Food traceability integrated intelligent electronic seal system

Abstract

The invention relates to the technical field of food seals, in particular to a food traceability integrated intelligent electronic seal system which comprises an optical seal monitoring module, a smell monitoring module, a data acquisition and storage module and an abnormity alarm module. The optical seal monitoring module monitors the physical change of the surface of the seal in real time and identifies the early damage of the seal; the smell monitoring module is used for detecting the concentration change of volatile organic compounds in a food package in real time and predicting whether the food is in an early stage of decay or deterioration or not; the data acquisition module is responsible for acquiring monitoring data in real time; and the abnormity alarm module is used for immediately sending out an alarm signal when monitoring that the seal is damaged or the food smell is abnormal based on the acquired monitoring data. According to the invention, the system can find problems in time at the early stage of seal damage or food deterioration, and the monitoring capability of food package integrity and safety is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of food sealing, and in particular to a food traceability integrated intelligent electronic sealing system. Background Art

[0002] With the continuous expansion and complexity of the global food supply chain, food safety and traceability have become the focus of attention of consumers and enterprises. During the production, transportation and storage of food, food quality control is particularly important. Especially under the conditions of long-distance transportation and long-term storage, the integrity of food packaging and changes in the internal environment directly affect the safety and quality of food. Therefore, more and more companies are adopting smart sealing technology and odor monitoring technology to ensure that the status of food in the entire supply chain is controllable and can be traced throughout the entire process.

[0003] There are already some food safety monitoring systems based on electronic seals on the market, but most of these systems focus on simple monitoring of the physical state of the seal, such as whether the seal is damaged. However, these technologies lack a detailed assessment of the degree of seal damage and cannot effectively identify early damage to the seal, resulting in difficulty in timely detection of food packaging problems. In addition, existing food odor monitoring technologies are limited to the collection and monitoring of gas concentrations and lack an effective corruption prediction model, making it difficult to provide timely warnings in the early stages of food corruption. Overall, existing smart seals and odor monitoring systems lack accuracy, real-time capabilities, and data integration capabilities in terms of food safety assurance, and are unable to meet the high standards of the modern food supply chain.

[0004] The present invention aims to provide a food traceability integrated intelligent electronic sealing system, which can efficiently integrate the data of seal and food status, provide accurate judgment basis for the abnormal alarm module, thereby greatly improving the safety, transparency and traceability of food transportation and storage. Summary of the invention

[0005] The present invention provides a food traceability integrated intelligent electronic sealing system.

[0006] The food traceability integrated intelligent electronic seal system includes an optical seal monitoring module, an odor monitoring module, a data acquisition module, and an abnormal alarm module, among which;

[0007] The optical seal monitoring module monitors the physical changes of the seal surface in real time and identifies early damage of the seal by analyzing the physical changes of the seal surface, specifically including:

[0008] Optical data acquisition: The optical sensor captures the image data of the seal surface in real time, including tiny cracks, color changes and surface damage on the seal surface;

[0009] Physical change analysis: Analyze physical changes on the seal surface, including color, texture, and structural anomalies, using image processing algorithms;

[0010] Damage degree identification: Based on the identified abnormal results, the damage degree of the seal is evaluated through a multi-level damage assessment model, and the health status of the seal is classified and marked;

[0011] The odor monitoring module detects the concentration changes of volatile organic compounds (VOCs) inside the food packaging in real time to predict whether the food is in the early stage of corruption or deterioration;

[0012] The data acquisition module is responsible for real-time acquisition of monitoring data, including the seal status and the internal environment of the food packaging;

[0013] The abnormal alarm module, based on the collected monitoring data, immediately sends out an alarm signal when it detects that the seal is damaged or the food smells abnormal.

[0014] Optionally, the optical data acquisition includes:

[0015] Optical sensor start: Start the integrated optical sensor to monitor the seal surface in real time at preset time intervals or in continuous mode, ready to capture its image data;

[0016] Capturing physical changes in seals: Using multispectral imaging technology, the reflection and scattered light data of the seal surface are obtained under visible light and near-infrared, respectively. By applying light of different wavelengths, tiny cracks, color changes and surface damage on the seal surface can be captured.

[0017] Optionally, the physical change analysis includes:

[0018] Color analysis: Using color differences based on RGB color space to calculate the color changes on the seal surface at different times;

[0019] Texture analysis: Identify changes in seal surface texture using a texture analysis algorithm based on the Gray Level Co-occurrence Matrix (GLCM);

[0020] Structural anomaly detection: An edge detection algorithm based on gradient changes (Sobel operator) is used to analyze the structural changes on the seal surface.

[0021] Optionally, the multi-level damage assessment model adopts a Bayesian network model, and the Bayesian network model includes:

[0022] Feature node modeling: For different physical changes of the seal, corresponding Bayesian network nodes are created. Each node represents a physical change of the seal, including color, texture and structural abnormalities, expressed as:

[0023] P(X1,X2,X3)=P(X1|Pa(X1))P(X2|Pa(X2))P(X3|Pa(X3));

[0024] Where X1 represents color change, X2 represents texture change, and X3 represents structural abnormality. Pa(X1), Pa(X2), and Pa(X3) are the parent nodes of nodes X1, X2, and X3, respectively. P(X1|Pa(X1)), P(X2|Pa(X2)), and P(X3|Pa(X3)) are the conditional probabilities of color change, texture change, and structural abnormality, respectively. P(X1, X2, X3) is the joint probability.

[0025] Introducing damage state weights: According to different physical changes, different weights are assigned to each damage state to indicate the impact on the degree of damage, expressed as:

[0026]

[0027] Where S represents the damage state of the seal (such as intact, slightly damaged, severely damaged), w1, w2, w3 are the corresponding weights, Z is the normalization constant, and P(S|X1, X2, X3) is the conditional probability of the damage state;

[0028] Dynamic update of damage state reasoning: Dynamically update the conditional probability table (CPT) in the Bayesian network so that the damage state reasoning is continuously optimized over time. The dynamic update mechanism is used to automatically adjust the judgment of the damage degree according to the newly collected image data, which is expressed as:

[0029]

[0030] Among them, S t is the damage state at the current time t, X 1：t is the observed value of the seal feature from time 1 to time t, H 1：t is the impact of historical data on the current damage state, P(X 1：t , H 1：t ) is the current damage state inferred based on current and historical data, P(S t |X 1：t , H 1：t ) is the conditional probability of the current damage state, P(S t-1 ) is the damage state at the previous moment, P(X t |S t ) is the conditional probability of the current feature, P(H t |S t ) is the conditional probability of historical data;

[0031] Classification and labeling of seal health status: Based on the evaluation results of different physical changes, the health status of the seal is divided into multiple categories, including intact, slightly damaged, and severely damaged. By calculating the contribution of each physical change to the damage status, the Bayesian network labels the health status of the seal according to the maximum a posteriori probability (MAP), which is expressed as:

[0032] S * = argmax S P(S|X1,X2,X3);

[0033] Among them, S* is the optimal damage state classification result derived based on the maximum a posteriori probability.

[0034] Optionally, the odor monitoring module includes:

[0035] Gas collection: Gas sensors are used to collect real-time data on the concentration of volatile organic compounds (VOCs) inside food packaging, including ethylene and carbon dioxide;

[0036] Gas data preprocessing: preprocessing the collected volatile organic compound (VOCs) concentration data, including noise filtering, data smoothing and abnormal data identification;

[0037] Odor analysis and prediction: Through the intelligent spoilage detection model, the pre-processed gas data is analyzed and predicted whether the food is in the early stages of spoilage or deterioration.

[0038] Optionally, the gas collection includes:

[0039] Gas molecule absorption signal detection: Detect volatile organic compounds (VOCs), including ethylene and carbon dioxide, inside food packaging through gas sensors;

[0040] Signal conversion and concentration calculation: convert the detected gas absorption or reaction signal into an electrical signal and calculate the concentration value of volatile organic compounds (VOCs) gas;

[0041] Gas concentration change monitoring: Continuously detect real-time changes in gas concentration, track fluctuations in ethylene and carbon dioxide concentrations over time, and record rising or falling trends in gas concentrations.

[0042] Optionally, the gas data preprocessing includes:

[0043] Noise filtering: Kalman filtering algorithm is used to filter the noise of volatile organic compound (VOCs) concentration data;

[0044] Data smoothing: Use the sliding average algorithm to smooth the volatile organic compound (VOCs) concentration data;

[0045] Abnormal data identification: An abnormal value in the data is identified using an anomaly detection algorithm based on Z-score.

[0046] Optionally, the intelligent corruption detection model adopts a Gaussian mixture model (GMM), and the Gaussian mixture model (GMM) includes:

[0047] Defining the probability distribution of gas characteristics: The Gaussian mixture model fits the gas concentration data through multiple Gaussian distributions (normal distributions), expressed as:

[0048]

[0049] Where P(x) is the probability density function of the gas characteristic x, π k is the mixing coefficient of the kth Gaussian distribution, is the kth Gaussian distribution with mean μ k and the covariance matrix ∑ k , K is the number of Gaussian distributions;

[0050] Estimate the mean and variance of the gas signature: Calculate the mean μ of each Gaussian distribution k and the covariance matrix ∑ k , determine the distribution characteristics of different gas states, expressed as:

[0051]

[0052] Where d is the dimension number of gas characteristics;

[0053] Calculate the responsibility of each gas feature belonging to the corresponding Gaussian distribution: For each gas feature x, calculate the responsibility of the feature belonging to the corresponding Gaussian distribution k, which represents the probability that the feature is generated by the corresponding Gaussian distribution, expressed as:

[0054]

[0055] Among them, γ(z k ) is the responsibility of the gas characteristic x to belong to the kth Gaussian distribution, is the probability density of the kth Gaussian distribution, is the sum of the weighted probability densities of all Gaussian distributions;

[0056] Maximum likelihood estimation for parameter update: Use maximum likelihood estimation (MLE) to update the parameters of each Gaussian distribution, expressed as:

[0057]

[0058] Among them, μ k is the mean of the kth Gaussian distribution, ∑ k is the covariance matrix of the kth Gaussian distribution, πk is the mixing coefficient of the kth Gaussian distribution, is the responsibility of the i-th data point to belong to the k-th Gaussian distribution, x (i) is the i-th observation data (gas characteristics), N is the total number of observation data;

[0059] Classification and prediction of corruption status: The calculated responsibility of each gas feature predicts the current corruption status of the food and classifies it, which is expressed as:

[0060] S′=argmax k γ(z k );

[0061] Among them, S′ is the optimal corruption state of the current gas characteristics, γ(z k ) is the responsibility of the kth Gaussian distribution.

[0062] Optionally, the data acquisition module includes:

[0063] Seal status collection: Based on the recognition results of early seal damage, seal status data is collected, including intact labels, slightly damaged labels, and severely damaged labels;

[0064] Collection of internal environment of food packaging: Based on the prediction results of food corruption or deterioration in the early stages, the internal environment data of food packaging is collected, including food safety, slight food corruption or severe food corruption.

[0065] Optionally, the abnormal alarm module includes:

[0066] Abnormal judgment: Based on the collected monitoring data, judge the seal status and food status, including whether the seal is slightly damaged or severely damaged, and whether the food is slightly corrupt or severely corrupt;

[0067] Alarm triggering: When the abnormal judgment result shows that the seal is slightly damaged or seriously damaged, or the food is slightly corrupted or seriously corrupted, the alarm signal is immediately triggered;

[0068] Alarm signal processing: Transmitted to the user end or management platform through wireless communication technology, informing relevant personnel to take corresponding measures, including checking and replacing seals, food quality inspection and processing, adjustment of sealing and storage conditions, and recording alarm information.

[0069] Beneficial effects of the present invention:

[0070] The present invention realizes comprehensive real-time monitoring of the seal status and the internal environment of the food package by integrating an optical seal monitoring module and an odor monitoring module. Through optical data acquisition and image processing technology, it can accurately identify tiny physical changes on the seal surface, such as cracks, color changes and structural damage. Combined with a multi-level damage assessment model, the health status of the seal is comprehensively assessed. At the same time, the odor monitoring module detects the concentration of volatile organic compounds (VOCs) inside the food package in real time, and accurately predicts the early signs of food corruption based on the Gaussian mixture model (GMM), ensuring that the system can promptly detect problems in the early stages of seal damage or food spoilage, thereby improving the monitoring capability of food packaging integrity and safety.

[0071] The present invention ensures efficient management and rapid response of monitoring data through the combination of a data acquisition module and an abnormal alarm module. The data acquisition module provides accurate classification information of seal and food status based on the identification results of seal damage and food corruption, such as intact, slightly damaged or severely damaged labels, food safety, slightly corrupt or severely corrupt, etc., thereby providing a basis for real-time monitoring for the abnormal alarm module. When a seal damage or abnormal food odor is detected, the abnormal alarm module can immediately send out an alarm signal to ensure that relevant personnel can receive early warning information in time and take corresponding measures to prevent potential safety problems from further expanding.

[0072] The present invention, through accurate identification of the health status of the seal and the risk of food spoilage, can take protective measures at an early stage to avoid seal damage leading to packaging failure or food spoilage. The combination of the intelligent early warning system and real-time data collection not only improves the management efficiency of food safety, but also enhances the transparency of the supply chain, effectively prevents the occurrence of food quality problems, and ensures the safety and traceability of food. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0074] Figure 1 A schematic diagram of system function modules according to an embodiment of the present invention;

[0075] Figure 2 Schematic diagram of an optical seal monitoring module according to an embodiment of the present invention. DETAILED DESCRIPTION

[0076] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0077] It should be noted that the references to "one embodiment", "an embodiment", "an exemplary embodiment", "some embodiments" and the like in the specification indicate that the embodiments described may include specific features, structures or characteristics, but not every embodiment may include the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the art to implement such feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0078] In general, a term can be understood, at least in part, from its use in context. For example, depending, at least in part, on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending, at least in part, on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0079] like Figure 1-Figure 2 As shown, the food traceability integrated intelligent electronic sealing system includes an optical sealing monitoring module, an odor monitoring module, a data acquisition module and an abnormal alarm module, wherein;

[0080] The optical seal monitoring module monitors the physical changes on the seal surface in real time and identifies early damage to the seal by analyzing the physical changes on the seal surface. Specifically, it includes:

[0081] Optical data acquisition: The optical sensor captures the image data of the seal surface in real time, including tiny cracks, color changes and surface damage on the seal surface;

[0082] Physical change analysis: Analyze physical changes on the seal surface, including color, texture, and structural anomalies, using image processing algorithms;

[0083] Damage degree identification: Based on the identified abnormal results, the damage degree of the seal is evaluated through a multi-level damage assessment model, and the health status of the seal is classified and marked;

[0084] The odor monitoring module detects the concentration changes of volatile organic compounds (VOCs) inside food packaging in real time to predict whether the food is in the early stages of corruption or deterioration;

[0085] The data acquisition module is responsible for real-time collection of monitoring data, including the seal status and the internal environment of the food packaging;

[0086] The abnormal alarm module, based on the collected monitoring data, will immediately send out an alarm signal when it detects that the seal is damaged or the food smells abnormal;

[0087] Through the above content, early signs of seal damage and food spoilage can be identified in a timely manner, the integrity and security of all monitoring data can be ensured, and alarms can be issued immediately when abnormal situations occur, thereby improving the safety and transparency of food transportation and storage, effectively preventing food quality problems, and ensuring the safety and reliability of the supply chain.

[0088] Optical data acquisition includes:

[0089] Optical sensor start: Start the integrated optical sensor to monitor the seal surface in real time at preset time intervals or in continuous mode, ready to capture its image data;

[0090] Capturing physical changes in seals: Using multispectral imaging technology, the reflected and scattered light data of the seal surface is obtained under visible light and near-infrared light. By applying light of different wavelengths, tiny cracks, color changes and surface damage on the seal surface can be captured;

[0091] Through the above, it is possible to accurately capture tiny physical changes on the seal surface, including cracks, color changes and surface damage. By collecting data in the visible and near-infrared bands, it is ensured that subtle changes in the seal status can be fully monitored under different spectral conditions.

[0092] Physical change analysis includes:

[0093] Color analysis: The color difference based on RGB color space is used to calculate the color change of the seal surface at different times, which is expressed as:

[0094]

[0095] Among them, ΔE is the color difference value, which indicates the amount of color change on the seal surface at different time points, and R1, G1, B1 and R2, G2, B2 represent the red, green, and blue channel values ​​at different positions on the seal surface in the initial state and the current state, respectively;

[0096] Texture analysis: The texture changes of the seal surface are identified by using a texture analysis algorithm based on the gray level co-occurrence matrix (GLCM), which is expressed as:

[0097]

[0098] Where P(i, j) is the probability of gray levels i and j appearing in the gray level co-occurrence matrix, and N is the number of gray levels;

[0099] Structural anomaly detection: The edge detection algorithm based on gradient change (Sobel operator) is used to analyze the structural changes on the seal surface, which can be expressed as:

[0100]

[0101] Among them, G x and G y They are the gradient changes in the horizontal and vertical directions respectively;

[0102] Through the above content, the seal surface can be accurately monitored in multiple dimensions, which improves the monitoring accuracy of the seal status and the ability to identify early damage, helps to take protective measures in time, and ensures the safety and seal integrity of food during transportation and storage.

[0103] The multi-level damage assessment model adopts the Bayesian network model, which includes:

[0104] Feature node modeling: For different physical changes of the seal, corresponding Bayesian network nodes are created. Each node represents a physical change of the seal, including color, texture and structural abnormalities, expressed as:

[0105] P(X1,X2,X3)=P(X1|Pa(X1))P(X2|Pa(X2))P(X3|Pa(X3));

[0106] Where X1 represents color change, X2 represents texture change, and X3 represents structural abnormality. Pa(X1), Pa(X2), and Pa(X3) are the parent nodes of nodes X1, X2, and X3, respectively, indicating the influencing factors or upstream features related to the feature. P(X1|Pa(X1)), P(X2|Pa(X2)), and P(X3|Pa(X3)) are the conditional probabilities of color change, texture change, and structural abnormality, respectively. P(X1, X2, X3) is the joint probability.

[0107] Introducing damage state weights: According to different physical changes, different weights are assigned to each damage state to indicate the impact on the degree of damage, expressed as:

[0108]

[0109] Where S represents the damage state of the seal (such as intact, slightly damaged, severely damaged), w1, w2, w3 are the corresponding weights, Z is the normalization constant, and P(S|X1, X2, X3) is the conditional probability of the damage state;

[0110] Dynamic update of damage state reasoning: Dynamically update the conditional probability table (CPT) in the Bayesian network so that the damage state reasoning is continuously optimized over time. The dynamic update mechanism is used to automatically adjust the judgment of the damage degree according to the newly collected image data, which is expressed as:

[0111]

[0112] Among them, S t is the damage state at the current time t, X 1：t is the observed value of the seal feature from time 1 to time t, H 1：t is the impact of historical data on the current damage state, P(X 1：t , H 1：t ) is the current damage state inferred based on current and historical data, P(S t |X 1：t , H 1：t ) is the conditional probability of the current damage state, P(S t-1 ) is the damage state at the previous moment, P(X t |S t ) is the conditional probability of the current feature, P(H t |S t ) is the conditional probability of historical data;

[0113] Classification and labeling of seal health status: Based on the evaluation results of different physical changes, the health status of the seal is divided into multiple categories, including intact, slightly damaged, and severely damaged. By calculating the contribution of each physical change to the damage status, the Bayesian network labels the health status of the seal according to the maximum a posteriori probability (MAP), which is expressed as:

[0114] S * = argmax S P(S|X1,X2,X3);

[0115] Among them, S * is the optimal damage state classification result derived based on the maximum a posteriori probability;

[0116] The setting basis and value range of weights w1, w2, and w3 specifically include:

[0117] The color change weight w1 is usually a visual change caused by early damage to the seal or environmental influences (such as sunlight, humidity), which usually does not directly lead to serious physical damage. Depending on the seal material and application scenario, the color change may be regarded as an early warning sign of minor damage, and the weight can be lower, with a value of 0.2;

[0118] The texture change weight w2 usually reflects physical changes such as wear, fine cracks or corrosion on the seal surface. This change may indicate that physical damage to the seal begins to appear in the mid-term stage, so the weight should be higher than the color change and the value is 0.5;

[0119] The structural anomaly weight w3 is usually a sign of severe physical damage to the seal, which may indicate that the seal has lost its function or is close to failure. Structural damage is the most important feature in the seal status assessment, so the weight should be set to the highest, with a value of 0.9.

[0120] Through the above content, the multi-dimensional characteristics of the seal (color, texture, structural abnormality) can be comprehensively considered, and through the dynamic update mechanism, the damage status of the seal can be accurately evaluated based on the combination of real-time monitoring data and historical data. After the weight mechanism is introduced, the importance of different features to damage assessment is reasonably distributed, ensuring the dominant role of key features such as structural abnormalities in the assessment. In addition, the model has good adaptability and interpretability, and can infer the health status of the seal in real time, continuously optimize the evaluation results over time, and timely detect the damage trend of the seal, thereby improving the accuracy and reliability of seal monitoring.

[0121] The odor monitoring module includes:

[0122] Gas collection: Gas sensors are used to collect real-time data on the concentration of volatile organic compounds (VOCs) inside food packaging, including ethylene and carbon dioxide;

[0123] Gas data preprocessing: preprocessing the collected volatile organic compound (VOCs) concentration data, including noise filtering, data smoothing and abnormal data identification;

[0124] Odor analysis and prediction: Through intelligent spoilage detection models, pre-processed gas data is analyzed and predicted whether food is in the early stages of spoilage or deterioration;

[0125] Through the above content, food quality problems can be warned in time, the accuracy and response speed of food safety monitoring are improved, and it is helpful to take protective measures in the early stages of corruption.

[0126] Gas collection includes:

[0127] Gas molecule absorption signal detection: Detect volatile organic compounds (VOCs), including ethylene and carbon dioxide, inside food packaging through gas sensors;

[0128] Signal conversion and concentration calculation: The detected gas absorption or reaction signal is converted into an electrical signal, and the concentration value of the volatile organic compound (VOCs) gas is calculated, which is expressed as:

[0129]

[0130] Where c is the gas concentration (VOCs, such as ethylene and carbon dioxide), I is the current intensity measured by the electrochemical sensor, and k is the sensitivity constant of the sensor;

[0131] Gas concentration change monitoring: Continuously detect the real-time changes in gas concentration, track the fluctuations of ethylene and carbon dioxide concentrations over time, and record the trend of gas concentration increase or decrease;

[0132] Through the above content, the concentration of volatile organic compounds (VOCs) inside food packaging can be accurately detected with high real-time and accuracy, ensuring high sensitivity and real-time response of monitoring, which helps to timely warn of possible food spoilage risks and improve the efficiency and reliability of food safety monitoring.

[0133] Gas data preprocessing includes:

[0134] Noise filtering: Kalman filtering algorithm is used to filter the noise of volatile organic compound (VOCs) concentration data, which is expressed as:

[0135]

[0136] in, is the estimated value at the current moment, is the estimated value at the previous moment, z k is the measured value (VOCs concentration) at the current moment, K k is the Kalman gain, which is used to weigh the measured data and the predicted data;

[0137] Data smoothing: The volatile organic compound (VOCs) concentration data is smoothed using a sliding average algorithm, expressed as:

[0138]

[0139] in, is the smoothing result at the nth moment, N is the window size, indicating the number of data points involved in the smoothing calculation, x i is the original data point;

[0140] Abnormal data identification: An abnormal value in the data is identified using an anomaly detection algorithm based on Z-score, which is expressed as:

[0141]

[0142] Among them, x is the current data point, μ is the average value of VOCs concentration data, and σ is the standard deviation of VOCs concentration data;

[0143] Through the above content, the quality of VOCs concentration data has been improved, making the data more stable and reliable, providing a solid data foundation for subsequent corruption predictions and improving the accuracy and efficiency of system monitoring.

[0144] The intelligent corruption detection model uses the Gaussian mixture model (GMM), which includes:

[0145] Defining the probability distribution of gas characteristics: The Gaussian mixture model fits the gas concentration data through multiple Gaussian distributions (normal distributions), expressed as:

[0146]

[0147] Where P(x) is the probability density function of the gas characteristic x, π k is the mixing coefficient of the kth Gaussian distribution, which represents the weight of each Gaussian distribution in the whole and satisfies is the kth Gaussian distribution with mean μ k and the covariance matrix ∑ k , K is the number of Gaussian distributions, used to represent different gas states;

[0148] Estimate the mean and variance of the gas signature: Calculate the mean μ of each Gaussian distribution k and the covariance matrix ∑ k , determine the distribution characteristics of different gas states, expressed as:

[0149]

[0150] Where d is the dimension number of gas characteristics;

[0151] Calculate the responsibility of each gas feature belonging to the corresponding Gaussian distribution: For each gas feature x, calculate the responsibility of the feature belonging to the corresponding Gaussian distribution k, which represents the probability that the feature is generated by the corresponding Gaussian distribution, expressed as:

[0152]

[0153] Among them, γ(z k ) is the responsibility of the gas characteristic x to belong to the kth Gaussian distribution, is the probability density of the kth Gaussian distribution, is the sum of the weighted probability densities of all Gaussian distributions;

[0154] Maximum likelihood estimation for parameter update: Use maximum likelihood estimation (MLE) to update the parameters of each Gaussian distribution, expressed as:

[0155]

[0156] Among them, μ k is the mean of the kth Gaussian distribution, ∑ k is the covariance matrix of the kth Gaussian distribution, π k is the mixing coefficient of the kth Gaussian distribution, is the responsibility of the i-th data point to belong to the k-th Gaussian distribution, x (i) is the i-th observation data (gas characteristics), N is the total number of observation data;

[0157] Classification and prediction of corruption status: The calculated responsibility of each gas feature predicts the current corruption status of the food and classifies it, which is expressed as:

[0158] S′=argmax k γ(z k );

[0159] Among them, S' is the optimal corruption state of the current gas characteristics, γ(z k ) is the responsibility degree of the kth Gaussian distribution;

[0160] Through the above, the changing trends of gas characteristics are adaptively captured, so as to detect the corruption or deterioration of food earlier and more accurately. Through these characteristics, GMM improves the accuracy and real-time response capability of corruption detection, ensuring the reliability of food safety monitoring.

[0161] The data acquisition module includes:

[0162] Seal status collection: Based on the recognition results of early seal damage, seal status data is collected, including intact labels, slightly damaged labels, and severely damaged labels;

[0163] Collection of food packaging internal environment: Based on the prediction results of food spoilage or deterioration in the early stages, the internal environment data of food packaging is collected, including food safety, slight food spoilage or severe food spoilage;

[0164] Through the above content, based on the results of early seal damage identification and food spoilage prediction, accurate classification information of seal status and food packaging environment is provided.

[0165] The abnormal alarm module includes:

[0166] Abnormal judgment: Based on the collected monitoring data, judge the seal status and food status, including whether the seal is slightly damaged or severely damaged, and whether the food is slightly corrupt or severely corrupt;

[0167] Alarm triggering: When the abnormal judgment result shows that the seal is slightly damaged or seriously damaged, or the food is slightly corrupted or seriously corrupted, the alarm signal is immediately triggered;

[0168] Alarm signal processing: transmit to the user end or management platform through wireless communication technology, notify relevant personnel to take corresponding measures, including checking and replacing seals, food quality inspection and processing, adjustment of sealing and storage conditions, and record alarm information for subsequent analysis and tracking;

[0169] Through the above content, it is possible to respond quickly to potential safety issues and provide clear guidance to relevant personnel (such as seal replacement, food quality testing and storage condition adjustment), thereby minimizing losses and improving the efficiency and reliability of food safety management. At the same time, the alarm module's instant warning function ensures that problems are resolved at an early stage to avoid further expansion of potential food safety risks.

[0170] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0171] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. Food traceability integrated intelligent electronic sealing system, characterized by: It includes an optical seal monitoring module, an odor monitoring module, a data acquisition module and an abnormal alarm module, among which; The optical seal monitoring module monitors the physical changes of the seal surface in real time and identifies early damage of the seal by analyzing the physical changes of the seal surface, specifically including: Optical data acquisition: The optical sensor captures the image data of the seal surface in real time, including tiny cracks, color changes and surface damage on the seal surface; Physical change analysis: Analyze physical changes on the seal surface, including color, texture, and structural anomalies, using image processing algorithms; Damage degree identification: Based on the identified abnormal results, the damage degree of the seal is evaluated through a multi-level damage assessment model, and the health status of the seal is classified and marked; The odor monitoring module detects the concentration changes of volatile organic compounds inside the food packaging in real time to predict whether the food is in the early stage of corruption or deterioration; The data acquisition module is responsible for real-time acquisition of monitoring data, including the seal status and the internal environment of the food packaging; The abnormal alarm module, based on the collected monitoring data, immediately sends out an alarm signal when it detects that the seal is damaged or the food smells abnormal.

2. The food traceability integrated intelligent electronic seal system according to claim 1, characterized in that: The optical data acquisition comprises: Optical sensor start: Start the integrated optical sensor to monitor the seal surface in real time at preset time intervals or in continuous mode, ready to capture its image data; Capturing physical changes in seals: Using multispectral imaging technology, the reflection and scattered light data of the seal surface are obtained under visible light and near-infrared, respectively. By applying light of different wavelengths, tiny cracks, color changes and surface damage on the seal surface can be captured.

3. The food traceability integrated intelligent electronic sealing system according to claim 2, characterized in that: The physical change analysis includes: Color analysis: Using color differences based on RGB color space to calculate the color changes on the seal surface at different times; Texture analysis: Identify changes in seal surface texture using a texture analysis algorithm based on a gray-level co-occurrence matrix; Structural anomaly detection: An edge detection algorithm based on gradient changes is used to analyze the structural changes on the seal surface.

4. The food traceability integrated intelligent electronic seal system according to claim 3, characterized in that: The multi-level damage assessment model adopts a Bayesian network model, which includes: Feature node modeling: For different physical changes of the seal, corresponding Bayesian network nodes are created. Each node represents a physical change of the seal, including color, texture and structural abnormalities, expressed as: P(X1,X2,X3)=P(X1|Pa(X1))P(X2|Pa(X2))P(X3|Pa(X3)); Where X1 represents color change, X2 represents texture change, X3 represents structural anomaly, Pa(X1), Pa(X2), and Pa(X3) are the parent nodes of nodes X1, X2, and X3 respectively, P(X1|Pa(X1)), P(X2|Pa(X2)), and P(X3|Pa(X3)) are the conditional probabilities of color change, texture change, and structural anomaly respectively, and P(X1,X2,X3) is the joint probability; Introducing damage state weights: According to different physical changes, different weights are assigned to each damage state to indicate the impact on the degree of damage, expressed as: Where S represents the damage state of the seal, w1, w2, w3 are the corresponding weights, Z is the normalization constant, and P(S|X1,X2,X3) is the conditional probability of the damage state; Dynamic update of damage state reasoning: Dynamically update the conditional probability table in the Bayesian network so that the damage state reasoning is continuously optimized over time. The dynamic update mechanism is used to automatically adjust the judgment of the damage degree according to the newly collected image data, which is expressed as: Among them, S t is the damage state at the current time T, X 1:t is the observed value of the seal feature from time 1 to time t, H 1:t is the impact of historical data on the current damage state, P(X 1:t ,H 1:t ) is the current damage state inferred based on current and historical data, P(S t |X 1:t ,H 1:t ) is the conditional probability of the current damage state, P(S t-1 ) is the damage state at the previous moment, P(X t |S t ) is the conditional probability of the current feature, P(H t |S t ) is the conditional probability of historical data; Classification and labeling of seal health status: Based on the evaluation results of different physical changes, the health status of the seal is divided into multiple categories, including intact, slightly damaged, and severely damaged. By calculating the contribution of each physical change to the damage status, the Bayesian network labels the health status of the seal according to the maximum posterior probability, which is expressed as: S * =argmax S P(S|X1,X2,X3); Among them, S * It is the optimal damage state classification result derived based on the maximum a posteriori probability.

5. The food traceability integrated intelligent electronic seal system according to claim 1, characterized in that: The odor monitoring module comprises: Gas collection: Gas sensors are used to collect real-time data on the concentration of volatile organic compounds inside food packaging, including ethylene and carbon dioxide; Gas data preprocessing: preprocessing the collected volatile organic compound concentration data, including noise filtering, data smoothing and abnormal data identification; Odor analysis and prediction: Through the intelligent spoilage detection model, the pre-processed gas data is analyzed and predicted whether the food is in the early stages of spoilage or deterioration.

6. The food traceability integrated intelligent electronic seal system according to claim 5, characterized in that: The gas collection includes: Gas molecule absorption signal detection: Detect volatile organic compounds inside food packaging, including ethylene and carbon dioxide, through gas sensors; Signal conversion and concentration calculation: convert the detected gas absorption or reaction signal into an electrical signal and calculate the concentration value of the volatile organic compound gas; Gas concentration change monitoring: Continuously detect real-time changes in gas concentration, track fluctuations in ethylene and carbon dioxide concentrations over time, and record rising or falling trends in gas concentrations.

7. The food traceability integrated intelligent electronic seal system according to claim 6, characterized in that: The gas data preprocessing includes: Noise filtering: Kalman filtering algorithm is used to filter the noise of volatile organic compound concentration data; Data smoothing: The VOC concentration data were smoothed using a sliding average algorithm; Abnormal data identification: An abnormal value in the data is identified using an anomaly detection algorithm based on Z-score.

8. The food traceability integrated intelligent electronic seal system according to claim 7, characterized in that: The intelligent corruption detection model adopts a Gaussian mixture model, which includes: Defining the probability distribution of gas characteristics: The Gaussian mixture model fits the gas concentration data through multiple Gaussian distributions, expressed as: Where P(x) is the probability density function of the gas characteristic x, π k is the mixing coefficient of the kth Gaussian distribution, is the kth Gaussian distribution with mean μ k and the covariance matrix ∑ k , K is the number of Gaussian distributions; Estimate the mean and variance of the gas signature: Calculate the mean μ of each Gaussian distribution k and the covariance matrix ∑ k , determine the distribution characteristics of different gas states, expressed as: Where d is the dimension number of gas characteristics; Calculate the responsibility of each gas feature belonging to the corresponding Gaussian distribution: For each gas feature x, calculate the responsibility of the feature belonging to the corresponding Gaussian distribution k, which represents the probability that the feature is generated by the corresponding Gaussian distribution, expressed as: Among them, γ(z k ) is the responsibility of the gas characteristic x to belong to the kth Gaussian distribution, is the probability density of the kth Gaussian distribution, is the sum of the weighted probability densities of all Gaussian distributions; Maximum likelihood estimation for parameter update: Use maximum likelihood estimation to update the parameters of each Gaussian distribution, expressed as: Among them, μ k is the mean of the kth Gaussian distribution, ∑ k is the covariance matrix of the kth Gaussian distribution, π k is the mixing coefficient of the kth Gaussian distribution, is the responsibility of the i-th data point to belong to the k-th Gaussian distribution, x (i) is the i-th observation data, and N is the total number of observation data; Classification and prediction of corruption status: The calculated responsibility of each gas feature predicts the current corruption status of the food and classifies it, which is expressed as: S’=argmax k γ(z k ); Among them, S' is the optimal corruption state of the current gas characteristics, γ(z k ) is the responsibility of the kth Gaussian distribution.

9. The food traceability integrated intelligent electronic seal system according to claim 1, characterized in that: The data acquisition module comprises: Seal status collection: Based on the recognition results of early seal damage, seal status data is collected, including intact labels, slightly damaged labels, and severely damaged labels; Collection of internal environment of food packaging: Based on the prediction results of food corruption or deterioration in the early stages, the internal environment data of food packaging is collected, including food safety, slight food corruption or severe food corruption.

10. The food traceability integrated intelligent electronic seal system according to claim 9, characterized in that: The abnormal alarm module comprises: Abnormal judgment: Based on the collected monitoring data, judge the seal status and food status, including whether the seal is slightly damaged or severely damaged, and whether the food is slightly corrupt or severely corrupt; Alarm triggering: When the abnormal judgment result shows that the seal is slightly damaged or seriously damaged, or the food is slightly corrupted or seriously corrupted, the alarm signal is immediately triggered; Alarm signal processing: Transmitted to the user end or management platform through wireless communication technology, informing relevant personnel to take corresponding measures, including checking and replacing seals, food quality inspection and processing, adjustment of sealing and storage conditions, and recording alarm information.

Images