remove outliers from physiological stress indicators, behavioral stress indicators, transportation environment indicators, and pathogen indicators. For missing values, mean filling, median filling, or nearest neighbor filling methods are selected to fill them in, and the units and formats of the indicators are standardized.

[0010] Preferably, the pre-stress scoring calculation module uses the gradient boosting tree algorithm to build the pre-stress scoring model, and obtains the comprehensive pre-stress score of the live animal through the following steps: A stress sample dataset was constructed by collecting physiological stress indicators, behavioral stress indicators, transportation environment indicators, and actual stress comprehensive scores in different cross-border transportation scenarios of live animals. The stress sample dataset was divided into a training set and a test set. The model was trained using the training set and the error rate of the model was calculated using the test set. Then, by inputting the physiological stress indicators, behavioral stress indicators, and transportation environment indicators, which have been standardized by the data preprocessing module, into the model, a comprehensive pre-stress score is obtained.

[0011] Preferably, when the pathogen outbreak association module obtains the outbreak risk score of diarrhea pathogens based on the pre-stress comprehensive score, it uses a machine learning algorithm combining random forest and support vector machine to construct a pathogen outbreak association model, and obtains the outbreak risk score of diarrhea pathogens through the following steps: The outbreak risk dataset was compiled by collecting pathogen indicators, pre-stress comprehensive scores, and actual outbreak risk scores in different cross-border transportation scenarios of live animals. The dataset was divided into training and testing sets. The training set was used to train the model, and the test set was used to calculate the error rate of the model. Then, by inputting the pathogenic bacteria index, which has been standardized by the data preprocessing module, and the comprehensive pre-stress score obtained by the pre-stress score calculation module into the model, an outbreak risk score is obtained.

[0012] Preferably, the calculation formula of the risk fusion algorithm used in the comprehensive risk assessment module is as follows: ; In the formula, For comprehensive risk scoring, To score the risk of outbreak, Cross-border transport transmission coefficient The destination susceptibility coefficient. , , These are the weighting coefficients, and + + =1.

[0013] Preferred cross-border transport propagation coefficient The calculation is performed using the following steps: Obtain the actual cross-border transport distance for live animals. Benchmark transport distance Actual transportation time Benchmark safe transport time Number of quarantine checkpoints along the transportation route The theoretical maximum number of quarantine nodes for this transportation route Then use the formula Calculate the cross-border transport propagation coefficient ; In the formula, Correction factor for carriage sealing , , These are the weighting coefficients. + + =1.

[0014] Preferred destination susceptibility coefficient The calculation is performed using the following steps: Obtain the destination breeding density for cross-border transport of live animals Suitable breeding density at the destination Monthly average temperature and humidity deviation of the destination Deviation of reference temperature and humidity from the threshold The average resistance index of the local flora of the destination to diarrhea-causing pathogens. Maximum resistance index Then use the formula Calculate the destination susceptibility coefficient ; In the formula, , , These are the weighting coefficients. + + =1.

[0015] Preferably, it also includes a warning information output module, which compares the comprehensive risk score calculated by the comprehensive risk assessment module with the warning threshold by setting a preset warning threshold. When the comprehensive risk score > the warning threshold, a warning information is pushed via SMS.

[0016] Preferably, it also includes a multi-dimensional data visualization module, which supports the intuitive display of pre-stress comprehensive score, outbreak risk score, comprehensive risk score, and pathogenic bacteria index in one or more combinations of line graphs, heat maps, and radar charts.

[0017] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention comprehensively covers four core categories of indicators—physiological stress indicators, behavioral stress indicators, transportation environment indicators, and pathogenic bacteria indicators—through a multi-source data acquisition module, ensuring the comprehensiveness of the data foundation for risk assessment. It achieves accurate quantification of the risk from pre-stress state to pathogenic bacteria outbreak risk by constructing a pre-stress scoring model using a gradient boosting tree algorithm and a pathogenic bacteria outbreak correlation model combining random forest and support vector machine. Furthermore, it calculates a comprehensive risk score, enabling early prediction of pathogenic bacteria outbreak risk and achieving proactive early warning and accurate quantitative assessment of risk.

[0018] 2. In the comprehensive risk scoring process, this invention incorporates the cross-border transportation transmission coefficient and the destination susceptibility coefficient into the comprehensive evaluation framework. The cross-border transportation transmission coefficient is quantitatively calculated using key indicators such as transportation distance, duration, number of quarantine points, and vehicle sealing. The destination susceptibility coefficient is precisely calculated by combining scenario-based indicators such as destination breeding density, temperature and humidity deviation, and local microbial resistance. This achieves a dual consideration of transportation transmission risk and destination spread risk, providing a more accurate decision-making basis for risk management. Attached Figure Description

[0019] Figure 1 This is a schematic diagram illustrating the principle of the present invention; Figure 2 This is a schematic diagram of the system architecture in this invention. Detailed Implementation

[0020] Example 1: With the increasing frequency of global livestock farming and pet trade, the scale of cross-border live animal transportation continues to expand. The impact of stress reactions during transportation on the animal's physical condition and the risk of carrying and spreading pathogens is becoming increasingly prominent. This implementation uses full-process, multi-dimensional data collection and fusion analysis to accurately assess the outbreak and spread risk of diarrhea pathogens during cross-border live animal transportation and to promptly output early warning information.

[0021] This embodiment provides a multi-source data fusion and early warning system for transboundary diarrhea pathogens. The system includes a multi-source data acquisition module, a data preprocessing module, a pre-stress score calculation module, a pathogen outbreak association module, and a comprehensive risk assessment module.

[0022] The multi-source data acquisition module is used to collect multi-dimensional data of the entire process of cross-border live animal transportation, including physiological stress indicators, behavioral stress indicators, transportation environment indicators, and pathogen indicators. Physiological stress indicators include cortisol levels, heart rate, respiratory rate, and body temperature. These physiological stress indicators are directly related to the intensity of the animal's stress response during transportation and are an important basis for assessing its physical tolerance and susceptibility to infection. The data collection process must ensure the real-time nature and accuracy of the data. Physiological monitoring equipment should be used to continuously monitor and record the live animals during transportation. Behavioral stress indicators include food intake, activity frequency, and sleep duration. These indicators indirectly reflect the impact of the transportation environment on animals and provide supplementary evidence for stress assessment. The data collection method combines manual observation with automated behavioral monitoring equipment to ensure the objectivity and continuity of the data. Transportation environmental indicators include transportation time, compartment density, temperature and humidity fluctuation range, bump intensity, ventilation conditions and compartment sealing. These indicators directly affect the growth and spread of pathogens and are also key external factors that cause stress in animals. By using environmental sensors and vehicle-mounted sensing equipment, the environment of the entire transportation process can be captured and recorded in real time. The pathogenicity index includes the type and concentration of pathogenic bacteria. This index is obtained by sampling and analyzing pathogenic bacteria in live animals and the transportation environment during transportation using microbial detection technology. It clarifies the type and concentration of pathogenic bacteria and provides a direct basis for subsequent outbreak risk assessment.

[0023] The data preprocessing module is used to standardize the collected multi-dimensional data throughout the entire process. This includes using box plots or Z-scores to remove outliers from physiological stress indicators, behavioral stress indicators, transportation environment indicators, and pathogen indicators. For missing values, it uses mean imputation, median imputation, or nearest neighbor imputation to fill in the gaps. It also standardizes the units and formats of the indicators. For abnormal data in physiological stress indicators, behavioral stress indicators, transportation environment indicators, and pathogen indicators, box plot method or Z-score method is used for identification and removal. The two methods are flexibly selected according to the data type and distribution characteristics. By screening out extreme values ​​that deviate from the normal range, abnormal data is avoided from interfering with the accuracy of subsequent calculation results, and the retained data is ensured to conform to the objective reality. For missing values ​​caused by equipment failure, detection error or other reasons during the data collection process, appropriate filling methods are selected according to the data characteristics: if the data is normally distributed, the mean filling method is used; if the data has a skewed distribution or outlier, the median filling method is used; if the data has temporal or spatial correlation, the nearest neighbor filling method is used. By using reasonable filling methods, the integrity of the dataset is ensured and deviations in model training and calculation due to missing data are avoided. All data after outlier removal and missing value filling are processed to unify units and formats. Since the original units of different indicators are different, all indicators need to be standardized to a dimensionless or dimensionless format to ensure that all indicators are comparable in subsequent model calculations.

[0024] The pre-stress scoring module obtains the comprehensive pre-stress score of live animals by establishing a pre-stress scoring model. Specifically, the comprehensive pre-stress score of live animals is obtained through the following steps: We collected physiological stress indicators, behavioral stress indicators, transportation environment indicators, and actual stress scores under different cross-border transportation scenarios for live animals. We then integrated these data to construct a stress sample dataset. The dataset needs to cover multiple transportation scenarios (such as different transportation distances, different compartment densities, and different animal species) to ensure the diversity and representativeness of the samples.

[0025] The constructed stress sample dataset was randomly divided into a training set and a test set in an 8:2 ratio, with 80% of the data used for model training and 20% of the data used for model testing and validation. A pre-stress scoring model is built using the gradient boosting tree algorithm. The training set data is input into the model for training. The model parameters are iteratively optimized to improve the model's ability to fit the data. Input the test set data into the trained model, calculate the model's error rate, and if the error rate exceeds 5%, retrain and validate until the model's error rate is ≤5%, ensuring that the model has high prediction accuracy. The physiological stress indicators, behavioral stress indicators, and transportation environment indicators output from the data preprocessing stage are input into the pre-stress scoring model. The model outputs a comprehensive pre-stress score for this cross-border live animal transportation scenario through comprehensive analysis and calculation of the three types of indicators. This score will serve as an important input parameter for subsequent pathogen outbreak risk assessment.

[0026] The pathogen outbreak association module obtains the outbreak risk score of diarrhea pathogens based on the pre-stress comprehensive score. Specifically, the outbreak risk score of diarrhea pathogens is obtained through the following steps: Collect pathogen indicators (pathogen types and detection concentrations), pre-stress comprehensive scores, and actual outbreak risk scores for different cross-border transportation scenarios of live animals, and integrate them to form a complete outbreak risk dataset. The dataset needs to cover transportation scenarios with different pathogen types, different stress levels, and different outbreak risk levels to ensure the comprehensiveness of model training. The outbreak risk dataset was divided into training and testing sets in an 8:2 ratio, with 80% of the data used for model training and 20% used for model performance validation. A pathogen outbreak association model was constructed using a machine learning algorithm combining random forest and support vector machine. By combining the ensemble learning advantages of random forest with the classification accuracy of support vector machine, the complex association between pathogen indicators, pre-stress comprehensive score and outbreak risk was effectively explored. The training set data was input into the model for training and the model parameters were optimized. Input the test set data into the trained model, calculate the model's error rate, and if the error rate is >5%, retrain until the model's error rate is ≤5%, ensuring that the model's prediction accuracy meets the system standard. The pathogenic bacteria indicators (pathogenic bacteria species and detection concentration) after data preprocessing and the pre-stress comprehensive score output by the pre-stress score calculation module are jointly input into the pathogenic bacteria outbreak association model. The model calculates the outbreak risk score of diarrhea pathogens in this transportation scenario through the fusion analysis of the two types of input data. This score directly reflects the probability of an outbreak of diarrhea pathogens in the current transportation scenario.

[0027] The comprehensive risk assessment module, based on the outbreak risk score, combines the cross-border transportation transmission coefficient and the destination susceptibility coefficient. It calculates a comprehensive risk score using a risk fusion algorithm to fully assess the transmission risk and impact range of a diarrheal pathogen outbreak. The specific process is as follows: Obtain the actual cross-border transport distance for live animals. Benchmark transport distance Actual transportation time Benchmark safe transport time Number of quarantine checkpoints along the transportation route The theoretical maximum number of quarantine nodes for this transportation route Then use the formula Calculate the cross-border transport propagation coefficient ; In the formula, Correction factor for car body sealing (when sealing is ≥90%) =1.0, 70%-90% =1.2, <70% =1.5), , , These are the weighting coefficients. + + =1. This formula quantifies the risk of pathogens spreading from the carrier to the external environment during transportation. Key influencing factors are identified: the longer the actual transportation distance and the longer the actual transportation time, the higher the probability of pathogen survival and transmission. Therefore, [the formula is used to] calculate [the risk of pathogens spreading from the carrier to the external environment]. (Ratio of actual distance to reference distance) The ratio of actual to baseline safe duration quantifies these two types of risks. The more quarantine checkpoints there are, the higher the probability of pathogens being intercepted and the lower the risk of transmission. Therefore, the ratio is used... The inhibitory factor is quantified by subtracting the actual and maximum quarantine node percentages from 1. Since the three types of factors have different degrees of influence, their weights are normalized. , , By assigning different priorities and weighted summation, the core transportation risk factors are obtained. Considering that the airtightness of the carriage affects the efficiency of pathogen aggregation and spread, a correction coefficient needs to be introduced. After adjusting the core risk factors, the formula for calculating the cross-border transportation transmission coefficient was finally derived. Obtain the destination breeding density for cross-border transport of live animals Suitable breeding density at the destination Monthly average temperature and humidity deviation of the destination ( Deviation of reference temperature and humidity from threshold The average resistance index of the local flora of the destination to diarrhea-causing pathogens. Maximum resistance index Then use the formula Calculate the destination susceptibility coefficient ; In the formula, , , These are the weighting coefficients. + + =1, this formula quantifies the susceptibility of a destination to diarrheal pathogens, screening key influencing factors from two aspects: environmental adaptability and biological resistance. Higher stocking density at the destination leads to more frequent contact between animals and faster transmission of pathogens. The ratio of actual to suitable stocking density quantifies this risk. The greater the deviation of temperature and humidity from the suitable range, the more easily animal immunity declines and pathogens proliferate. The ratio of actual temperature and humidity deviation from the baseline value quantifies the environmental risk. The stronger the resistance of the local flora, the more it can inhibit the reproduction of foreign pathogens, and the lower the susceptibility risk. Therefore, it is used... (1 minus the actual and maximum resistance index percentages) quantifies this resistance factor. Since the three types of factors are independent and have different priorities, they are normalized by weighting. , , By reflecting the importance of each factor and summing them by weight, the formula for calculating the susceptibility coefficient of the destination is finally derived. A risk fusion algorithm is used to calculate a comprehensive risk score, ensuring that the comprehensive risk score can fully and accurately reflect the overall risk level of diarrheal pathogens in cross-border transportation. The calculation formula is as follows: ; In the formula, For comprehensive risk scoring, To score the risk of outbreak, Cross-border transport transmission coefficient The destination susceptibility coefficient. , , These are the weighting coefficients, and + + =1. This formula aims to accurately quantify the overall risk of transboundary diarrhea pathogens and avoid the one-sidedness of assessments based on a single indicator. It needs to integrate three core dimensions: direct risk, transmission risk, and spread risk. As a direct result of the quantification of the risk of pathogenic outbreaks, it is the core source of overall risk. This reflects the ease with which pathogens can spread during transportation, and determines the scope of risk diffusion. This reflects the destination's susceptibility to pathogens and influences the likelihood of risk spread within the destination. Since these three factors have different priorities in impacting the overall risk, a weighted summation logic is used. Furthermore, to ensure a controllable scoring range and avoid results distortion due to differences in indicator magnitudes, normalized weights are employed. , , Finally, the formula for the risk fusion algorithm was derived.

[0028] This embodiment also includes an early warning information output module, which performs early warning judgment and information push based on the comprehensive risk score calculated by the comprehensive risk assessment module and in combination with a preset early warning threshold. The specific process is as follows: A clear warning threshold is set in advance, which is the core basis for determining whether to activate the warning. The warning threshold can be flexibly adjusted according to different transportation scenarios, animal species and destination prevention and control needs. The comprehensive risk score output by the comprehensive risk assessment module Real-time comparison with preset warning thresholds: If the comprehensive risk score is less than or equal to the warning threshold, it indicates that the current risk is within a controllable range and the system will not activate the warning. If the comprehensive risk score is greater than the warning threshold, it indicates that there is a high risk of outbreak and spread of diarrheal pathogens and the warning will be triggered immediately. Once the early warning mechanism is triggered, early warning information is sent to relevant responsible persons (such as transportation managers, quarantine personnel, destination aquaculture managers, public health prevention and control personnel, etc.) via SMS to ensure that relevant personnel are aware of the risk situation in a timely manner and take targeted prevention and control measures (such as strengthening quarantine and testing, optimizing the transportation environment, and making advance preparations for destination prevention and control) to reduce the risk and harm.

[0029] This embodiment also includes a multi-dimensional data visualization module, which supports the intuitive display of pre-stress comprehensive score, outbreak risk score, comprehensive risk score, and pathogen indicators in one or more combinations of line graphs, heat maps, and radar charts, making it convenient for relevant personnel to quickly understand and grasp the risk situation.

[0030] This embodiment comprehensively collects multi-dimensional key data on physiological stress indicators, behavioral stress indicators, transportation environment indicators, and pathogen indicators throughout the entire process of cross-border live animal transportation. Combined with standardized preprocessing methods such as box plots and Z-scores to remove outliers and fill in missing values, it lays a high-quality data foundation for subsequent analysis. Furthermore, by accurately calculating the pre-stress comprehensive score and outbreak risk score, the reliability of the core assessment indicators is ensured. The comprehensive risk score further integrates outbreak risk, scientifically quantified cross-border transportation transmission coefficient, and destination susceptibility coefficient, comprehensively covering key risks such as pathogen outbreaks, cross-border transmission, and destination suitability. The system incorporates multiple risk dimensions, making risk assessment more comprehensive and scientific. Simultaneously, the early warning output module can promptly send alerts via SMS when risks exceed limits. The multi-dimensional data visualization module presents core data and scoring results in intuitive formats such as line graphs and heat maps. This facilitates relevant management personnel in quickly and accurately grasping the risk situation and taking timely and targeted prevention and control measures, effectively reducing the probability of diarrheal pathogen outbreaks. It also helps prevent the risk of cross-border transmission of pathogens, providing efficient, accurate, and convenient technical support for the safe management of cross-border live animal transportation, and improving the level of disease prevention and control and safety management efficiency in the field of cross-border live animal transportation.

[0031] The embodiments disclosed in this invention are preferred embodiments, but are not limited thereto. Those skilled in the art can easily understand the spirit of this invention based on the above embodiments and make different extensions and variations, but as long as they do not depart from the spirit of this invention, they are all within the protection scope of this invention.

Claims

1. A multi-source data fusion and early warning system for transboundary diarrhea pathogens, characterized in that, The system includes a multi-source data acquisition module, a data preprocessing module, a pre-stress score calculation module, a pathogen outbreak correlation module, and a comprehensive risk assessment module; The multi-source data acquisition module is used to collect multi-dimensional data of the entire process of cross-border live animal transportation, including physiological stress indicators, behavioral stress indicators, transportation environment indicators, and pathogen indicators. The data preprocessing module is used to standardize the collected multi-dimensional data throughout the entire process. The pre-stress scoring calculation module obtains the comprehensive pre-stress score of live animals by establishing a pre-stress scoring model; The pathogen outbreak association module obtains the outbreak risk score of diarrhea pathogens based on the pre-stress comprehensive score; The comprehensive risk assessment module calculates a comprehensive risk score based on the outbreak risk score using a risk fusion algorithm.

2. The cross-border diarrheal pathogen multi-source data fusion and early warning system according to claim 1, characterized in that, The physiological stress indicators include cortisol level, heart rate, respiratory rate, and body temperature; the behavioral stress indicators include food intake, activity frequency, and sleep duration; the transportation environment indicators include transportation duration, compartment density, temperature and humidity fluctuation range, bump intensity, ventilation conditions, and compartment sealing; and the pathogen indicators include pathogen types and detection concentrations.

3. The cross-border diarrheal pathogen multi-source data fusion and early warning system according to claim 1, characterized in that, The standardization process includes using box plots or Z-scores to remove outliers from physiological stress indicators, behavioral stress indicators, transportation environment indicators, and pathogen indicators. For missing values, mean filling, median filling, or nearest neighbor filling methods are selected to fill them in, and the units and formats of the indicators are standardized.

4. The cross-border diarrheal pathogen multi-source data fusion and early warning system according to claim 2, characterized in that, The pre-stress scoring calculation module uses the gradient boosting tree algorithm to build the pre-stress scoring model and obtains the comprehensive pre-stress score of the live animal through the following steps: A stress sample dataset was constructed by collecting physiological stress indicators, behavioral stress indicators, transportation environment indicators, and actual stress comprehensive scores in different cross-border transportation scenarios of live animals. The stress sample dataset was divided into a training set and a test set. The model was trained using the training set and the error rate of the model was calculated using the test set. Then, by inputting the physiological stress indicators, behavioral stress indicators, and transportation environment indicators, which have been standardized by the data preprocessing module, into the model, a comprehensive pre-stress score is obtained.

5. The cross-border diarrheal pathogen multi-source data fusion and early warning system according to claim 4, characterized in that, When the pathogen outbreak association module obtains the outbreak risk score of diarrhea pathogens based on the pre-stress comprehensive score, it uses a machine learning algorithm combining random forest and support vector machine to construct a pathogen outbreak association model, and obtains the outbreak risk score of diarrhea pathogens through the following steps: The outbreak risk dataset was compiled by collecting pathogen indicators, pre-stress comprehensive scores, and actual outbreak risk scores in different cross-border transportation scenarios of live animals. The dataset was divided into training and testing sets. The training set was used to train the model, and the test set was used to calculate the error rate of the model. Then, by inputting the pathogenic bacteria index, which has been standardized by the data preprocessing module, and the pre-stress comprehensive score obtained by the pre-stress score calculation module into the model, an outbreak risk score is obtained.

6. The cross-border diarrheal pathogen multi-source data fusion and early warning system according to claim 1, characterized in that, The calculation formula for the risk fusion algorithm used in the comprehensive risk assessment module is as follows: ； In the formula, For comprehensive risk scoring, To score the risk of outbreak, Cross-border transport transmission coefficient The destination susceptibility coefficient. , , These are the weighting coefficients, and + + =1.

7. The cross-border diarrheal pathogen multi-source data fusion and early warning system according to claim 6, characterized in that, Cross-border transport transmission coefficient The calculation is performed using the following steps: Obtain the actual cross-border transport distance for live animals. Benchmark transport distance Actual transportation time Benchmark safe transport time Number of quarantine checkpoints along the transportation route The theoretical maximum number of quarantine nodes for this transportation route Then use the formula Calculate the cross-border transport propagation coefficient ; In the formula, Correction factor for carriage sealing , , These are the weighting coefficients. + + =1.

8. A multi-source data fusion and early warning system for transboundary diarrheal pathogens according to claim 6, characterized in that, Destination Susceptibility Index The calculation is performed using the following steps: Obtain the destination breeding density for cross-border transport of live animals Suitable breeding density at the destination Monthly average temperature and humidity deviation of the destination Deviation of reference temperature and humidity from the threshold Average resistance index of local flora to diarrhea pathogens at the destination Maximum resistance index Then use the formula Calculate the destination susceptibility coefficient ; In the formula, , , These are the weighting coefficients. + + =1.

9. A multi-source data fusion and early warning system for transboundary diarrheal pathogens according to claim 1, characterized in that, It also includes a warning information output module, which compares the comprehensive risk score calculated by the comprehensive risk assessment module with the warning threshold by setting a preset warning threshold. When the comprehensive risk score is greater than the warning threshold, a warning information is pushed via SMS.

10. A multi-source data fusion and early warning system for transboundary diarrheal pathogens according to claim 1, characterized in that, It also includes a multi-dimensional data visualization module, which supports the intuitive display of pre-stress comprehensive score, outbreak risk score, comprehensive risk score, and pathogen indicators in one or more combinations of line graphs, heat maps, and radar charts.

Images