Food safety risk assessment system and method based on AI

Abstract

The invention discloses a food safety risk assessment system and method based on AI, and relates to the technical field of food safety risk assessment, and the system comprises a source statistical unit, a transportation tracking unit, a data statistical unit, a risk assessment unit and an adjustment processing unit. The monitoring unit is used for monitoring the temperature of a shipping place and the temperature of a transportation driving area, analyzing statistical data after statistics of source data is completed, formulating a real-time quality control scheme for a source and detecting the quality of each environment, and the transportation tracking unit formulates a transportation route in the transportation process and then monitors the temperature of the shipping place and the temperature of the transportation driving area. And data generated in the transportation process is counted and transmitted. In the food risk assessment process, the food source and the specific condition in the transportation process can be fully analyzed, it is guaranteed that the risk assessment result is more accurate, and the analysis efficiency can be improved.

Description

Technical Field

[0001] This invention relates to the field of food safety risk assessment technology, specifically to an AI-based food safety risk assessment system and method. Background Technology

[0002] Currently, food safety risk assessment is a scientific evaluation of the potential adverse effects of biological, chemical, and physical hazards in food on human health. Risk assessment is a very important part of food production and marketing. The existing food safety risk assessment methods are relatively simple and cannot be tailored to different types of food. Furthermore, the existing assessment methods contain errors that deviate from the actual situation, resulting in inaccurate final assessment results. Additionally, the methods cannot be adjusted and addressed immediately after the risk assessment is completed, which could lead to greater risks. Summary of the Invention

[0003] The purpose of this invention is to provide an AI-based food safety risk assessment system and method to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: an AI-based food safety risk assessment system, comprising a source statistics unit, a transportation tracking unit, a data statistics unit, a risk assessment unit, and an adjustment and processing unit; The source statistics unit collects data from the food transportation source, analyzes the collected data after the source data collection is completed, and formulates real-time quality control plans for the source to enable quality detection in various environments. The transportation tracking unit plans the transportation route during the transportation process, monitors the temperature at the point of shipment and the areas through which the transportation has passed, and statistically transmits the data generated during the transportation process. The data statistics unit receives the data transmitted by the transportation tracking unit, analyzes the data after receiving it, and identifies abnormal data during the analysis. The risk assessment unit establishes risk assessment levels. During the risk assessment process, the degree of risk is divided according to the established levels. After the level division is completed, a control plan is formulated to control the risk. During the handling process, the problems that cause the risk are collected centrally. The adjustment and processing unit formulates control plans at different levels and develops simulation control models to simulate and adjust the control plans in advance, adjust the transportation plans of the transportation tracking unit, and analyze the problems that arise during the simulation process.

[0005] Preferably, the source statistics unit centrally collects environmental data from the source production area before food is shipped, as well as quality testing data and drug records for unprocessed food. It also collects data on storage environment and storage time. After all data is collected, a data statistics model is established to collect various data from the production process in the source production area. During the collection process, the collected and stored data is timestamped and classified. For identical data, data with earlier storage time is deleted based on the timestamp.

[0006] Preferably, the transportation tracking unit includes a route planning module, a regional climate module, and a real-time monitoring module. The route planning module centrally plans routes from food production areas to their destinations, and separately plans perishable routes, dry goods routes, high-risk routes, commonly used routes, and emergency routes. After the routes are planned, the road conditions of the routes are checked one day in advance. If abnormal road conditions are found, the routes are adjusted. For perishable routes, the mileage during transportation is analyzed. When the mileage is within 500 kilometers, a temperature-controlled refrigerated truck is used, and the interior of the truck is observed every 20 minutes. Transportation departs one hour in advance. When the mileage is between 500 and 1,000 kilometers, with the use of temperature-controlled refrigerated trucks, temperature-controlled transfer stations are set up along the route for transfers. The condition of the truck interior is observed every 30 minutes, and the truck departs two hours in advance. Congestion information for each time period of the route is collected in advance, and the transportation route is adjusted to avoid congestion. When the mileage exceeds 1,000 kilometers, in addition to using temperature-controlled refrigerated trucks and transfer stations, the transfer time is controlled. The condition of the truck interior is observed every 40 minutes, and the transfer time is adjusted according to the temperature of the area where the transfer station is located. For every degree the temperature of the area increases, the transfer time is shortened by 5 minutes. When formulating dry cargo routes, separate short-haul, medium-to-long-haul, and ultra-long-haul routes are developed. The short-haul route covers distances within 300 kilometers, the medium-to-long-haul route covers distances between 300 and 1,000 kilometers, and the ultra-long-haul route covers distances within 1,000 kilometers. The short-haul route prioritizes the use of small trucks and national and provincial highways during transportation. The medium-to-long-haul route prioritizes national highways and uses medium-sized trucks. The ultra-long-haul route prioritizes a combination of rail and air transport. The dryness of the loaded cargo is monitored in real time. When designing high-risk routes, priority is given to air, rail, and highway transportation. Quarantine checkpoints along the route are planned, and quarantine time is controlled in advance during transportation. Positioning and monitoring agencies are set up at the transportation end. Common routes are developed to provide solutions for general food transportation, and emergency route plans are also developed. When problems occur on the above transportation routes, the characteristics of the above routes are extracted during the development of emergency routes. After the characteristics are extracted, the emergency routes are used and adjusted according to the actual transportation conditions.

[0007] Preferably, the regional climate module collects temperature data in the transportation area and the cargo storage area during food transportation, and establishes a food temperature model, including frozen food, refrigerated food, room temperature food, and high-risk special food. The temperature of frozen food is set to be less than or equal to 18 degrees Celsius, with a fluctuation of less than or equal to 2 degrees Celsius; the temperature of refrigerated food is set to be between 0 and 4 degrees Celsius, with a fluctuation of less than or equal to 3 degrees Celsius; the temperature of room temperature food is set to be between 10 and 25 degrees Celsius, with a maximum of 30 degrees Celsius and a minimum of 5 degrees Celsius; and the temperature of high-risk special food is set to be between 2 and 8 degrees Celsius, with a fluctuation of less than or equal to 1 degree Celsius. When the temperature in the transportation area is too high or too low, the temperature inside the transportation and storage area is regulated.

[0008] Preferably, the real-time monitoring module monitors the driving route from the origin to the destination, the transportation time, the temperature of the storage area and the temperature of the driving area during the transportation process. After the monitoring is completed, a monitoring data collection model is formulated, and the collected data is classified and processed, including driving route data, transportation data, storage data and temperature data, etc., and the classified data is timestamped.

[0009] Preferably, the data statistics unit includes a data analysis module and an anomaly identification module. The data analysis module performs comprehensive analysis on the data within the monitoring data collection model, and cleans and preprocesses the data, including removing duplicate and abnormal data and encoding and identifying the data. The analysis includes the following parameters: delivery time (delivery time = arrival time - dispatch time), delivery cost (delivery cost = transportation cost + labor cost + packaging cost + vehicle-related costs), delivery on-time rate (delivery on-time rate = number of on-time deliveries / total number of deliveries), delivery error rate (delivery on-time rate = number of on-time deliveries / total number of deliveries), and food loss rate (food loss rate = quantity of lost food / total quantity of food). The abnormal data generated during the analysis process by the anomaly identification module is identified and classified into an origin database and a transportation database. The origin database collects abnormal data generated during the analysis of origin data, while the transportation database collects abnormal data generated during transportation. After the analysis and collection are completed, a predictive model is developed based on the analysis results to pre-assess the risks of food.

[0010] Preferably, the risk assessment unit includes a risk level module, a scheme formulation module, and an unscheduled spot check module. When assessing food risks, the risk level module formulates different hazard levels, and the classification criteria are based on the degree of harm to the human body and the market share. The standards for the degree of harm and the market share threshold are adjusted according to the type of food, and a first level, a second level, and a third level are formulated respectively, with the first level having the highest risk, the second level second, and the third level having the lowest risk. When the analysis and judgment result is Level 1, data from the source production area and transportation data will be investigated simultaneously, relevant personnel will be consulted, and a comprehensive investigation and systematic testing of related food products will be conducted. When the analysis and judgment result is Level 2, data from the source production area and transportation data will be investigated simultaneously, a large-scale investigation of related food products will be conducted, and sampling tests will be carried out on related food products. When the analysis and judgment result is Level 3, data from the source production area and transportation data will be investigated, a certain range of investigation of related food products will be conducted, simple testing of related food products will be carried out, and a population switching plan will be formulated to determine the impact of food products on different population groups in advance. Once the plan is formulated, a corresponding risk level control plan will be set.

[0011] Preferably, the scheme formulation module extracts and analyzes the data within the data statistical model and the monitoring data collection model. After the analysis, it formulates a source production area detection scheme and a transportation route detection scheme, respectively, to detect the source production area and the transportation route. At the same time, it formulates a food safety risk control scheme, which are a first control scheme, a second control scheme, and a third control scheme. When the risk level is level one, the first control scheme is used to control and adjust the risk. When the risk level is level two, the second control scheme is used to control the risk. When the risk level is level two, the third control scheme is used to control the risk.

[0012] During the processing, the random inspection module connects to an external control center. The control center triggers the random inspection module to conduct random inspections. The inspections are carried out according to the pre-established random inspection plans for the source production area and the transportation route. The inspection results are fed back to the control center for adjustment and processing through the plan formulation module.

[0013] Preferably, the adjustment processing unit includes an adjustment scheme module and a problem analysis module. After the risk assessment is completed, an adjustment scheme for the source production area and a transportation tracking adjustment scheme are formulated based on the results of the risk assessment. The source production area adjustment scheme includes a first source adjustment scheme, a second source adjustment scheme, and a third source adjustment scheme, with the first source adjustment scheme having the largest adjustment range and the third source adjustment scheme having the smallest adjustment range. Similarly, the transportation tracking adjustment scheme includes a first transportation adjustment scheme, a second transportation adjustment scheme, and a third transportation adjustment scheme, with the first transportation adjustment scheme having the largest adjustment range and the third transportation adjustment scheme having the smallest adjustment range.

[0014] The problem analysis module simulates the implementation of the formulated adjustment plan and collects the problems that arise during the simulation. After the problem collection is completed, it is analyzed through expert review. After the analysis is completed, the adjustment plan for the source production area and the adjustment plan for transportation tracking are adjusted.

[0015] The AI-based food safety risk assessment method includes the following steps: Step 1: The source statistics unit collects data from the food transportation source. After the source data is collected, the data is analyzed and a real-time quality control plan is developed for the source. This allows for quality testing of various environments. The transportation tracking unit then plans the transportation route during the transportation process, monitors the temperature at the point of shipment and the areas the transportation has passed through, and collects and transmits the data generated during the transportation process. Step 2: The data statistics unit receives the data transmitted by the transportation tracking unit. After receiving the data, it analyzes it and identifies abnormal data. The risk assessment unit determines the risk assessment level and classifies the risk level according to the determined level. After the level is classified, a control plan is formulated to control the risk. During the handling, the problems that cause the risk are collected centrally. Step 3: Formulate control plans at different levels by adjusting the processing units, and develop a simulation control model to simulate and adjust the control plans in advance. Adjust the transportation plan of the transportation tracking unit and analyze the problems that arise during the simulation process.

[0016] Compared with the prior art, the beneficial effects of the present invention are: This invention, in the process of risk assessment of food, can fully analyze the specific circumstances of the food's source and transportation process, ensuring more accurate risk assessment results and improving analysis efficiency. It classifies risk levels during risk assessment and implements corresponding control schemes based on different levels. These schemes ensure efficient and reasonable control measures, lower costs, and allow for real-time module adjustments, guaranteeing the timeliness and accuracy of the analysis and preventing discrepancies between the actual control measures and the actual situation, thus avoiding situations where adjustments fail to achieve the expected results. Attached Figure Description

[0017] Figure 1 A schematic diagram of the overall structure is provided for embodiments of the present invention; Figure 2 This is a schematic diagram of the device structure provided in an embodiment of the present invention; Figure 3 A partial cross-sectional view of the trigger ring is provided for an embodiment of the present invention.

[0018] In the diagram: 1. Inspection frame; 2. Arched frame; 3. Turbine fan head; 4. Inlet filter plate; 5. Single-phase motor; 6. Sphere; 7. Air guide plate; 8. Air outlet; 9. Rotating rod; 10. Connecting rod; 11. Trigger ring; 12. Movable top rod; 13. Silicone block; 14. Mounting slot; 15. Touch switch. Detailed Implementation

[0019] 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.

[0020] Example 1: Please see Figure 1 The present invention provides a technical solution: an AI-based food safety risk assessment system, including a source statistics unit, a transportation tracking unit, a data statistics unit, a risk assessment unit, and an adjustment and processing unit; The source statistics unit collects data from the food transportation source, analyzes the collected data after the source data collection is completed, and formulates real-time quality control plans for the source to enable quality detection in various environments. The transportation tracking unit plans the transportation route during the transportation process, monitors the temperature at the point of shipment and the areas through which the transportation has passed, and statistically transmits the data generated during the transportation process. The data statistics unit receives the data transmitted by the transportation tracking unit, analyzes the data after receiving it, and identifies abnormal data during the analysis. The risk assessment unit establishes risk assessment levels. During the risk assessment process, the degree of risk is divided according to the established levels. After the level division is completed, a control plan is formulated to control the risk. During the handling process, the problems that cause the risk are collected centrally. The adjustment and processing unit formulates control plans at different levels and develops simulation control models to simulate and adjust the control plans in advance, adjust the transportation plans of the transportation tracking unit, and analyze the problems that arise during the simulation process.

[0021] The source statistics unit collects environmental data (e.g., soil, water quality, and air) from the source production area before food is shipped, as well as quality testing data and records of drugs used in unprocessed food. It also collects data on storage environment and storage time. After collecting all the data, a data statistics model is established to collect various data from the production process in the source production area. During the collection process, the collected and stored data is timestamped and classified. For identical data, the data with the earlier storage time is deleted based on the timestamp. Furthermore, the data statistics model internally constructs statistical schemes for different foods in their unprocessed state, and builds a fine-tuning module within the data statistics model. When using the corresponding statistical scheme, adjustments are made through the fine-tuning module. After fine-tuning, the real-time situation of the place of origin is collected, and the statistical scheme used after collection and analysis is fine-tuned. Moreover, the analysis is conducted online by experts.

[0022] The transportation tracking unit includes a route planning module, a regional climate module, and a real-time monitoring module. The route planning module centrally plans routes from food production areas to their destinations, separately designating routes for perishable goods, dry goods, high-risk goods, common goods, and emergency routes. After route planning, road conditions are checked one day in advance. If abnormalities are found, the route is adjusted. For perishable routes, the mileage during transport is analyzed. When the mileage is within 500 kilometers, a temperature-controlled refrigerated truck is used, and the interior of the truck is observed every 20 minutes. Transportation departs one hour in advance. For distances between 500 and 1000 kilometers, when using temperature-controlled refrigerated trucks, temperature-controlled transfer stations are set up along the route for transfers. The condition of the truck interior is observed every 30 minutes, and the truck departs two hours in advance. Congestion information for each time period of the route is collected in advance, and the transportation route is adjusted to avoid congestion. When the distance traveled is over 1000 kilometers, in addition to using temperature-controlled refrigerated trucks and transfer stations, the transfer time is controlled. The condition of the truck interior is observed every 40 minutes, and the transfer time is adjusted according to the temperature of the area where the transfer station is located. For every degree the temperature of the area increases, the transfer time is shortened by 5 minutes. When formulating dry cargo routes, separate short-haul, medium-to-long-haul, and ultra-long-haul routes are developed. The short-haul route covers distances within 300 kilometers, the medium-to-long-haul route covers distances between 300 and 1,000 kilometers, and the ultra-long-haul route covers distances within 1,000 kilometers. The short-haul route prioritizes the use of small trucks and national and provincial highways during transportation. The medium-to-long-haul route prioritizes national highways and uses medium-sized trucks. The ultra-long-haul route prioritizes a combination of rail and air transport. The dryness of the loaded cargo is monitored in real time. When designing high-risk routes, priority is given to air, rail, and highway transportation. Quarantine checkpoints along the route are planned, and quarantine time is controlled in advance during transportation. Positioning and monitoring agencies are set up at the transportation end. Common routes are developed to provide solutions for general food transportation, and emergency route plans are also developed. When problems occur on the above transportation routes, the characteristics of the above routes are extracted during the development of emergency routes. After the characteristics are extracted, the emergency routes are used and adjusted according to the actual transportation conditions.

[0023] During food transportation, the regional climate module collects temperature data from both the transportation and storage areas, and establishes food temperature models for frozen, refrigerated, room-temperature, and high-risk special foods. The temperature for frozen foods is set to be ≤18 degrees Celsius with a fluctuation of ≤2 degrees Celsius; the temperature for refrigerated foods is set to be ≤0 to 4 degrees Celsius with a fluctuation of ≤3 degrees Celsius; the temperature for room-temperature foods is set to be ≤10 to 25 degrees Celsius, with a maximum of ≤30 degrees Celsius and a minimum of ≤5 degrees Celsius; and the temperature for high-risk special foods is set to be ≤2 to 8 degrees Celsius with a fluctuation of ≤1 degree Celsius. When the temperature in the transportation area is too high or too low, the module regulates the temperature within the transportation and storage areas.

[0024] During transportation, the real-time monitoring module monitors the travel route from the origin to the destination, the transportation time, the temperature of the storage area, and the temperature of the travel area. After the monitoring is completed, a monitoring data collection model is developed, and the collected data is classified and processed, including travel route data, transportation data, storage data, and temperature data, etc., and the classified data is timestamped.

[0025] The data statistics unit includes a data analysis module and an anomaly identification module. The data analysis module performs comprehensive analysis on the data within the monitoring data collection model, and cleans and preprocesses the data, including removing duplicate and abnormal data and encoding and identifying the data. The analysis includes the following parameters: delivery time (delivery time = arrival time - dispatch time), delivery cost (delivery cost = transportation cost + labor cost + packaging cost + vehicle-related costs, and vehicle-related costs include depreciation cost, maintenance cost, cargo insurance and capital occupation cost), delivery on-time rate (delivery on-time rate = number of on-time deliveries / total number of deliveries), delivery error rate (delivery on-time rate = number of on-time deliveries / total number of deliveries), and food loss rate (food loss rate = quantity of food lost / total quantity of food). A comprehensive analysis model was developed, and the specific construction steps are as follows: (1) Route optimization analysis: The specific analysis method is to use GIS system and Python and other tools, combined with real-time traffic congestion data, the order of delivery points and the time requirements for food preservation, to optimize the delivery route. The purpose of the analysis is to find the optimal delivery route and reduce delivery time and cost. (2) Quality dynamic early warning analysis: The specific analysis method is that when the temperature and humidity during transportation exceed the safe range and the food testing data is abnormal, the system pushes early warning information in real time and triggers the backup plan at the same time. The purpose of the analysis is to promptly discover and deal with food quality problems and ensure food safety. (3) Procurement dynamic optimization analysis: The specific analysis method is to adjust based on the historical consumption data and number of diners of the demand side, and combine the fluctuation of food prices, production capacity and food quality in the place of origin to automatically generate the optimal procurement list. The purpose of the analysis is to reduce procurement costs and ensure the freshness of food.

[0026] The abnormal data generated during the analysis process by the anomaly identification module is identified and classified into an origin database and a transportation database. The origin database collects abnormal data generated during the analysis of origin data, while the transportation database collects abnormal data generated during transportation. After the analysis and collection are completed, a predictive model is developed based on the analysis results to pre-assess the risks of food.

[0027] The risk assessment unit includes a risk level module, a scheme formulation module, and an unscheduled spot check module. When assessing food risks, the risk level module formulates different hazard levels, and the classification criteria are based on the degree of harm to the human body and the share of the food entering the market. The standards for the degree of harm and the share threshold are adjusted according to the type of food, and three levels are formulated respectively, with the first level having the highest risk, the second level second, and the third level having the lowest risk. When the analysis and judgment result is Level 1, data from the source production area and transportation data will be investigated simultaneously, relevant personnel will be consulted, and a comprehensive investigation and systematic testing of related food products will be conducted. When the analysis and judgment result is Level 2, data from the source production area and transportation data will be investigated simultaneously, a large-scale investigation of related food products will be conducted, and sampling tests will be carried out on related food products. When the analysis and judgment result is Level 3, data from the source production area and transportation data will be investigated, a certain range of investigation of related food products will be conducted, simple testing of related food products will be carried out, and a population switching plan will be formulated to determine the impact of food products on different population groups in advance. Once the plan is formulated, a corresponding risk level control plan will be set.

[0028] The scheme formulation module extracts and analyzes the data within the data statistical model and the monitoring data collection model. After the analysis, it formulates source production area detection schemes and transportation route detection schemes respectively, and conducts detection on the source production area and transportation route. At the same time, it formulates food safety risk control schemes, namely the first control scheme, the second control scheme, and the third control scheme. When the risk level is the first level, the first control scheme is used to control and adjust the risk; when the risk level is the second level, the second control scheme is used to control the risk; and when the risk level is the third level, the third control scheme is used to control the risk.

[0029] During the processing, the random inspection module connects to an external control center. The control center triggers the random inspection module to conduct random inspections. The inspections are carried out according to the pre-established random inspection plans for the source production area and the transportation route. The inspection results are fed back to the control center for adjustment and processing through the plan formulation module.

[0030] The adjustment processing unit includes an adjustment scheme module and a problem analysis module. After the risk assessment is completed, an adjustment scheme for the source production area and a transportation tracking adjustment scheme are formulated based on the results of the risk assessment. The source production area adjustment scheme includes a first source adjustment scheme, a second source adjustment scheme, and a third source adjustment scheme, with the first source adjustment scheme having the largest adjustment range and the third source adjustment scheme having the smallest adjustment range. Similarly, the transportation tracking adjustment scheme includes a first transportation adjustment scheme, a second transportation adjustment scheme, and a third transportation adjustment scheme, with the first transportation adjustment scheme having the largest adjustment range and the third transportation adjustment scheme having the smallest adjustment range.

[0031] The problem analysis module simulates the implementation of the formulated adjustment plan and collects the problems that arise during the simulation. After the problem collection is completed, it is analyzed through expert review. After the analysis is completed, the adjustment plan for the source production area and the adjustment plan for transportation tracking are adjusted.

[0032] Example 2: Please see Figure 2 and Figure 3, the present invention provides a technical solution: The irregular sampling module includes a sampling frame 1, and an arched frame 2 is installed at the bottom end inside the sampling frame 1. A plurality of air inlets are opened around the middle of the bottom end of the inner cavity of the sampling frame 1, and air inlet filter plates 4 are installed inside the air inlets. A single-phase motor 5 is installed at the bottom end of the sampling frame 1. A rotating hole is opened in the middle of the arched frame 2. A rotating rod 9 is installed at the top end of the single-phase motor 5. A circular hole is opened at the bottom end of the sampling frame 1. The top end of the rotating rod 9 passes through the circular hole and the rotating hole and is installed with a connecting rod 10. A vortex fan head 3 is installed at the bottom end of the rotating rod 9 inside the arched frame 2. Arc-shaped plates are installed around the bottom end of the inner cavity of the sampling frame 1. A plurality of air delivery holes 8 are opened around the arched frame 2. A guide air plate 7 is installed at the upper end around the arched frame 2. The left cross-section of the guide air plate 7 is in an inverted L-shaped structure. An annular guide plate is installed around the inner cavity of the arched frame 2 below the air delivery holes 8. A plurality of circular balls 6 are placed at the bottom end of the inner cavity of the sampling frame 1. A plurality of small balls are placed at the bottom end of the inner cavity of the sampling frame 1. Both the circular balls 6 and the small balls are hollow structures. A trigger ring 11 is installed on the left side of the connecting rod 10. A plurality of installation slots 14 are equidistantly opened around the rear side of the trigger ring 11. A plurality of movable ejector rods 12 are equidistantly arranged around the inner cavity of the trigger ring 11. The opposite ends of the movable ejector rods 12 around the circumference are connected to the inner cavity of the installation slots 14 through return hinges. A touch switch 15 is installed on one side of the inner cavity of the installation slots 14. A silica gel block 13 is installed at the rear end of the movable ejector rod 12. The circular dimension formed by the opposite ends of the movable ejector rods 12 around the circumference is smaller than the dimension of the circular ball 6. The circular dimension formed by the opposite ends of the movable ejector rods 12 around the circumference is larger than the dimension of the small ball. The touch switch 15 is electrically connected to an external control center; The specific implementation method is as follows: When performing irregular sampling, the single-phase motor 5 is started through an external control center to drive the rotating rod 9 to rotate. The rotating rod 9 drives the vortex fan head 3 to rotate, and an air flow is formed inside the arched frame 2, so that the air flow passes through the air delivery holes 8 and hits the outside of the arched frame 2. The air flow is sprayed onto the bottom end inside the sampling frame 1 through the guide air plate 7 and contacts the circular balls 6 and the small balls, blowing up the circular balls 6 and the small balls, making the circular balls 6 and the small balls float inside the sampling frame 1. When the rotating rod 9 drives the connecting rod 10 to rotate, the trigger ring 11 is driven to rotate. When rotating, the trigger ring 11 contacts the floating circular ball 6 and pushes the movable ejector rod 12 to move around the return hinge as the axis, then drives the silica gel block 13 to move, making the silica gel block 13 contact and squeeze the touch switch 15, triggering the external control center. The sampling model is triggered through the control center to perform sampling operations. Through the implementation of the irregular detection scheme supporting the irregular detection module, the real situation of each link of the food can be sampled, avoiding the generation of loopholes due to regularity during regular sampling, which is easy for people to seize the loopholes and cause造假 and preparation in advance, resulting in the final detection result not being real enough, and adding different numbers of small balls adaptively each time of use. It should be noted that the term "造假" in the original text seems to be incorrect. It might be a misspelling. If it is "造假" (which means "to fabricate; to fake"), it is an inappropriate term in this context. It is recommended to check and correct it to a more appropriate expression such as "cheating" or "forgery" to make the translation more accurate and the content more understandable.

[0033] Example 3: The AI-based food safety risk assessment method includes the following steps: Step 1: The source statistics unit collects data from the food transportation source. After the source data is collected, the data is analyzed and a real-time quality control plan is developed for the source. This allows for quality testing of various environments. The transportation tracking unit then plans the transportation route during the transportation process, monitors the temperature at the point of shipment and the areas the transportation has passed through, and collects and transmits the data generated during the transportation process. Step 2: The data statistics unit receives the data transmitted by the transportation tracking unit. After receiving the data, it analyzes it and identifies abnormal data. The risk assessment unit determines the risk assessment level and classifies the risk level according to the determined level. After the level is classified, a control plan is formulated to control the risk. During the handling, the problems that cause the risk are collected centrally. Step 3: Formulate control plans at different levels by adjusting the processing units, and develop a simulation control model to simulate and adjust the control plans in advance. Adjust the transportation plan of the transportation tracking unit and analyze the problems that arise during the simulation process.

[0034] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0035] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An AI-based food safety risk assessment system, characterized by: Including source statistics unit, transportation tracking unit, data statistics unit, risk assessment unit and adjustment processing unit; The source statistics unit generates data statistics of the food transportation source, analyzes the statistical data, and formulates a real-time quality control plan for the source to detect the quality of each environment; The transportation tracking unit formulates the transportation route during transportation, monitors the temperature of the delivery location and the area where the transportation travels, and transmits the data generated during transportation; The data statistics unit receives the data transmitted by the transportation tracking unit, analyzes the data after receiving, and identifies abnormal data during analysis; The risk assessment unit formulates risk assessment levels, divides the risk level according to the formulated levels during risk assessment, formulates a control scheme to control the risk after the level division is completed, and collects the problems that cause the risk during processing; The adjustment processing unit formulates different levels of control schemes and simulates the control model to simulate and adjust the control scheme in advance, and adjusts the transportation scheme of the transportation tracking unit. 2.The AI-based food safety risk assessment system according to claim 1, characterized in that: The source statistics unit collects the environmental data of the source area before the food is sent, collects the quality detection data of the unprocessed food and the record of the use of drugs for unprocessed food, and collects the data of the storage environment and storage time, establishes a data statistics model to collect various data during the production process of the source area after collecting various data, and classifies the collected data by time stamp during the collection process. Delete the data with earlier storage time according to the same data according to the time stamp. 3.The AI-based food safety risk assessment system according to claim 2, characterized in that: The transportation tracking unit includes a route planning module, a regional climate module, and a real-time monitoring module. The route planning module centrally formulates the route from the food production area to the destination, and formulates perishable route, dry goods route, high-risk route, common route and emergency route. After the route is formulated, the road conditions of the driving route are queried one day in advance, and the driving route is adjusted when the road conditions are found to be abnormal. The kilometer number during transportation is analyzed and processed for the perishable route, and the inside of the vehicle compartment is observed every 20 minutes, and the vehicle is transported one hour in advance. When the driving distance is between 500 and 1000 kilometers, the temperature preservation vehicle compartment is used, and the inside of the vehicle compartment is observed every 30 minutes along the way, and the vehicle is transported two hours in advance. The congestion situation of each time period of the driving route is collected in advance, and the transportation route is adjusted to avoid. When the driving distance is more than 1000 kilometers, the temperature preservation vehicle compartment and the transfer station are used at the same time, and the transfer time is controlled, the inside of the vehicle compartment is observed every 40 minutes, and the temperature of the area where the transfer station is located is adjusted according to the temperature of the area where the transfer station is located. The temperature of the area where the transfer station is located is increased by one degree, and the transfer time is shortened by five minutes. When formulating dry cargo routes, separate short-haul, medium-to-long-haul, and ultra-long-haul routes are developed. The short-haul route covers distances within 300 kilometers, the medium-to-long-haul route covers distances between 300 and 1,000 kilometers, and the ultra-long-haul route covers distances within 1,000 kilometers. The short-haul route prioritizes the use of small trucks and national and provincial highways during transportation. The medium-to-long-haul route prioritizes national highways and uses medium-sized trucks. The ultra-long-haul route prioritizes a combination of rail and air transport. The dryness of the loaded cargo is monitored in real time. When designing high-risk routes, priority is given to air, rail, and highway transportation. Quarantine checkpoints along the route are planned, and quarantine time is controlled in advance during transportation. Positioning and monitoring agencies are set up at the transportation end. Common routes are developed to provide solutions for general food transportation, and emergency route plans are also developed. When problems occur on the above transportation routes, the characteristics of the above routes are extracted during the development of emergency routes. After the characteristics are extracted, the emergency routes are used and adjusted according to the actual transportation conditions. 4.The AI-based food safety risk assessment system according to claim 3, characterized in that: During food transportation, the regional climate module collects temperature data from both the transportation and storage areas, and establishes food temperature models for frozen, refrigerated, room-temperature, and high-risk special foods. The temperature for frozen foods is set to be ≤18 degrees Celsius with a fluctuation of ≤2 degrees Celsius; the temperature for refrigerated foods is set to be ≤0 to 4 degrees Celsius with a fluctuation of ≤3 degrees Celsius; the temperature for room-temperature foods is set to be ≤10 to 25 degrees Celsius, with a maximum of ≤30 degrees Celsius and a minimum of ≤5 degrees Celsius; and the temperature for high-risk special foods is set to be ≤2 to 8 degrees Celsius with a fluctuation of ≤1 degree Celsius. When the temperature in the transportation area is too high or too low, the module regulates the temperature within the transportation and storage areas. 5.The AI-based food safety risk assessment system according to claim 4, characterized in that: During transportation, the real-time monitoring module monitors the travel route from the origin to the destination, the transportation time, the temperature of the storage area, and the temperature of the travel area. After the monitoring is completed, a monitoring data collection model is developed, and the collected data is classified and processed, including travel route data, transportation data, storage data, and temperature data, etc., and the classified data is timestamped. 6.The AI-based food safety risk assessment system according to claim 5, characterized in that: The data statistics unit includes a data analysis module and an anomaly identification module. The data analysis module performs comprehensive analysis on the data within the monitoring data collection model and cleans and preprocesses the data, including removing duplicate and abnormal data and encoding and identifying the data. The module also analyzes the delivery time, delivery cost, delivery on-time rate, delivery error rate, and food loss rate. The abnormal data generated during the analysis process by the anomaly identification module is identified and classified into an origin database and a transportation database. The origin database collects abnormal data generated during the analysis of origin data, while the transportation database collects abnormal data generated during transportation. After the analysis and collection are completed, a predictive model is developed based on the analysis results to pre-assess the risks of food. 7.The AI-based food safety risk assessment system according to claim 6, characterized in that: The risk assessment unit includes a risk level module, a scheme formulation module, and an unscheduled spot check module. When assessing food risks, the risk level module formulates different hazard levels, and the classification criteria are based on the degree of harm to the human body and the share of the food entering the market. The standards for the degree of harm and the share threshold are adjusted according to the type of food, and three levels are formulated respectively, with the first level having the highest risk, the second level second, and the third level having the lowest risk. When the analysis and judgment result is Level 1, data from the source production area and transportation data will be investigated simultaneously, relevant personnel will be consulted, and a comprehensive investigation and systematic testing of related food products will be conducted. When the analysis and judgment result is Level 2, data from the source production area and transportation data will be investigated simultaneously, a large-scale investigation of related food products will be conducted, and sampling tests will be carried out on related food products. When the analysis and judgment result is Level 3, data from the source production area and transportation data will be investigated, a certain range of investigation of related food products will be conducted, simple testing of related food products will be carried out, and a population switching plan will be formulated to determine the impact of food products on different population groups in advance. Once the plan is formulated, a corresponding risk level control plan will be set. 8.The AI-based food safety risk assessment system according to claim 7, characterized in that: The scheme formulation module extracts and analyzes the data within the data statistical model and the monitoring data collection model. After the analysis, it formulates source production area detection schemes and transportation route detection schemes respectively, and conducts detection on the source production area and transportation route. At the same time, it formulates food safety risk control schemes, namely the first control scheme, the second control scheme, and the third control scheme. When the risk level is the first level, the first control scheme is used to control and adjust the risk; when the risk level is the second level, the second control scheme is used to control the risk; when the risk level is the third level, the third control scheme is used to control the risk. During the processing, the random inspection module connects to an external control center. The control center triggers the random inspection module to conduct random inspections. The inspections are carried out according to the pre-established random inspection plans for the source production area and the transportation route. The inspection results are fed back to the control center for adjustment and processing through the plan formulation module. 9.The AI-based food safety risk assessment system according to claim 8, characterized in that: The adjustment processing unit includes an adjustment scheme module and a problem analysis module. After the risk assessment is completed, an adjustment scheme for the source production area and a transportation tracking adjustment scheme are formulated based on the results of the risk assessment. The source production area adjustment scheme includes a first source adjustment scheme, a second source adjustment scheme, and a third source adjustment scheme, with the first source adjustment scheme having the largest adjustment range and the third source adjustment scheme having the smallest adjustment range. Similarly, the transportation tracking adjustment scheme includes a first transportation adjustment scheme, a second transportation adjustment scheme, and a third transportation adjustment scheme, with the first transportation adjustment scheme having the largest adjustment range and the third transportation adjustment scheme having the smallest adjustment range. The problem analysis module simulates the implementation of the formulated adjustment plan and collects the problems that arise during the simulation. After the problem collection is completed, it is analyzed through expert review. After the analysis is completed, the adjustment plan for the source production area and the adjustment plan for transportation tracking are adjusted.

10. An AI-based food safety risk assessment method, characterized by: The risk assessment method is applicable to the AI-based food safety risk assessment system of claim 9, and includes the following steps: Step 1: The source statistics unit collects data from the food transportation source. After the source data is collected, the data is analyzed and a real-time quality control plan is developed for the source. This allows for quality testing of various environments. The transportation tracking unit then plans the transportation route during the transportation process, monitors the temperature at the point of shipment and the areas the transportation has passed through, and collects and transmits the data generated during the transportation process. Step 2: The data statistics unit receives the data transmitted by the transportation tracking unit. After receiving the data, it analyzes it and identifies abnormal data. The risk assessment unit determines the risk assessment level and classifies the risk level according to the determined level. After the level is classified, a control plan is formulated to control the risk. During the handling, the problems that cause the risk are collected centrally. Step 3: Formulate control plans at different levels by adjusting the processing units, and develop a simulation control model to simulate and adjust the control plans in advance. Adjust the transportation plan of the transportation tracking unit and analyze the problems that arise during the simulation process.

Images