A dynamic traceability verification method for cold-chain logistics transportation

Abstract

The application discloses a cold-chain logistics transportation dynamic traceability verification method and belongs to the technical field of cold-chain transportation. The cold-chain logistics transportation dynamic traceability verification method comprises the following steps: dividing a driving track into detection sections with different driving directions, and establishing a prediction model of external environment data and a measured spatial temperature gradient; when entering a next detection section, a reference temperature gradient is calculated by using a prediction model of a previous section; a data credibility or a refrigeration / heat preservation system abnormality is judged according to a comparison between a transition time when a second temperature difference reaches a steady state and a threshold value calculated based on thermal inertia; when the credibility of the prediction model is lower than the threshold value, a controllable thermal disturbance is applied by short-time starting a defrosting heater; and the authenticity of a sensor is verified according to a disturbance duration. The application realizes physical-level dynamic verification of temperature data by using driving direction changes and controllable thermal disturbance, effectively identifies sensor fraud and failure, and improves the credibility of cold-chain traceability data.

Description

Technical Field

[0001] This invention relates to the field of cold chain transportation technology, and more specifically, to a dynamic traceability verification method for cold chain logistics transportation. Background Technology

[0002] Cold chain logistics refers to a supply chain system that maintains goods at a specified low temperature throughout all stages of production, storage, transportation, sales, and consumption. For temperature-sensitive items such as fresh food and pharmaceutical products (e.g., vaccines, biological agents), the integrity of the cold chain is directly related to product quality and safety and public health. In recent years, with consumption upgrades and stricter pharmaceutical regulations, the cold chain logistics market has continued to expand, but the accompanying frequent quality and safety incidents have exposed serious deficiencies in the cold chain traceability system.

[0003] While IoT monitoring theoretically enables end-to-end cold chain monitoring, data falsification and monitoring failures are commonplace in actual operation, severely undermining the credibility of traceability systems. Temperature sensors are the core component of cold chain monitoring, but the accuracy of their data collection is highly dependent on their installation location and physical environment. During actual transportation, various intentional or accidental probe cheating behaviors occur. For example: deliberately placing temperature probes near air conditioning vents, air coolers, or ice surfaces to read localized low-temperature values ​​far below the core temperature of the goods, when the goods are actually in a high-temperature environment; wrapping sensors with foam, blankets, or other materials to isolate them from the external environment, preventing them from sensing true temperature changes; configuring multiple sensors in the system and only uploading data from the optimally positioned probe to the platform, blocking readings from other sensors that reflect the true environment.

[0004] In view of this, we propose a dynamic traceability verification method for cold chain logistics transportation. Summary of the Invention

[0005] 1. Technical problems to be solved The purpose of this invention is to provide a dynamic traceability verification method for cold chain logistics transportation to solve the problems mentioned in the background art.

[0006] 2. Technical Solution This invention is achieved through the following technical solution: A method for dynamic traceability verification of cold chain logistics transportation includes the following steps: S1. Collect the temperature at different locations inside the carriage using temperature sensors, calculate the average temperature and the measured temperature gradient inside the carriage, and upload the average temperature. S2. Collect external environmental data; S3. Divide the driving trajectory into multiple detection segments, and the driving direction of the vehicle is different in two adjacent detection segments; S4. Based on the mapping relationship between external environmental data and the measured spatial temperature gradient of the vehicle under steady state in the current detection section, establish a prediction model to predict the temperature gradient inside the vehicle compartment. S5. Calculate the Euclidean distance between the measured spatial temperature gradient and the predicted temperature gradient, and use it as the first temperature difference. S6. When the first temperature difference exceeds the set threshold, upload complete temperature detection data; if the temperature sensor is not faulty, increase the acquisition frequency. S7. When the vehicle enters the next detection segment, the prediction model corresponding to the previous detection segment is used to calculate the reference temperature gradient, calculate the second temperature difference between the spatial temperature gradient and the reference temperature gradient, and record the transition time required for the second temperature difference to reach a steady state. S8. If the transition time is less than the set first time threshold, a low confidence flag will be uploaded at the current time node; if the transition time is greater than the set second time threshold, complete temperature detection data will be uploaded and the collection frequency will be increased. S9. After the second temperature difference reaches a steady state, repeat S4-S9.

[0007] As an optional solution to the technical solution of this application, in step S1, the measured spatial temperature gradient is the temperature gradient in the width direction of the carriage.

[0008] As an optional solution to the technical solution of this application, in step S2, the external environment data includes: the temperature outside the carriage, the intensity of solar radiation, the direction of travel, the vehicle speed, and the solar azimuth angle.

[0009] As an optional solution to the technical solution of this application, in step S3, the angle between the driving directions on two adjacent detection road sections is not less than 10°-30°.

[0010] As an optional solution to the technical solution of this application, in step S4, the prediction model is: ; ; ; In the formula, W is the weight matrix, and b is the bias. The relative angle of incidence of the sun at the current moment; The weight matrix and bias are calculated using the least squares method.

[0011] As an optional solution to the technical solution of this application, in step S6, the temperature sensor fault is determined by the following steps: A1. Calculate the rate of temperature change for each temperature sensor within a set time window; A2. Calculate the mean and standard deviation of the temperature change rate for all temperature sensors; A3. Temperature sensors whose temperature change rate differs from the mean temperature change rate by more than three times the standard deviation are considered faulty.

[0012] As an optional solution to the technical solution of this application, in step S8, the first time threshold is calculated in the following way: ; ; ; In the formula, Let Veff be the temperature difference between the inside and outside of the carriage, heff be the effective heat capacity volume of the carriage, Aeff be the equivalent heat transfer coefficient of the carriage, and τ be the equivalent heat transfer area of ​​the carriage. air density, τmin, α1, and α2 are the specific heat capacity of air, and are determined through experimental calibration.

[0013] As an optional solution to the technical solution of this application, in step S8, the second time threshold is calculated in the following way: ; ; τmax,β1,β2 were determined through experimental calibration.

[0014] As an optional solution to the technical solution of this application, step S4 further includes: S401. The confidence level of the prediction model is calculated using the following formula; ; In the formula, t is the current time, ts is the update time of the prediction model parameters, and τm is the confidence half-life; S402. When the reliability of the prediction model is lower than the set reliability threshold, a controllable thermal disturbance is applied to the carriage. S403. Calculate the duration of the disturbance from the start of the controllable thermal disturbance to the time when the space temperature gradient reaches a steady state. S404. If the duration of the disturbance is less than the set first disturbance threshold, then upload the low confidence flag at the current node. S404. If the duration of the disturbance exceeds the set second disturbance threshold, upload the complete temperature detection data and increase the acquisition frequency.

[0015] 3. Beneficial effects Compared with the prior art, the beneficial effects of the present invention are: 1) This invention, by deploying multiple temperature sensors, not only calculates the average temperature inside the vehicle compartment but also simultaneously calculates the measured spatial temperature gradient along the width of the compartment, incorporating the spatial distribution characteristics of the temperature field into the monitoring scope. Based on changes in temperature characteristics, the system adjusts the data acquisition frequency and upload strategy. Under normal operating conditions, the system only uploads the average temperature, reducing communication load. When an anomaly is detected or the reliability decreases, the system automatically increases the monitoring granularity, uploads complete data, and identifies the confidence level, providing tiered and reliable data support for subsequent quality traceability and responsibility determination. This closed-loop system significantly enhances the credibility and reliability of the cold chain logistics traceability system.

[0016] 2) This invention comprehensively considers external environmental factors such as outside temperature, solar radiation intensity, driving direction, vehicle speed, and solar azimuth angle. It uses the least squares method to establish a prediction model to predict the temperature gradient inside the vehicle. By comparing the measured spatial temperature gradient with the predicted temperature gradient, the Euclidean distance between the two is calculated as the first temperature difference. The data acquisition and uploading strategy is adjusted according to the first temperature difference. This invention can identify situations where temperature sensor probes are blocked or moved, resulting in low reliability of temperature data. Furthermore, it can perform high-frequency data collection for abnormal vehicle temperature control, reducing the difficulty of tracing the source and increasing the reliability of the tracing results.

[0017] 3) This invention addresses the situation where vehicles maintain a fixed driving direction for extended periods and the reliability of the prediction model decays over time. When the reliability of the prediction model falls below a set threshold, a controllable thermal disturbance is applied to the passenger compartment by briefly activating the defrost heater. The duration of the disturbance, during which the spatial temperature gradient reaches a steady state, is monitored. This proactively tests the reliability of temperature data, overcoming the shortcomings of relying on external thermal disturbances and being unable to monitor situations where vehicles maintain a fixed driving direction for extended periods. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the overall structure of a dynamic traceability verification method for cold chain logistics transportation. Detailed Implementation

[0019] The technical solution of the present invention will now be clearly and completely described in conjunction with the accompanying drawings.

[0020] Please see Figure 1 This invention provides a technical solution: A method for dynamic traceability verification of cold chain logistics transportation includes the following steps: S1. Collect the temperature at different locations inside the carriage using temperature sensors, calculate the average temperature and the measured temperature gradient inside the carriage, and upload the average temperature. S2. Collect external environmental data; S3. Divide the driving trajectory into multiple detection segments, and the driving direction of the vehicle is different in two adjacent detection segments; S4. Based on the mapping relationship between external environmental data and the measured spatial temperature gradient of the vehicle under steady state in the current detection section, establish a prediction model to predict the temperature gradient inside the vehicle compartment. S5. Calculate the Euclidean distance between the measured spatial temperature gradient and the predicted temperature gradient, and use it as the first temperature difference. S6. When the first temperature difference exceeds the set threshold, upload complete temperature detection data; if the temperature sensor is not faulty, increase the acquisition frequency. S7. When the vehicle enters the next detection segment, the prediction model corresponding to the previous detection segment is used to calculate the reference temperature gradient, calculate the second temperature difference between the spatial temperature gradient and the reference temperature gradient, and record the transition time required for the second temperature difference to reach a steady state. S8. If the transition time is less than the set first time threshold, a low confidence flag will be uploaded at the current time node; if the transition time is greater than the set second time threshold, complete temperature detection data will be uploaded and the collection frequency will be increased. S9. After the second temperature difference reaches a steady state, repeat S4-S9.

[0021] In step S1, the average temperature and the measured temperature gradient inside the carriage are calculated using the following formula: ; ; ; In the formula, Ti(t) is the data of the i-th temperature sensor at time t, Xi is the x-coordinate of the i-th temperature sensor, and n is the number of temperature sensors.

[0022] Preferably, the measured spatial temperature gradient is the temperature gradient along the width of the carriage.

[0023] In step S2, the external environmental data includes: the temperature outside the carriage. Solar radiation intensity Driving direction Speed and solar azimuth .

[0024] In step S3, the angle between the driving directions on two adjacent detection road segments shall not be less than 10°-30°.

[0025] In step S4, the prediction model is: ; ; ; In the formula, W is the weight matrix, and b is the bias. The relative angle of incidence of the sun at the current moment. , Characterizes the distribution of solar radiation in the left-right / front-back direction of the carriage; The weight matrix and bias are calculated using the least squares method: ; In the formula, To measure the space temperature gradient, λ is the regularization coefficient.

[0026] The data was collected under conditions where the cooling power remained stable (the compressor ran continuously and the frequency fluctuation was <5%).

[0027] In step S6, the following steps are used to determine if the temperature sensor is faulty: A1. Calculate the rate of temperature change for each temperature sensor within a set time window; A2. Calculate the mean and standard deviation of the temperature change rate for all temperature sensors; A3. Temperature sensors whose temperature change rate differs from the mean temperature change rate by more than three times the standard deviation are considered faulty.

[0028] In step S8, the first time threshold is calculated in the following way: ; ; ; In the formula, Let Veff be the temperature difference between the inside and outside of the carriage, heff be the effective heat capacity volume of the carriage, Aeff be the equivalent heat transfer coefficient of the carriage, and τ be the equivalent heat transfer area of ​​the carriage. air density, τmin, α1, and α2 are the specific heat capacity of air, and are determined through experimental calibration.

[0029] In step S8, the second time threshold is calculated in the following way:

[0030]

[0031] τmax,β1,β2 were determined through experimental calibration.

[0032] During cold chain transportation, vehicles may maintain a fixed direction of travel for extended periods. The small temperature fluctuations captured by the temperature sensor pose a significant risk of data manipulation, leading to a decrease in the reliability of the prediction model's results over time. Therefore, the reliability of the prediction model needs to be tested. Step S4 also includes: S401. The confidence level of the prediction model is calculated using the following formula; ; In the formula, t is the current time, ts is the update time of the prediction model parameters, and τm is the confidence half-life; S402. When the reliability of the prediction model is lower than the set reliability threshold, a controllable thermal disturbance is applied to the carriage. S403. Calculate the duration of the disturbance from the start of the controllable thermal disturbance to the time when the space temperature gradient reaches a steady state. S404. If the duration of the disturbance is less than the set first disturbance threshold, then upload the low confidence flag at the current node. S404. If the duration of the disturbance exceeds the set second disturbance threshold, upload the complete temperature detection data and increase the acquisition frequency.

[0033] In step S402, the controllable thermal disturbance is the short-term activation of the defrost heater; the first disturbance threshold and the second disturbance threshold are calibrated based on the activation duration of the defrost heater and the thermal inertia of the carriage thermal system.

[0034] In step S403, when the rate of change of the space temperature gradient is less than 5% for 60 consecutive seconds, the space temperature gradient is considered to have reached a steady state.

[0035] If the duration of the disturbance is less than the set first disturbance threshold, it means that the controllable thermal disturbance was not detected by the temperature sensor, indicating that the temperature sensor may be blocked or the temperature data may be falsified, resulting in low reliability of the temperature data.

[0036] If the duration of the disturbance exceeds the set second disturbance threshold, it indicates a problem with the refrigeration or insulation system inside the vehicle.

Claims

1. A method for dynamic traceability verification of cold chain logistics transportation, characterized in that: Includes the following steps: S1. Collect the temperature at different locations inside the carriage using temperature sensors, calculate the average temperature and the measured temperature gradient inside the carriage, and upload the average temperature. S2. Collect external environmental data; S3. Divide the driving trajectory into multiple detection segments, and the driving direction of the vehicle is different in two adjacent detection segments; S4. Based on the mapping relationship between external environmental data and the measured spatial temperature gradient of the vehicle under steady state in the current detection section, establish a prediction model to predict the temperature gradient inside the vehicle compartment. S5. Calculate the Euclidean distance between the measured spatial temperature gradient and the predicted temperature gradient, and use it as the first temperature difference. S6. When the first temperature difference exceeds the set threshold, upload complete temperature detection data; if the temperature sensor is not faulty, increase the acquisition frequency. S7. When the vehicle enters the next detection segment, the prediction model corresponding to the previous detection segment is used to calculate the reference temperature gradient, calculate the second temperature difference between the spatial temperature gradient and the reference temperature gradient, and record the transition time required for the second temperature difference to reach a steady state. S8. If the transition time is less than the set first time threshold, a low confidence flag will be uploaded at the current time node; if the transition time is greater than the set second time threshold, complete temperature detection data will be uploaded and the collection frequency will be increased. S9. After the second temperature difference reaches a steady state, repeat S4-S9.

2. The method for dynamic traceability verification of cold chain logistics transportation according to claim 1, characterized in that: In step S1, the measured spatial temperature gradient is the temperature gradient along the width of the carriage.

3. The method for dynamic traceability verification of cold chain logistics transportation according to claim 1, characterized in that: In step S2, the external environmental data includes: the temperature outside the carriage, the intensity of solar radiation, the direction of travel, the vehicle speed, and the solar azimuth angle.

4. The method for dynamic traceability verification of cold chain logistics transportation according to claim 1, characterized in that: In step S3, the angle between the driving directions on two adjacent detection road segments shall not be less than 10°-30°.

5. The method for dynamic traceability verification of cold chain logistics transportation according to claim 1, characterized in that: In step S4, the prediction model is: ； ； ； In the formula, W is the weight matrix, and b is the bias. The relative angle of incidence of the sun at the current moment; The weight matrix and bias are calculated using the least squares method.

6. The method for dynamic traceability verification of cold chain logistics transportation according to claim 1, characterized in that: In step S6, the following steps are used to determine if the temperature sensor is faulty: A1. Calculate the rate of temperature change for each temperature sensor within a set time window; A2. Calculate the mean and standard deviation of the temperature change rate for all temperature sensors; A3. Temperature sensors whose temperature change rate differs from the mean temperature change rate by more than three times the standard deviation are considered faulty.

7. The method for dynamic traceability verification of cold chain logistics transportation according to claim 1, characterized in that: In step S8, the first time threshold is calculated in the following way: ； ； ； In the formula, Let Veff be the temperature difference between the inside and outside of the carriage, heff be the effective heat capacity volume of the carriage, Aeff be the equivalent heat transfer coefficient of the carriage, and τ be the equivalent heat transfer area of ​​the carriage. air density, τmin, α1, and α2 are the specific heat capacity of air, and are determined through experimental calibration.

8. The method for dynamic traceability verification of cold chain logistics transportation according to claim 7, characterized in that: In step S8, the second time threshold is calculated in the following way: ； ； τmax,β1,β2 were determined through experimental calibration.

9. The method for dynamic traceability verification of cold chain logistics transportation according to claim 1, characterized in that: Step S4 also includes: S401. The confidence level of the prediction model is calculated using the following formula; ； In the formula, t is the current time, ts is the update time of the prediction model parameters, and τm is the confidence half-life; S402. When the reliability of the prediction model is lower than the set reliability threshold, a controllable thermal disturbance is applied to the carriage. S403. Calculate the duration of the disturbance from the start of the controllable thermal disturbance to the time when the space temperature gradient reaches a steady state. S404. If the duration of the disturbance is less than the set first disturbance threshold, then upload the low confidence flag at the current node. S404. If the duration of the disturbance exceeds the set second disturbance threshold, upload the complete temperature detection data and increase the acquisition frequency.

Images