A campus food safety big data management method and system

Abstract

The application belongs to the technical field of campus food safety management, and specifically discloses a kind of campus food safety big data management method and system, comprising: monitoring food material quality parameters and binding identity traceability identification, obtaining current risk state through real-time deviation analysis, constructing deviation change curve based on historical data, analyzing quality deviation length to predict risk trend, determining food material safety risk according to double threshold comparison, and tracing it, determining potential risk student group and obtaining its medical data, comprehensively quality risk index and health risk diffusion degree through risk assessment matrix output safety risk grade, finally according to risk grade start grading early warning;The application calculates the comprehensive quality risk index, and evaluates the health risk diffusion degree by combining the student medical data, and makes a rule determination in the risk assessment matrix according to the grade comparison principle, which constitutes a closed-loop evaluation from risk prediction to consequence verification.

Description

Technical Field

[0001] This invention belongs to the field of campus food safety management technology, and relates to a big data management method and system for campus food safety. Background Technology

[0002] Campus food safety management is a crucial intersection of adolescent health protection and public health management. Due to the dense population and concentrated dining environment on campuses, and the fact that students are still developing and have relatively weak immune systems, food safety risks can easily spread rapidly, leading to mass health incidents. Therefore, establishing an intelligent food safety risk prevention and control system capable of providing early warnings and real-time intervention is of great practical significance for protecting student health and maintaining teaching order.

[0003] For example, Chinese invention patent CN107609007A discloses a food safety alert system. This system collects data through an information acquisition unit, matches it with inspection results released by the State Food and Drug Administration through an information matching unit, and then pushes safety reminders to users through an information output unit. This improves the efficiency of food safety information dissemination.

[0004] The existing technologies mentioned above have the following shortcomings: 1. Currently, they mainly rely on published static sampling results for information matching and push, which cannot continuously monitor and analyze food quality parameters in real time. At the same time, they cannot build dynamic models based on historical data to quantify the deterioration trend of risks, resulting in the system lacking early warning and risk foresight, and can only respond passively.

[0005] 2. Currently, information is mainly broadcast to a general user group without establishing a precise traceability path from problematic ingredients to specific consumers. There is a lack of correlation analysis between identity traceability markers and food ingredient data, which makes it impossible to accurately identify potential at-risk groups. At the same time, its decision-making relies solely on single official inspection data and does not incorporate multi-source information such as student health feedback for cross-verification. It is also impossible to build an assessment mechanism for quality risks and health impacts, resulting in insufficient accuracy of early warning targets and a lack of targeted intervention measures. Summary of the Invention

[0006] In view of this, in order to solve the problems mentioned in the background technology, a big data management method and system for campus food safety is proposed.

[0007] The objective of this invention can be achieved through the following technical solution: This invention provides a big data management method for campus food safety, including: S1, monitoring the quality parameters of each food ingredient in the current monitoring period, binding the food ingredient with an identity traceability identifier, and performing deviation analysis between the quality parameters and the benchmark quality parameters to obtain the real-time deviation.

[0008] S2. Based on the real-time deviation, and combined with the historical data of each ingredient, construct a deviation change curve and calculate the quality deviation growth rate of each ingredient.

[0009] S3. Based on the comparison results of the real-time deviation and quality deviation growth rate with the preset threshold, determine whether there is a safety risk to the food ingredients.

[0010] S4. When there is a safety risk in the food ingredients, the use of the food ingredients is tracked based on the identity traceability identifier to identify the student groups at potential risk and obtain the medical data of the student groups at potential risk within a preset observation period.

[0011] S5. Calculate the comprehensive quality risk index based on the real-time deviation and quality deviation growth of risky ingredients, assess the health risk diffusion degree in combination with the medical treatment data, and substitute both into the preset risk assessment matrix to obtain the safety risk level.

[0012] S6. Trigger the corresponding graded early warning mechanism according to the level of safety risk.

[0013] The present invention also provides a campus food safety big data management system, including: a quality monitoring module, which monitors the quality parameters of each ingredient in the current monitoring period, binds the ingredient with an identity traceability identifier, and performs deviation analysis between the quality parameters and the benchmark quality parameters to obtain the real-time deviation.

[0014] The trend analysis module, based on the real-time deviation and combined with the historical data of each ingredient, constructs a deviation change curve and calculates the growth rate of quality deviation for each ingredient.

[0015] The risk assessment module determines whether there is a safety risk in the food ingredients based on the comparison results of the real-time deviation and quality deviation growth rate with preset thresholds.

[0016] The traceability module tracks the use of food ingredients based on the identified traceability markers when food ingredients pose a safety risk, identifies potential high-risk student groups, and obtains their medical records within a preset observation period.

[0017] The safety assessment module calculates a comprehensive quality risk index based on the real-time deviation and quality deviation growth of risky ingredients, and assesses the spread of health risks by combining the medical data. The two are then substituted into a preset risk assessment matrix to obtain the safety risk level.

[0018] The early warning execution module triggers corresponding graded early warning mechanisms based on the level of security risk.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention monitors the quality parameters of food ingredients and calculates the real-time deviation, and combines historical data to construct a deviation change curve to analyze the growth rate of quality deviation, thereby realizing dynamic trend prediction of food safety risks and significantly improving the timeliness and predictability of risk identification.

[0020] (2) This invention identifies the food safety risks of food ingredients and tracks the assigned food sales windows based on the identity traceability mark. Then, by matching consumption records with food recipe data, it accurately identifies potential risk student groups, providing a clear target range for subsequent targeted monitoring and precise intervention, thereby improving the pertinence of risk control.

[0021] (3) This invention calculates the comprehensive quality risk index and assesses the health risk diffusion by combining student medical data. The two are then judged according to the principle of level comparison in the risk assessment matrix, realizing a closed-loop assessment from risk prediction to consequence verification, thereby improving the accuracy of safety risk level determination.

[0022] (4) This invention determines safety risks by setting dual thresholds for deviation and growth, and triggers graded early warnings based on the output of the risk assessment matrix, thereby avoiding overreaction in low-risk situations and ensuring the efficiency of emergency response in high-risk situations, thus improving the rationality of resource utilization and the level of intelligence of emergency response. Attached Figure Description

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

[0024] Figure 1 This is a schematic diagram showing the connections between the steps of the method of the present invention.

[0025] Figure 2 A schematic diagram showing the steps for identifying the student group at potential risk for this invention.

[0026] Figure 3 This is a schematic diagram showing the connections of the various modules in the system of the present invention. Detailed Implementation

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] Please see Figure 1 As shown, the present invention provides a big data management method for campus food safety. The method includes: S1, monitoring the quality parameters of each food ingredient in the current monitoring period, binding the food ingredient with an identity traceability identifier, and performing deviation analysis between the quality parameters and the benchmark quality parameters to obtain the real-time deviation.

[0029] The traceability identifier contains core data fields that uniquely track the source and batch of ingredients, including but not limited to: batch number, supplier's unified social credit code, purchase date and time, and associated third-party testing report number. By scanning the identifier, the system can instantly obtain the above information, providing a data foundation for subsequent traceability.

[0030] It should be added that the benchmark quality parameters include: benchmark microbial content range, benchmark pesticide residue range, and benchmark number of supplier violations. The benchmark microbial content range refers to the legally permissible fluctuation range of microbial indicators such as total bacterial count and coliform bacteria, determined according to national mandatory standards based on the biological characteristics of different food categories. The setting of this range provides a legally valid technical basis for determining the risk of microbial contamination.

[0031] The benchmark pesticide residue range refers to the safe limits for various pesticide residues established according to national standards for various agricultural products. This range serves as a benchmark for assessing the risk of pesticide chemical pollution, ensuring that test data are standardized and comparable.

[0032] The benchmark supplier violation count refers to the tolerance threshold for supplier behavior obtained through statistical analysis of historical procurement and acceptance data from the school cafeteria within a set statistical period. Specifically, it can be calculated as 1.5-2 times the average number of supplier violations within the historical period, or a fixed value can be directly set by the food safety administrator based on management strategies. This benchmark value reflects the dynamic management requirements for supplier behavior compliance.

[0033] It is worth noting that by simultaneously considering both microbial content and pesticide residue levels, a dual coverage of both biological and chemical hazards to food products is achieved, establishing a comprehensive assessment mechanism for direct health risks. Furthermore, by incorporating the number of supplier violations, the level of supply chain quality management is quantified into a calculable risk parameter, thereby establishing a clear risk transmission path from management behavior to quality performance. This enhances the coverage and reliability of risk identification.

[0034] For example, the analysis of the real-time deviation includes: obtaining the microbial content, pesticide residue, and number of supplier behavior violations from the quality parameters, and comparing the microbial content, pesticide residue, and number of supplier behavior violations with their respective benchmark quality parameters.

[0035] Specifically, the supplier's violations include, but are not limited to: providing false test reports, failing to meet temperature control standards during transportation, repeatedly providing food items nearing their expiration date, and having missing or inconsistent accompanying documents.

[0036] For the microbial content, the range standardization calculation is performed with the upper limit of the benchmark microbial content range as the benchmark maximum value and the lower limit as the benchmark minimum value to obtain the benchmark deviation of microbial content.

[0037] Specifically, the formula for calculating the deviation of the microbial content benchmark is as follows: In the formula Indicates the deviation from the baseline for microbial content. Indicates microbial content, This indicates the lower limit of the baseline microbial content range. This represents the upper limit of the baseline microbial content range. When the microbial content is below the safety lower limit, the deviation is negative, indicating a lower risk. When within the baseline range, the microbial content baseline deviation is between 0 and 1. When the microbial content exceeds the safety upper limit, the microbial content baseline deviation will be greater than 1, thus correctly indicating a significant increase in risk.

[0038] The range standardization formula was used for calculation. By uniformly mapping the microbial content indicators of different dimensions to the [0,1] interval, a standardized quantitative basis for multi-source data fusion was established. This linear transformation established a positive correlation between content and risk level, realizing the gradient characterization of risk level.

[0039] Similarly, the deviation from the baseline of pesticide residues was obtained by analyzing the deviation from the baseline of microbial content.

[0040] The relative deviation of the number of supplier violations from the benchmark number of supplier violations is analyzed to obtain the supplier violation benchmark deviation.

[0041] The maximum value among the deviations from the microbial content benchmark, the pesticide residue benchmark, and the supplier violation benchmark is selected as the real-time deviation for each ingredient.

[0042] Considering the potential weakening of risk signals in multi-indicator evaluation systems, this invention preferably selects the maximum value from the deviations of each dimension as the real-time deviation. The principle is that the extreme value screening mechanism can retain the peak signals in each risk dimension, achieving undiminished transmission of prominent single risks. By eliminating the mutual compensation effect of evaluation indicators, it ensures that a complete risk characterization can be obtained for any high-risk state in any dimension.

[0043] Selecting the maximum value as the real-time deviation provides the input data with the highest risk sensitivity for subsequent quality deviation growth analysis. Since the essence of growth analysis is the quantification of the rate of change of risk status, using the maximum value as the benchmark input ensures that the calculated slope parameter can accurately reflect the development trend of the most severe risk dimension, thereby significantly improving the dynamic risk assessment model's ability to capture high-risk signals and its timely warning.

[0044] S2. Based on the real-time deviation, and combined with the historical data of each ingredient, construct a deviation change curve and calculate the quality deviation growth rate of each ingredient.

[0045] For example, the calculation of the quality deviation growth rate of each ingredient includes: obtaining the historical deviation sequence of each ingredient from the historical data, and adding the real-time deviation to the corresponding historical deviation sequence to obtain the deviation change sequence of each ingredient.

[0046] Based on the aforementioned deviation change sequence, a deviation change curve for each ingredient is constructed with the monitoring period as the horizontal axis and the deviation as the vertical axis.

[0047] From the deviation change curve, extract the curve segment with the current monitoring cycle as the end point and a preset duration as the time window, and calculate the slope of the curve segment.

[0048] If the slope is greater than 0, the slope is taken as the quality deviation growth rate; if the slope is less than or equal to 0, 0 is taken as the quality deviation growth rate.

[0049] It's important to add that the slope is used as the measure of quality deviation growth because the mathematical properties of the slope parameter fully characterize the trend direction and intensity of quality deviation. When the slope is positive, it indicates that the real-time deviation is monotonically increasing, objectively reflecting that the food quality is in a stage of deterministic deterioration. When the slope is negative, it indicates that the real-time deviation is decreasing, reflecting that the trend of quality deterioration has been curbed or there are signs of improvement. This quantitative method realizes the transformation from traditional qualitative judgment to precise quantitative analysis, providing calculable and comparable numerical evidence for risk assessment.

[0050] Therefore, the magnitude of the absolute value of the slope directly characterizes the severity of risk evolution. A larger positive slope value indicates that the quality deviation is accelerating, revealing that risk accumulation has entered a non-linear growth stage. Through the above method, the system achieves a quantitative perception of the rate of risk evolution, completing a technological leap from static risk assessment to dynamic risk evolution cognition.

[0051] S3. Based on the comparison results of the real-time deviation and quality deviation growth rate with the preset threshold, determine whether there is a safety risk to the food ingredients.

[0052] For example, determining whether the food has a safety risk includes: comparing the real-time deviation with a preset deviation threshold, and comparing the quality deviation growth rate with a preset quality deviation growth rate threshold.

[0053] It is understood that the preset deviation threshold is a critical value used to determine whether the current quality status of the food constitutes a safety risk. This threshold represents the overall quality status reflected by the real-time deviation of the food at a certain moment, indicating that it has moved from an acceptable state to an unacceptable risk state. In one specific embodiment, the preset deviation threshold is set by collecting historical real-time deviation data of all batches of food over the past year, calculating the 75th percentile of the dataset, and confirming it with food safety experts. After the system is running, the threshold is automatically recalculated and updated quarterly based on new data.

[0054] The preset quality deviation growth threshold is a critical value used to determine whether the trend of deterioration in food quality constitutes a safety risk. This threshold indicates that even if the real-time deviation itself is not high, its rate of deterioration is fast enough to indicate a sharp increase in risk in the short term, requiring immediate attention. It is obtained by analyzing historical data and tracing the quality deviation growth sequence of all food cases ultimately identified as high-risk before the risk outbreak. For example, high-risk cases that were ultimately confirmed to cause health events are selected from historical data, and the quality deviation growth sequence of these cases in the 24 hours before the risk outbreak is traced. The average value plus 1.5 times the standard deviation is calculated and used as the quality deviation growth threshold.

[0055] If the real-time deviation exceeds its preset deviation threshold or the quality deviation growth exceeds its preset quality deviation growth threshold, then the food ingredient is determined to have a safety risk; otherwise, the food ingredient is determined not to have a safety risk.

[0056] S4. When there is a safety risk in the food ingredients, the use of the food ingredients is tracked based on the identity traceability identifier to identify the student groups at potential risk and obtain the medical data of the student groups at potential risk within a preset observation period.

[0057] It should be noted that the preset observation period refers to a fixed time range calculated from the end of the meal in which the potentially hazardous food ingredient was consumed. In one specific embodiment, the preset observation period is set to 4 to 24 hours based on the incubation period of common bacterial food poisoning.

[0058] In one embodiment, the medical data is obtained by connecting to the campus medical room or health center management system via a data interface, and by querying the registration records of the potentially at-risk student group within the preset observation period. For example, the data interface is a standardized interface based on RESTful API or HL7 protocol. The system initiates query requests to the health management system at preset intervals, and the request message contains a list of student identifiers and a time range, and receives the returned structured medical record data.

[0059] Please see Figure 2 As shown, for example, identifying the student group at potential risk includes: querying their outbound records based on the identity traceability identifier of the food with safety risks, and determining the specific food service window to which they are assigned.

[0060] The system retrieves all student consumption records from the cafeteria consumption management system during the theoretical usage period of food with safety risks at the food service window, and extracts student identifiers and details of purchased meals from the consumption records.

[0061] Based on a pre-stored food recipe database, the details of the purchased meals are matched with the ingredients at risk of safety risks, and student identifiers are selected from the purchased meals that contain the ingredients at risk of safety risks.

[0062] In one embodiment of the present invention, the food recipe database is used to establish the association between food items and ingredients. The database stores the correspondence between the food items sold in the canteen and the ingredients they contain.

[0063] Specifically, the database can be a relational data table containing at least the following fields: dish identifier, ingredient identifier, and estimated usage. The dish identifier uniquely identifies a dish, the ingredient identifier uniquely identifies a type of ingredient, and the estimated usage records the standard amount of the ingredient required to prepare the dish, which can be used for more accurate risk assessment. Meanwhile, the data in the dish recipe database comes from standard recipes pre-entered by the cafeteria.

[0064] The selected student identifiers are aggregated to obtain a group of students at potential risk.

[0065] S5. Calculate the comprehensive quality risk index based on the real-time deviation and quality deviation growth of risky ingredients, assess the health risk diffusion degree in combination with the medical treatment data, and substitute both into the preset risk assessment matrix to obtain the safety risk level.

[0066] For example, the calculation of the comprehensive quality risk index includes: for each food ingredient with safety risks, multiplying its real-time deviation degree with its quality deviation growth degree to obtain a single quality risk value for each food ingredient with safety risks.

[0067] When there are multiple food ingredients with safety risks, the individual quality risk values ​​are sorted in descending order, and the value at the top of the sort is selected as the baseline quality risk value; otherwise, the individual quality risk value is used as the baseline quality risk value.

[0068] The average quality risk value is obtained by averaging the individual quality risk values.

[0069] The benchmark quality risk value and the average quality risk value are weighted and combined to obtain the comprehensive quality risk index.

[0070] As an example, the formula for calculating the comprehensive quality risk index is: In the formula As a comprehensive quality risk index, As the baseline quality risk value, This represents the average quality risk value. and These are the weights for the benchmark quality risk value and the average quality risk value, used to quantify their impact on the overall quality risk. , .

[0071] It should be added that by calculating the comprehensive quality risk index through weighted fusion, on the one hand, the weight allocation can reflect the actual weight of the benchmark quality risk value and the average quality risk value in different dimensions of overall risk, reflecting the difference in contribution of the benchmark value (representing the highest risk individual) and the average value (representing the overall risk level) to the degree of comprehensive quality risk. On the other hand, it can directly integrate the information of the benchmark value and the average value, comprehensively considering their synergistic impact on food safety risk, and avoiding the one-sidedness of assessment by a single indicator.

[0072] The weights can be set based on food safety management requirements and actual operational experience, or they can be obtained through experimental data. For example, data on the baseline quality risk value, average quality risk value, and final risk level in historical food safety incidents can be collected first. The correlation coefficient between the two and the risk events can be calculated. The contribution to the overall risk can be determined through regression analysis or principal component analysis. After normalization, the contribution is converted into the weights of the baseline quality risk value and the average quality risk value, and the sum of the weights is 1. This allows for precise quantification of the overall quality risk index.

[0073] For example, the assessment of the spread of health risks includes: obtaining the number of students from the potentially at-risk group who sought medical attention within a preset observation period after meals from the medical data.

[0074] The number of students at potential risk was counted, and the ratio of the number of students seeking medical treatment to the number of students at potential risk was used as the degree of health risk diffusion.

[0075] For example, the preset risk assessment matrix includes: obtaining the distribution of the comprehensive quality risk index of ingredients from historical data, using a clustering algorithm to divide the data into three levels, and obtaining the quality risk index threshold range corresponding to each quality risk level.

[0076] In one specific embodiment, the specific method for obtaining the quality risk index threshold range corresponding to each quality risk level is as follows: the historical comprehensive quality risk index dataset is divided into three clusters using the K-means clustering algorithm, and the optimal clustering scheme is determined by the silhouette coefficient method or the elbow rule, and then the boundary value of each cluster is obtained.

[0077] The cluster boundary values ​​are sorted by numerical value. The upper boundary of the smallest cluster is used as the boundary threshold between low and medium levels, and the upper boundary of the middle cluster is used as the boundary threshold between medium and high levels. Based on the boundary thresholds, threshold intervals corresponding to the three levels are established.

[0078] The threshold range obtained by the cluster analysis method ensures that the classification objectively reflects the natural distribution characteristics of historical data, avoids the subjective bias of manually setting thresholds, and enables the classification results to adapt to the data distribution characteristics of different campus environments, thereby improving the scientificity and adaptability of the risk assessment matrix.

[0079] Based on public health statistical models and the distribution of health data in historical data, the health risk diffusion degree is divided into three levels: low, medium and high, and the threshold range of health risk diffusion degree corresponding to each risk diffusion level is determined.

[0080] It should be added that the threshold range for determining the health risk diffusion degree corresponding to each risk diffusion level includes: based on the authoritative definition of mass health events in the field of public health, the lower threshold of high-level health risk diffusion degree is set as a fixed value, and when the health risk diffusion degree reaches or exceeds this threshold, it is judged as a high-risk diffusion state.

[0081] Based on the distribution characteristics of visit rates during non-event periods in the historical data, the statistical quantile is calculated as the lower threshold for the medium-level health risk diffusion, specifically the 95th percentile of the historical baseline value is taken as the medium-level threshold. Health risk diffusion ranges below the medium-level lower threshold are classified as low-level, thus completing the division of health risk diffusion threshold intervals into three levels.

[0082] It is worth noting that the public health statistical model integrates general authoritative standards with individualized health data characteristics of specific campus environments, improving the classification results of health risk diffusion to be both universal and targeted.

[0083] The threshold ranges of the quality risk index corresponding to each quality risk level are combined with the threshold ranges of the health risk diffusion degree corresponding to each risk diffusion level to form multiple assessment areas. A corresponding safety risk level is preset for each assessment area to obtain a risk assessment matrix.

[0084] In a preferred embodiment, the risk assessment matrix is ​​defined using the highest-priority principle: if either the quality risk level or the health risk diffusion level is high, the safety risk level is set to high. If both are low, the safety risk level is set to low. In all other cases, the safety risk level is set to medium.

[0085] For example, obtaining the safety risk level includes: comparing the comprehensive quality risk index with the quality risk index threshold range corresponding to each quality risk level in the preset risk assessment matrix to obtain the quality risk level.

[0086] The risk diffusion level is obtained by comparing the health risk diffusion degree with the health risk diffusion degree threshold range corresponding to each risk diffusion level in the preset risk assessment matrix.

[0087] The quality risk level is compared with the risk diffusion level, and the higher level is taken as the safety risk level.

[0088] S6. Trigger the corresponding graded early warning mechanism according to the level of safety risk.

[0089] Please see Figure 3 As shown, the present invention provides a campus food safety big data management system, which includes: a quality monitoring module, a trend analysis module, a risk determination module, a traceability module, a safety assessment module, and an early warning execution module.

[0090] In the above, the quality monitoring module is connected to the trend analysis module, the risk assessment module, and the safety assessment module, respectively. The trend analysis module is connected to the risk assessment module and the safety assessment module, respectively. The traceability module is connected to the risk assessment module and the safety assessment module, respectively. The safety assessment module is also connected to the early warning execution module.

[0091] The quality monitoring module monitors the quality parameters of each ingredient in the current monitoring period, binds the ingredient with an identity traceability identifier, and performs deviation analysis between the quality parameters and the benchmark quality parameters to obtain the real-time deviation.

[0092] The trend analysis module, based on the real-time deviation and combined with the historical data of each ingredient, constructs a deviation change curve and calculates the quality deviation growth rate of each ingredient.

[0093] The risk assessment module determines whether there is a safety risk in the food ingredients based on the comparison results of the real-time deviation and quality deviation growth rate with preset thresholds.

[0094] When there is a safety risk in the food, the traceability module tracks the use of the food based on the identity traceability identifier, identifies potential high-risk student groups, and obtains the medical data of potential high-risk student groups within a preset observation period.

[0095] The safety assessment module calculates a comprehensive quality risk index based on the real-time deviation and quality deviation growth of risky ingredients, and assesses the health risk diffusion degree by combining the medical treatment data. The two are then substituted into a preset risk assessment matrix to obtain the safety risk level.

[0096] The early warning execution module triggers a corresponding graded early warning mechanism based on the level of security risk.

[0097] In one specific embodiment, the tiered early warning mechanism adopts differentiated measures based on the level of safety risk: At low risk, the canteen manager is notified to strengthen random inspections and conduct self-inspections within a specified period. At medium risk, the use of the affected ingredients is immediately suspended, the logistics management department is notified to supervise the handling, and health follow-ups are conducted on students dining there. At high risk, the affected window is immediately closed, the emergency plan is activated, the regulatory department is notified, and parents are notified, while medical intervention is initiated. By clearly defining the early warning targets, handling measures, and time limits for different risk levels, an executable closed-loop management system is formed.

[0098] In addition, different levels of warnings correspond to different time limits for handling and information release channels. For example, low-risk warnings require self-inspection to be completed within 24 hours and released through the internal system, while high-risk warnings require immediate activation of emergency response and simultaneous release of warning information through multiple channels such as official announcements and SMS.

[0099] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0100] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0101] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0102] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0103] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0104] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A big data management method for campus food safety, characterized in that: The method includes: S1. Monitor the quality parameters of each ingredient in the current monitoring period, bind the ingredient with an identity traceability identifier, and perform deviation analysis between the quality parameters and the benchmark quality parameters to obtain the real-time deviation. S2. Based on the real-time deviation, and combined with the historical data of each ingredient, construct the deviation change curve and calculate the quality deviation growth rate of each ingredient. S3. Based on the comparison results of the real-time deviation and quality deviation growth rate with the preset threshold, determine whether there is a safety risk to the food ingredients; S4. When there is a safety risk in the food, the use of the food is tracked based on the identity traceability mark to identify the student group at potential risk and obtain the medical data of the student group at potential risk within the preset observation period. S5. Calculate the comprehensive quality risk index based on the real-time deviation and quality deviation growth of risky ingredients, assess the health risk diffusion degree in combination with the medical data, and substitute the two into the preset risk assessment matrix to obtain the safety risk level. S6. Trigger the corresponding graded early warning mechanism according to the level of safety risk; The analysis of the real-time deviation includes: Microbial content, pesticide residue, and number of supplier violations are obtained from the quality parameters, and the microbial content, pesticide residue, and number of supplier violations are compared with their benchmark quality parameters. For the microbial content, the range standardization calculation is performed with the upper limit of the benchmark microbial content range as the benchmark maximum value and the lower limit as the benchmark minimum value to obtain the benchmark deviation of microbial content. Similarly, the deviation from the baseline of pesticide residues was obtained by analyzing the deviation from the baseline of microbial content. The relative deviation of the number of supplier violations from the benchmark number of supplier violations is analyzed to obtain the supplier violation benchmark deviation. The maximum value among the deviations from the microbial content benchmark, the pesticide residue benchmark, and the supplier violation benchmark is selected as the real-time deviation for each ingredient. The calculation of the quality deviation growth rate of each ingredient includes: The historical deviation sequence of each ingredient is obtained from the historical data, and the real-time deviation is added to the corresponding historical deviation sequence to obtain the deviation change sequence of each ingredient. Based on the aforementioned deviation change sequence, a deviation change curve for each ingredient is constructed with the monitoring period as the horizontal axis and the deviation as the vertical axis. From the deviation change curve, extract the curve segment with the current monitoring cycle as the end point and the preset duration as the time window, and calculate the slope of the curve segment; If the slope is greater than 0, the slope is taken as the mass deviation growth rate; if the slope is less than or equal to 0, 0 is taken as the mass deviation growth rate. The preset risk assessment matrix includes: The distribution of the comprehensive quality risk index of ingredients was obtained from historical data. The data was divided into three levels using a clustering algorithm to obtain the threshold range of the quality risk index corresponding to each quality risk level. Based on the distribution of health data in public health statistical models and historical data, the health risk diffusion degree is divided into three levels: low, medium and high, and the threshold range of health risk diffusion degree corresponding to each risk diffusion level is determined. The threshold range of the quality risk index corresponding to each quality risk level is combined with the threshold range of the health risk diffusion degree corresponding to each risk diffusion level to form multiple assessment areas, and a corresponding safety risk level is preset for each assessment area to obtain a risk assessment matrix. The acquisition of the security risk level includes: The comprehensive quality risk index is compared with the quality risk index threshold range corresponding to each quality risk level in the preset risk assessment matrix to obtain the quality risk level. The risk diffusion level is obtained by comparing the health risk diffusion degree with the health risk diffusion degree threshold range corresponding to each risk diffusion level in the preset risk assessment matrix. The quality risk level is compared with the risk diffusion level, and the higher level is taken as the safety risk level.

2. The campus food safety big data management method according to claim 1, characterized in that: The determination of whether food ingredients pose a safety risk includes: The real-time deviation is compared with a preset deviation threshold, and the quality deviation growth rate is compared with a preset quality deviation growth threshold. If the real-time deviation exceeds its preset deviation threshold or the quality deviation growth exceeds its preset quality deviation growth threshold, then the food ingredient is determined to have a safety risk; otherwise, the food ingredient is determined not to have a safety risk.

3. The campus food safety big data management method according to claim 1, characterized in that: The identified student groups at potential risk include: Based on the identification and traceability markers of food ingredients with safety risks, their outbound records can be queried to determine the specific sales window to which they are assigned. The system retrieves all student consumption records from the cafeteria consumption management system during the theoretical usage period of food with safety risks at the food sales window, and extracts student identifiers and details of purchased meals from the consumption records. Based on a pre-stored food recipe database, the details of the purchased food items are matched with the ingredients at risk of safety risks to identify students whose food items contain such ingredients. The selected student identifiers are aggregated to obtain a group of students at potential risk.

4. The campus food safety big data management method according to claim 1, characterized in that: The calculation of the comprehensive quality risk index includes: For each food ingredient with safety risks, the real-time deviation rate and the growth rate of quality deviation are multiplied to calculate the single quality risk value of each food ingredient with safety risks. When there are multiple food ingredients with safety risks, the single quality risk values ​​are sorted in descending order, and the value at the first position in the sort is selected as the baseline quality risk value; otherwise, the single quality risk value is used as the baseline quality risk value. The average quality risk value is obtained by averaging the individual quality risk values. The benchmark quality risk value and the average quality risk value are weighted and combined to obtain the comprehensive quality risk index.

5. The campus food safety big data management method according to claim 1, characterized in that: The assessment of the spread of health risks includes: The number of students at potential risk who sought medical attention within a pre-set observation period after meals was obtained from the aforementioned medical data. The number of students at potential risk was counted, and the ratio of the number of students seeking medical treatment to the number of students at potential risk was used as the degree of health risk diffusion.

6. A campus food safety big data management system, used to execute the steps of the campus food safety big data management method as described in any one of claims 1-5, characterized in that: The system includes: The quality monitoring module monitors the quality parameters of each ingredient in the current monitoring period, binds the ingredient with an identity traceability identifier, and performs deviation analysis between the quality parameters and the benchmark quality parameters to obtain the real-time deviation. The trend analysis module, based on the real-time deviation and combined with the historical data of each ingredient, constructs a deviation change curve and calculates the quality deviation growth rate of each ingredient. The risk assessment module determines whether there is a safety risk in the food ingredients based on the comparison results of the real-time deviation and quality deviation growth rate with preset thresholds. The traceability module tracks the use of food ingredients based on the identity traceability identifier when there is a safety risk, identifies potential risk student groups, and obtains medical data of potential risk student groups within a preset observation period. The safety assessment module calculates a comprehensive quality risk index based on the real-time deviation and quality deviation growth of risky ingredients, assesses the spread of health risks by combining the medical data, and substitutes both into a preset risk assessment matrix to obtain the safety risk level. The early warning execution module triggers corresponding graded early warning mechanisms based on the level of security risk.

Images