Food cold-chain logistics quality monitoring and early warning method and system

Abstract

The invention relates to the technical field of cold-chain logistics, in particular to a food cold-chain logistics quality monitoring and early warning method and system, and the method comprises the steps: obtaining an original quality data index in food cold-chain logistics through a food quality information multi-sensor collection platform; based on the original quality data indexes, constructing a food cold-chain logistics quality prediction model; according to a food cold-chain logistics quality prediction model, predicting the original quality data index to obtain food quality data; based on the food quality data, constructing a food quality data multi-level analysis index weight model; according to the food quality data multi-level analysis index weight model, calculating a food quality data multi-level analysis index weight; based on the food quality data multilevel analysis index weight, constructing a food cold-chain logistics quality monitoring model based on an analytic hierarchy process; according to the food cold-chain logistics quality monitoring model based on the analytic hierarchy process, abnormity judgment is carried out on food quality, and the accuracy of the prediction model is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of cold chain logistics, and more specifically, to a food cold chain logistics quality monitoring and early warning method and system. Background Art

[0002] With the improvement of people's living standards and the increasing attention to food safety, food cold chain logistics plays an increasingly important role in modern society. However, the current food cold chain logistics quality monitoring and early warning technology still faces many challenges.

[0003] Traditional food cold chain logistics quality monitoring mainly relies on single equipment such as temperature recorders. This method has obvious defects such as single data dimension and poor real-time performance. Although technologies such as GPS positioning and RFID have been introduced in recent years, these technologies are often used in isolation and it is difficult to form a comprehensive quality monitoring system. Some advanced companies have tried to use Internet of Things technology for multi-dimensional data collection, but they are still insufficient in data analysis and early warning.

[0004] The quality prediction models currently used in the industry are mostly based on simple statistical methods or empirical formulas, which are difficult to accurately reflect the impact of the complex and changeable cold chain logistics environment on food quality. Although some researchers have proposed prediction methods based on machine learning, these methods are often only targeted at a specific type of food or a specific environmental parameter, lacking versatility and comprehensiveness.

[0005] In terms of quality assessment, existing technologies mostly use a single threshold or a simple scoring system, which makes it difficult to fully consider the combined impact of multiple factors such as food type, packaging materials, and transportation environment. Some studies have proposed multi-factor assessment models, but the scientificity and operability of the models still need to be improved.

[0006] In addition, most existing early warning systems are passive, that is, they can only issue an alarm after a problem occurs. This method is difficult to effectively prevent the occurrence of quality problems and often causes unnecessary economic losses. Although some forward-looking studies have proposed the concept of active early warning, it still faces problems such as insufficient accuracy and high false alarm rate in actual application.

[0007] In general, the existing food cold chain logistics quality monitoring and early warning technology has the following main problems: insufficient data collection dimensions, making it difficult to fully reflect the food quality status; low accuracy of quality prediction models, making it difficult to cope with complex and changeable cold chain environments; oversimplified quality assessment methods, making it difficult to accurately reflect the combined impact of multiple factors; and delayed early warning mechanisms, making it difficult to effectively prevent quality problems. These problems have seriously restricted the quality management level of food cold chain logistics, and a more comprehensive, intelligent, and efficient quality monitoring and early warning method is urgently needed. Summary of the invention

[0008] The present invention aims to solve the above technical problems and provide a food cold chain logistics quality monitoring and early warning method and system based on multi-dimensional data collection, intelligent prediction analysis and active early warning. This method realizes multi-dimensional data collection of the entire cold chain logistics process by building a multi-sensor acquisition platform; improves the accuracy of quality prediction by introducing machine learning algorithms; realizes a comprehensive assessment of food quality by establishing a multi-level analysis model; and effectively prevents the occurrence of quality problems by designing an active early warning mechanism.

[0009] The present invention provides a food cold chain logistics quality monitoring and early warning method, comprising:

[0010] The acquisition steps include:

[0011] Acquire original quality data indicators in food cold chain logistics through a food quality information multi-sensor acquisition platform, wherein the food quality information multi-sensor acquisition platform includes a cold chain logistics video monitoring sub-platform, a geographic positioning sub-platform, a temperature and humidity sensor sub-platform, and a multi-spectral image acquisition sub-platform;

[0012] Processing steps include:

[0013] Based on the original quality data indicators, a food cold chain logistics quality prediction model is constructed;

[0014] According to the food cold chain logistics quality prediction model, the original quality data indicators are predicted to obtain food quality data;

[0015] Based on the food quality data, a multi-level analysis index weight model for food quality data is constructed;

[0016] Calculating the weights of multi-level analysis indicators of food quality data according to the multi-level analysis indicator weight model of food quality data;

[0017] Based on the multi-level analysis index weights of the food quality data, a food cold chain logistics quality monitoring model based on the hierarchical analysis method is constructed;

[0018] Output steps include:

[0019] Based on the food cold chain logistics quality monitoring model based on the analytic hierarchy process, abnormal food quality is judged;

[0020] When a change in food quality is detected, an alarm signal is triggered and a notification message of the abnormal type is pushed.

[0021] Preferably, the obtaining step specifically includes:

[0022] Collecting geospatial data of food cold chain logistics through the geolocation sub-platform;

[0023] Collect video information of the food cold chain logistics process through the cold chain logistics video monitoring sub-platform;

[0024] The temperature and humidity data of the food cold chain logistics are collected through the temperature and humidity sensor sub-platform, wherein the temperature and humidity data include the ambient temperature, ambient humidity and food temperature of the food cold chain logistics;

[0025] Multispectral images of food cold chain logistics are collected through the multispectral image collection sub-platform.

[0026] Preferably, the step of constructing a food cold chain logistics quality prediction model specifically includes:

[0027] Establishing a time series prediction model based on support vector machine for food cold chain logistics, the time series prediction model includes a temperature time series support vector machine model, a humidity time series support vector machine model, a food temperature time series support vector machine model and a video time series support vector machine model;

[0028] The original quality data indicators are input into a corresponding time series prediction model for prediction to obtain the food quality data.

[0029] Preferably, the step of constructing a multi-level analysis index weight model for food quality data specifically includes:

[0030] Establishing a multi-level analysis indicator weight model for food quality data, wherein the multi-level analysis indicator weight model for food quality data includes a primary indicator score and a secondary indicator score;

[0031] The secondary indicators in the multi-level analysis indicator weight model of food quality data are assigned values ​​using the indicator 1-9 scaling method;

[0032] The relative weights of the primary indicators are determined using expert evaluation methods;

[0033] Calculate the relative weight of the secondary indicator A i and the relative weight of the first-level indicator B i , and obtain the matrices A and B of the secondary and primary indicators.

[0034] Preferably, the step of constructing a food cold chain logistics quality monitoring model based on the analytic hierarchy process specifically includes:

[0035] A food quality determination formula is established, and the food quality determination formula is:

[0036] f=∑(C i , D i,j ·X i ·c j ·Y j ·wj ·t j ·u j ·v j ),

[0037] Where, f represents the food quality grade; C i represents the food quality grade in the i-th time period of food transportation; D i,j represents the weight of the food quality grade in the i-th time period of food transportation; X i represents the type of food in the i-th time period of food transportation; c j Y represents the weight of the food quality level in the i-th time period of food transportation; j represents the type of food packaging material in the i-th time period of food transportation; w j represents the weight of the geographical environment in the i-th time period of food transportation on food quality; t j represents the weight of the time of food cold chain logistics on food quality in the i-th time period of food transportation; u j represents the weight of the video of food cold chain logistics in the i-th time period of food transportation on food quality; v j It represents the weight of the temperature and humidity data indicators of food cold chain logistics in the i-th time period of food transportation on food quality; i is a natural number.

[0038] Preferably, the step of judging abnormality of food quality specifically includes:

[0039] The collected food quality data, the multi-level analysis index weight model of food quality data and the food cold chain logistics quality monitoring model based on the hierarchical analysis method are used as input, and the weights c of the food quality grades are calculated. j As output;

[0040] According to the weight of food quality level c j The relationship with the food quality grade threshold to determine whether the food quality grade has changed;

[0041] When the test result c j <0, it is determined that the food quality grade has changed, where j = 1, 2, ..., m; the quality data of the current food, the output result of the food quality data determination formula and the corresponding food quality grade are used to generate a notification message and trigger an alarm signal.

[0042] As an advantage, the method further comprises the following steps:

[0043] When any of the video time series support vector machine model, the temperature time series support vector machine model, the food temperature time series support vector machine model, and the humidity time series support vector machine model is abnormal, the food cold chain logistics process is automatically suspended;

[0044] Conduct video evidence collection and store video data information;

[0045] Automatically record video information, temperature and humidity information, and geographic positioning information of the food cold chain logistics process.

[0046] As a preference, the method further comprises the following steps:

[0047] A food cold chain quality parameter influencing factor model is constructed, and the calculation formula of the food cold chain quality parameter influencing factor model is:

[0048] f ji =h i ·v i ·w i ,

[0049] Among them, f ji is the influencing factor of food quality parameters, h i is the cold chain transportation environment temperature of the i-th section in the food cold chain logistics process, v i is the logistics speed of the i-th section in the food cold chain logistics process, w i is the relative humidity of the i-th section in the food cold chain logistics process;

[0050] Establish the food freshness parameter calculation formula:

[0051] A ji =A0·(1-N ji / N i )·(M ji / M0),

[0052] Among them, A ji is the food freshness parameter, N ji is the cumulative consumption of the jth category in the i-th section of the food cold chain logistics process, M ji is the influencing factor of food quality parameters in the i-th section of the food cold chain logistics process, A0 is the initial value of the food freshness parameter, A i is the freshness parameter of the i-th section in the food cold chain logistics process, N i is the total quantity of the jth category in the ith section of the food cold chain logistics process, and M0 is the initial value of the influencing factor of food quality parameters.

[0053] As a preference, the method further comprises the following steps:

[0054] A food cold chain logistics quality monitoring model is constructed, and the calculation formula of the food cold chain logistics quality monitoring model is:

[0055] Q i =A i ·(1-B / f ji ),

[0056] Among them, Q i is the food cold chain logistics quality parameter of the i-th section in the food cold chain logistics process, A i is the freshness parameter of the i-th section in the food cold chain logistics process, B is the threshold of the food cold chain logistics quality monitoring model, and f ji are the factors affecting food quality parameters;

[0057] Based on the food cold chain logistics quality monitoring model, predict and analyze the food cold chain logistics quality parameters;

[0058] Data preprocessing of food cold chain logistics quality parameters, including outlier detection and data standardization of food cold chain logistics quality parameters;

[0059] Food cold chain logistics quality threshold prediction based on food cold chain logistics quality monitoring model, the cold chain food quality threshold prediction model is used to predict the food cold chain logistics quality monitoring threshold.

[0060] The food cold chain logistics quality monitoring and early warning system used to implement the method comprises:

[0061] Food quality information multi-sensor acquisition platform, used to obtain raw quality data indicators in food cold chain logistics;

[0062] Food cold chain logistics quality prediction module, used to build a food cold chain logistics quality prediction model and make predictions;

[0063] The food quality data multi-level analysis indicator weight calculation module is used to construct the food quality data multi-level analysis indicator weight model and calculate the weight;

[0064] Food cold chain logistics quality monitoring module, used to build a food cold chain logistics quality monitoring model based on hierarchical analysis method;

[0065] The abnormality judgment and early warning module is used to judge the abnormality of food quality and trigger an alarm signal and push a notification message when an abnormality is detected;

[0066] Video forensics module, used for video forensics and data storage when anomalies are detected;

[0067] Quality parameter influencing factor analysis module, used to build a food cold chain quality parameter influencing factor model and calculate food freshness parameters;

[0068] The quality monitoring and predictive analysis module is used to build a food cold chain logistics quality monitoring model and perform predictive analysis and data preprocessing.

[0069] The beneficial effects of the present invention are mainly reflected in the following aspects:

[0070] The method of the present invention not only overcomes the problems of incomplete data collection, low prediction accuracy, simplified evaluation methods, and lagging early warning mechanisms in the prior art, but also achieves synergistic efficiencies of multiple technical effects through the organic combination of various technical links. For example, multi-dimensional data collection provides rich training data for machine learning predictions, improving the accuracy of the prediction model; and accurate prediction results provide a reliable basis for quality assessment, thereby supporting more accurate early warnings.

[0071] In addition, the method of the present invention is highly adaptable and scalable. By adjusting the model parameters and threshold settings, the method can easily adapt to different types of food, different transportation environments, and different quality management requirements. This flexibility enables the method to play a role in various complex cold chain logistics scenarios, greatly improving its practical value.

[0072] In general, the method provided by the present invention is a comprehensive upgrade and innovation of the existing food cold chain logistics quality monitoring and early warning technology. It can not only significantly improve the food safety level, but also reduce operating costs and improve economic benefits by optimizing the cold chain logistics process. This is of great significance for promoting technological progress and sustainable development of the food cold chain logistics industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 The present invention is a flow chart of the method.

[0074] Figure 2 This is a flow chart of the food cold chain logistics quality prediction module of the present invention.

[0075] Figure 3 The present invention is a flowchart of a product quality data multi-level analysis index weight calculation module.

[0076] Figure 4 This is a flow chart of the food cold chain logistics quality monitoring module of the present invention.

[0077] Figure 5 It is a flow chart of the abnormality judgment and early warning module of the present invention.

[0078] Figure 6 It is a flow chart of the video evidence collection module of the present invention.

[0079] Figure 7 It is a flow chart of the quality parameter influencing factor analysis module of the present invention.

[0080] Figure 8 It is a flow chart of the quality monitoring and prediction analysis module of the present invention. DETAILED DESCRIPTION

[0081] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects thereof are described in detail below in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0082] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0083] Please refer to Figure 1-8 The present invention provides a method and system for monitoring and warning the quality of food cold chain logistics. The method acquires comprehensive quality data through a multi-sensor acquisition platform, uses a machine learning algorithm for predictive analysis, and establishes a quality monitoring model in combination with the analytic hierarchy process, thereby realizing intelligent monitoring and timely warning of the entire process of food cold chain logistics.

[0084] Specifically, the food cold chain logistics quality monitoring and early warning method of the present invention comprises the following steps:

[0085] First, the original quality data indicators in the food cold chain logistics are obtained through the food quality information multi-sensor acquisition platform. The multi-sensor acquisition platform integrates the cold chain logistics video monitoring sub-platform, geographic positioning sub-platform, temperature and humidity sensor sub-platform and multi-spectral image acquisition sub-platform, which can fully collect various key data in the food cold chain logistics process.

[0086] Among them, the cold chain logistics video monitoring sub-platform uses high-definition cameras to collect video images of the logistics process in real time. The geo-positioning sub-platform uses satellite positioning systems such as GPS / Beidou to obtain real-time location information of transport vehicles. The temperature and humidity sensing sub-platform collects environmental temperature and humidity and food temperature data through a distributed temperature and humidity sensor network. The multi-spectral image acquisition sub-platform uses a multi-spectral camera to obtain spectral information on the food surface for analyzing food freshness.

[0087] This multi-dimensional data collection method can fully reflect the various influencing factors in the food cold chain logistics process, providing a rich data basis for subsequent analysis. Compared with traditional single temperature monitoring, the multi-sensor fusion collection of the present invention greatly improves the comprehensiveness and accuracy of the data.

[0088] Next, based on the acquired original quality data indicators, this method constructs a food cold chain logistics quality prediction model. The prediction model uses the support vector machine (SVM) algorithm to establish a time series prediction model including temperature, humidity, food temperature and video data.

[0089] Specifically, the mathematical expression of the temperature time series support vector machine model is:

[0090]

[0091] Where f(x) is the predicted temperature value, x i is the historical temperature data, α i and is the Lagrange multiplier, K(x i ,x) is the kernel function, and b is the bias term. Similarly, corresponding SVM time series prediction models are also established for humidity, food temperature, and video data. Compared with traditional statistical prediction methods, this prediction method based on machine learning has stronger nonlinear fitting and generalization capabilities, and can more accurately predict the trend of food quality changes.

[0092] After obtaining the food quality data, this method further constructs a multi-level analysis index weight model for food quality data. The model uses the analytic hierarchy process (AHP) to establish a hierarchical structure including primary and secondary indicators.

[0093] Preferably, the first-level indicators include food safety, freshness, nutritional value, etc., and the second-level indicators are further refined into specific indicators such as temperature fluctuation, humidity change, microbial content, color change, etc. For the second-level indicators, a 1-9 scale method is used for pairwise comparison and assignment. For example, the impact of temperature fluctuation on food safety may be assigned 9 points, while the impact of humidity change may be assigned 5 points.

[0094] After the relative weights of the first-level indicators are determined through expert evaluation, the complete weight system can be calculated. Assuming there are m first-level indicators and n second-level indicators, the final weight matrix can be expressed as:

[0095]

[0096] where w ij It represents the comprehensive weight of the jth secondary indicator under the i-th primary indicator.

[0097] This multi-level weight analysis method can comprehensively consider the impact of various factors on food quality and provide a scientific basis for subsequent quality assessment.

[0098] Based on the above weight model, the present invention further constructs a food cold chain logistics quality monitoring model based on hierarchical analysis method. The core of the model is to establish a comprehensive food quality judgment formula:

[0099]

[0100] Where, f represents the final food quality grade, C irepresents the food quality level in the i-th time period, D i , j is the corresponding weight, X i Indicates the type of food, c j is the quality level weight, Y j Indicates the type of packaging material, w j ,t j 、u j and v j They respectively represent the impact weights of geographical environment, time, video, and temperature and humidity data on food quality.

[0101] This comprehensive judgment formula takes into account multiple key factors in the food cold chain logistics process and can more comprehensively and accurately evaluate the food quality status. For example, for perishable fruits, the temperature factor may be given a higher weight, while for vegetables that need to retain moisture, the humidity factor may receive more attention.

[0102] In practical applications, this method also sets up a corresponding abnormal judgment mechanism. When a change in food quality is detected, the system will trigger an alarm signal and push a notification message of the abnormal type. Specifically, a quality change threshold can be set, for example, when the weight of the quality level c j When it is less than 0.8 (assuming the weight of the normal quality level is 1), it is judged as abnormal quality. This timely abnormality detection and early warning mechanism can effectively prevent the expansion of food quality problems.

[0103] The method of the present invention is not limited to quality monitoring, but also includes comprehensive data collection and preprocessing processes. For example, when acquiring raw data, the geolocation sub-platform can collect GPS coordinate data every 5 minutes to accurately record the driving trajectory of the transport vehicle. The temperature and humidity sensing sub-platform can collect temperature and humidity data every 1 minute to ensure real-time monitoring of temperature and humidity changes.

[0104] In the data preprocessing stage, this method uses a series of technical means to improve data quality. For example, for temperature data, a moving average filtering algorithm can be used to remove abnormal fluctuations:

[0105]

[0106] Where T′ i is the smoothed temperature value, T i is the original temperature data, and n is the sliding window size (can be 2-5).

[0107] In addition, this method also takes into account the special needs of different types of food. For example, for frozen meat, a strict temperature control range (-18°C to -15°C) can be set; while for fresh fruit, the combined effect of temperature and humidity may need to be considered, such as setting a suitable temperature and humidity range (temperature 0-4°C, relative humidity 85% to 95%).

[0108] In general, the food cold chain logistics quality monitoring and early warning method provided by the present invention realizes intelligent monitoring and accurate early warning of the entire process of food cold chain logistics through innovative technologies such as multi-dimensional data collection, machine learning prediction, multi-level analysis and comprehensive quality assessment. This method can not only detect potential quality problems in a timely manner, but also provide data support for the optimized management of food cold chain logistics, effectively improving food safety levels and economic benefits. The food cold chain logistics quality monitoring and early warning method of the present invention also includes a series of sophisticated abnormal judgment and processing mechanisms in practical applications. Specifically, the steps of this method for abnormal judgment of food quality are as follows:

[0109] First, the collected food quality data, the aforementioned food quality data multi-level analysis index weight model and the food cold chain logistics quality monitoring model based on the analytic hierarchy process are used as input to output the weight cj of the food quality grade. This step actually converts the complex multi-dimensional data into a quantifiable quality indicator through the model.

[0110] Next, this method uses the weight c of the food quality grade j The food quality level is compared with the preset food quality level threshold to determine whether the food quality level has changed. Preferably, multiple thresholds can be set to correspond to different degrees of quality changes. For example, 0.9 can be set as the threshold for slight quality change, 0.8 as the threshold for moderate quality change, and 0.7 as the threshold for severe quality change. j When the value is less than these thresholds, the system will determine that the food quality level has changed to a corresponding extent.

[0111] It is worth noting that the value range of j in the present invention is 1 to m, where m represents the total number of food quality grades. For example, food quality can be divided into five grades: high quality, good, general, poor, and unqualified, and the value of m is 5. This detailed classification can more accurately reflect the changing trend of food quality.

[0112] When the system detects a change in food quality level, a series of response mechanisms will be triggered. First, the system will generate a notification message with the current food quality data, the output of the food quality data determination formula, and the corresponding food quality level. This information will be integrated into a detailed quality report, including specific quality indicators, change trends, and possible cause analysis.

[0113] At the same time, the system will also trigger an alarm signal. Preferably, different levels of alarms can be set according to the degree of quality change. For example, for a slight quality change, the system may only send a reminder message; while for a serious quality problem, the system may send an audible and visual alarm and immediately notify the relevant person in charge.

[0114] The method of the present invention also includes a complete abnormal situation processing mechanism. When any of the video time series support vector machine model, the temperature time series support vector machine model, the food temperature time series support vector machine model, and the humidity time series support vector machine model is abnormal, the system will automatically suspend the food cold chain logistics process. This timely response can effectively prevent the quality problem from further deteriorating.

[0115] While suspending the logistics process, this method will also start the video evidence collection program to comprehensively record the abnormal situation. The system will automatically store video data information to provide important evidence for subsequent cause analysis and accountability. In addition to video information, the system will also automatically record the temperature and humidity information and geographic location information of the food cold chain logistics process. These multi-dimensional data can help managers fully understand the environmental conditions and specific locations where the abnormalities occurred.

[0116] In order to more accurately evaluate food quality, the present invention also constructs a food cold chain quality parameter influencing factor model. The calculation formula of the model is:

[0117] f ji =h i ·v i ·w i ,

[0118] Among them, f ji represents the influencing factor of food quality parameters, h i represents the cold chain transportation environment temperature of the i-th section in the food cold chain logistics process, v i Indicates the logistics speed of this section, w i Represents relative humidity. This model considers the combined impact of three key factors, temperature, speed and humidity, on food quality.

[0119] For example, for perishable fruits, a higher weight may be given to the temperature factor, while for vegetables that need to stay fresh, more attention may be paid to the speed factor. By adjusting the weights of these factors, more accurate quality assessment standards can be developed for different types of food.

[0120] In addition, the present invention also establishes a food freshness parameter calculation formula:

[0121]

[0122] Among them, A ji Represents food freshness parameter, N ji M represents the cumulative consumption of the jth category in the i-th section of the food cold chain logistics process, ji It represents the influencing factor of the food quality parameter in this section, A0 represents the initial value of the food freshness parameter, A i represents the freshness parameter of the i-th segment, N i represents the total number of the jth category in the ith segment, and M0 represents the initial value of the influencing factor of food quality parameters.

[0123] This formula takes into account the consumption of food and factors affecting its quality, and can more accurately reflect the changes in the freshness of food during the cold chain logistics process. For example, for fresh fruit, appropriate initial freshness parameters and influencing factors can be set according to its respiration intensity, so as to track its freshness changes more accurately.

[0124] In order to comprehensively monitor the quality of food cold chain logistics, the present invention also constructs a food cold chain logistics quality monitoring model. The calculation formula of the model is:

[0125]

[0126] Among them, Q i A represents the quality parameter of the food cold chain logistics in the i-th section of the food cold chain logistics process. i represents the freshness parameter of this segment, B represents the threshold of the food cold chain logistics quality monitoring model, and f ji Represents the influencing factors of food quality parameters.

[0127] This model combines freshness parameters and quality influencing factors to comprehensively evaluate the quality status of food during the cold chain logistics process. By setting an appropriate threshold B, potential quality problems can be discovered in a timely manner. For example, for high-value seafood products, a lower threshold B can be set to ensure stricter quality control.

[0128] Based on this quality monitoring model, this method also conducts a series of predictive analysis and data preprocessing. First, predictive analysis of food cold chain logistics quality parameters can predict possible quality problems in advance. Second, through data preprocessing, including the detection of outliers in food cold chain logistics quality parameters and data standardization, the reliability and comparability of the data can be improved.

[0129] Finally, this method also uses the cold chain food quality threshold prediction model to predict the food cold chain logistics quality monitoring threshold. This dynamic threshold adjustment mechanism can adaptively adjust the quality monitoring standards according to different food types, seasonal changes and other factors, thereby achieving more intelligent and accurate quality control.

[0130] In general, the food cold chain logistics quality monitoring and early warning method provided by the present invention realizes intelligent monitoring, accurate evaluation and timely early warning of the whole process of food cold chain logistics by constructing a series of sophisticated models and algorithms. This method can not only effectively ensure food safety, but also provide important data support for the optimization management of cold chain logistics, and has important practical application value.

[0131] The present invention also provides a system for implementing the above-mentioned food cold chain logistics quality monitoring and early warning method. The system includes multiple functional modules, each of which implements specific steps in the method in a targeted manner, thereby forming a complete food cold chain logistics quality monitoring and early warning solution.

[0132] First, the system includes a food quality information multi-sensor acquisition platform 1. This platform is the data source of the entire system and is responsible for obtaining the original quality data indicators in the food cold chain logistics. The food quality information multi-sensor acquisition platform 1 integrates a variety of sensor devices, including but not limited to high-definition cameras, GPS positioning modules, temperature and humidity sensors, and multi-spectral cameras. These devices are distributed at various key nodes of cold chain logistics and collect multi-dimensional quality-related data in real time.

[0133] For example, on a typical cold chain transport vehicle, temperature and humidity sensors can be installed to monitor the environmental parameters in the vehicle compartment, while a GPS module can be used to track the vehicle location, and the camera in the vehicle compartment can be used to monitor the food status. For some special foods, such as fresh fruits, multispectral cameras can also be used to evaluate their surface features to more accurately determine the freshness.

[0134] Next, the system includes a food cold chain logistics quality prediction module 2. This module receives data from the multi-sensor acquisition platform and uses advanced machine learning algorithms to build a food cold chain logistics quality prediction model. The food cold chain logistics quality prediction module 2 is mainly based on the support vector machine (SVM) algorithm and establishes multiple time series prediction models including temperature, humidity, food temperature and video data.

[0135] In practical applications, the food cold chain logistics quality prediction module 2 may adopt different prediction strategies for different types of food. For example, for perishable aquatic products, more attention may be paid to the prediction of temperature changes; while for vegetables that need to retain moisture, more attention may be paid to the prediction of humidity changes.

[0136] The system also includes a food quality data multi-level analysis indicator weight calculation module 3. This module is responsible for constructing a food quality data multi-level analysis indicator weight model and calculating the weight of each indicator. The food quality data multi-level analysis indicator weight calculation module 3 adopts the hierarchical analysis method (AHP) to determine the relative importance of each quality indicator by combining expert evaluation and mathematical calculation.

[0137] In actual operation, the weight calculation module 3 for multi-level analysis indicators of food quality data may regularly update the weight model to adapt to the special needs of different seasons, different regions or different types of food. For example, in summer, temperature control may be given a higher weight; and in the case of a long transportation distance, the weight of the time factor may be increased accordingly.

[0138] Next is the food cold chain logistics quality monitoring module 4. This module is responsible for building a food cold chain logistics quality monitoring model based on the hierarchical analysis method. The food cold chain logistics quality monitoring module 4 comprehensively considers multiple factors, including food type, packaging materials, environmental conditions, etc., and uses a complex mathematical model to evaluate the real-time quality status of food.

[0139] In specific implementation, the food cold chain logistics quality monitoring module 4 may set different quality assessment standards for different food categories. For example, for fresh fruits, more attention may be paid to sensory indicators such as color and hardness; while for frozen meat, more attention may be paid to the stability of the core temperature.

[0140] The system also includes an abnormality judgment and early warning module 5. This module is responsible for judging the abnormality of food quality and triggering an alarm signal and pushing a notification message when an abnormality is detected. The abnormality judgment and early warning module 5 can respond to different degrees of quality abnormalities by setting multiple thresholds.

[0141] For example, when a slight quality change is detected, the abnormality judgment and early warning module 5 may only send a reminder message to the relevant personnel; when a serious quality problem is found, it may trigger an emergency alarm and immediately notify the management personnel to take emergency measures.

[0142] In order to preserve evidence of abnormal situations, the system is also equipped with a video evidence module 6. When the system detects an abnormality, the video evidence module 6 will automatically start to record the on-site situation at that time. This includes not only video images, but also multi-dimensional data such as temperature and humidity data at the corresponding time point, location information, etc. These data can provide important basis for subsequent cause analysis and accountability.

[0143] The system also includes a quality parameter influencing factor analysis module 7. This module is responsible for building a food cold chain quality parameter influencing factor model and calculating food freshness parameters. The quality parameter influencing factor analysis module 7 takes into account the comprehensive impact of multiple factors such as temperature, speed, humidity, etc. on food quality, and can more accurately evaluate the quality changes of food during transportation.

[0144] Finally, the system includes a quality monitoring and prediction analysis module 8. This module is responsible for building a food cold chain logistics quality monitoring model, performing prediction analysis and data preprocessing. The quality monitoring and prediction analysis module 8 can not only monitor the food quality status in real time, but also predict the quality change trend in the future, providing an important reference for management decisions.

[0145] Through the organic combination of these functional modules, the system of the present invention realizes intelligent monitoring and accurate early warning of the entire process of food cold chain logistics. The system can not only detect and handle quality anomalies in a timely manner, but also provide data support for the optimization management of cold chain logistics, effectively improving food safety levels and economic benefits.

[0146] It is worth noting that the above modules do not operate in isolation, but closely collaborate and communicate data. For example, the data collected by the food quality information multi-sensor acquisition platform 1 will be fed back to multiple subsequent modules at the same time; and the output of the quality monitoring and prediction analysis module 8 may in turn affect the threshold setting of the abnormal judgment and early warning module 5. This synergy between modules enables the entire system to operate more intelligently and efficiently.

[0147] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A food cold chain logistics quality monitoring and early warning method, characterized in that: include: The acquisition steps include: Acquire original quality data indicators in food cold chain logistics through a food quality information multi-sensor acquisition platform, wherein the food quality information multi-sensor acquisition platform includes a cold chain logistics video monitoring sub-platform, a geographic positioning sub-platform, a temperature and humidity sensor sub-platform, and a multi-spectral image acquisition sub-platform; Processing steps include: Based on the original quality data indicators, a food cold chain logistics quality prediction model is constructed; According to the food cold chain logistics quality prediction model, the original quality data indicators are predicted to obtain food quality data; Based on the food quality data, a multi-level analysis index weight model for food quality data is constructed; Calculating the weights of multi-level analysis indicators of food quality data according to the multi-level analysis indicator weight model of food quality data; Based on the multi-level analysis index weights of the food quality data, a food cold chain logistics quality monitoring model based on the hierarchical analysis method is constructed; Output steps include: Based on the food cold chain logistics quality monitoring model based on the analytic hierarchy process, abnormal food quality is judged; When a change in food quality is detected, an alarm signal is triggered and a notification message of the abnormal type is pushed.

2. The food cold chain logistics quality monitoring and early warning method according to claim 1 is characterized in that: The acquisition step specifically includes: Collecting geospatial data of food cold chain logistics through the geolocation sub-platform; Collect video information of the food cold chain logistics process through the cold chain logistics video monitoring sub-platform; The temperature and humidity data of the food cold chain logistics are collected through the temperature and humidity sensor sub-platform, wherein the temperature and humidity data include the ambient temperature, ambient humidity and food temperature of the food cold chain logistics; Multispectral images of food cold chain logistics are collected through the multispectral image collection sub-platform.

3. The food cold chain logistics quality monitoring and early warning method according to claim 1 is characterized in that: The steps of constructing a food cold chain logistics quality prediction model specifically include: Establishing a time series prediction model based on support vector machine for food cold chain logistics, the time series prediction model includes a temperature time series support vector machine model, a humidity time series support vector machine model, a food temperature time series support vector machine model and a video time series support vector machine model; The original quality data indicators are input into a corresponding time series prediction model for prediction to obtain the food quality data.

4. The food cold chain logistics quality monitoring and early warning method according to claim 1 is characterized in that: The steps of constructing a multi-level analysis index weight model for food quality data specifically include: Establishing a multi-level analysis indicator weight model for food quality data, wherein the multi-level analysis indicator weight model for food quality data includes a primary indicator score and a secondary indicator score; The secondary indicators in the multi-level analysis indicator weight model of food quality data are assigned values ​​using the indicator 1-9 scaling method; The relative weights of the primary indicators are determined using expert evaluation methods; Calculate the relative weight of the secondary indicator A i and the relative weight of the first-level indicator B i , and obtain the matrices A and B of the secondary and primary indicators.

5. The food cold chain logistics quality monitoring and early warning method according to claim 1 is characterized in that: The steps of constructing a food cold chain logistics quality monitoring model based on the analytic hierarchy process specifically include: A food quality determination formula is established, and the food quality determination formula is: f=∑(C i ·D i,j ·X i ·c j ·Y j ·w j ·t j ·u j ·v j ), Where, f represents the food quality grade; C i represents the food quality grade in the i-th time period of food transportation; D i,j represents the weight of the food quality grade in the i-th time period of food transportation; X i represents the type of food in the i-th time period of food transportation; c j Y represents the weight of the food quality level in the i-th time period of food transportation; i represents the type of food packaging material in the i-th time period of food transportation; w j represents the weight of the geographical environment in the i-th time period of food transportation on food quality; t j represents the weight of the time of food cold chain logistics on food quality in the i-th time period of food transportation; u j represents the weight of the video of food cold chain logistics in the i-th time period of food transportation on food quality; v j It represents the weight of the temperature and humidity data indicators of food cold chain logistics in the i-th time period of food transportation on food quality; i is a natural number.

6. The food cold chain logistics quality monitoring and early warning method according to claim 1 is characterized in that: The step of judging abnormality of food quality specifically includes: The collected food quality data, the multi-level analysis index weight model of food quality data and the food cold chain logistics quality monitoring model based on the hierarchical analysis method are used as input, and the weights c of the food quality grades are calculated. j As output; According to the weight of food quality level c j The relationship with the food quality grade threshold to determine whether the food quality grade has changed; When the test result c j <0, it is determined that the food quality grade has changed, where j = 1, 2, ..., m; the quality data of the current food, the output result of the food quality data determination formula and the corresponding food quality grade are used to generate a notification message and trigger an alarm signal.

7. The food cold chain logistics quality monitoring and early warning method according to claim 1 is characterized in that: The following steps are also included: When any of the video time series support vector machine model, the temperature time series support vector machine model, the food temperature time series support vector machine model, and the humidity time series support vector machine model is abnormal, the food cold chain logistics process is automatically suspended; Conduct video evidence collection and store video data information; Automatically record video information, temperature and humidity information, and geographic positioning information of the food cold chain logistics process.

8. The food cold chain logistics quality monitoring and early warning method according to claim 1 is characterized in that: The following steps are also included: A food cold chain quality parameter influencing factor model is constructed, and the calculation formula of the food cold chain quality parameter influencing factor model is: f ji =h i ·v i ·w i , Among them, f ji is the influencing factor of food quality parameters, h i is the cold chain transportation environment temperature of the i-th section in the food cold chain logistics process, v i is the logistics speed of the i-th section in the food cold chain logistics process, w i is the relative humidity of the i-th section in the food cold chain logistics process; Establish the food freshness parameter calculation formula: A ji =A0·(1-N ji / N i )·(M ji / M0), Among them, A ji is the food freshness parameter, N ji is the cumulative consumption of the jth category in the i-th section of the food cold chain logistics process, M ji is the influencing factor of food quality parameters in the i-th section of the food cold chain logistics process, A0 is the initial value of the food freshness parameter, A i is the freshness parameter of the i-th section in the food cold chain logistics process, N i is the total quantity of the jth category in the ith section of the food cold chain logistics process, and M0 is the initial value of the influencing factor of food quality parameters.

9. The food cold chain logistics quality monitoring and early warning method according to claim 1 is characterized in that: The following steps are also included: A food cold chain logistics quality monitoring model is constructed, and the calculation formula of the food cold chain logistics quality monitoring model is: Q i =A i ·(1-B / f ji ), Among them, Q i is the food cold chain logistics quality parameter of the i-th section in the food cold chain logistics process, A i is the freshness parameter of the i-th section in the food cold chain logistics process, B is the threshold of the food cold chain logistics quality monitoring model, and f ji are the factors affecting food quality parameters; Based on the food cold chain logistics quality monitoring model, predict and analyze the food cold chain logistics quality parameters; Data preprocessing of food cold chain logistics quality parameters, including outlier detection and data standardization of food cold chain logistics quality parameters; Food cold chain logistics quality threshold prediction based on food cold chain logistics quality monitoring model, the cold chain food quality threshold prediction model is used to predict the food cold chain logistics quality monitoring threshold.

10. A food cold chain logistics quality monitoring and early warning system for implementing the method described in any one of claims 1 to 9, characterized in that: include: Food quality information multi-sensor acquisition platform, used to obtain raw quality data indicators in food cold chain logistics; Food cold chain logistics quality prediction module, used to build a food cold chain logistics quality prediction model and make predictions; The food quality data multi-level analysis indicator weight calculation module is used to construct the food quality data multi-level analysis indicator weight model and calculate the weight; Food cold chain logistics quality monitoring module, used to build a food cold chain logistics quality monitoring model based on hierarchical analysis method; The abnormality judgment and early warning module is used to judge the abnormality of food quality and trigger an alarm signal and push a notification message when an abnormality is detected; Video forensics module, used for video forensics and data storage when anomalies are detected; Quality parameter influencing factor analysis module, used to build a food cold chain quality parameter influencing factor model and calculate food freshness parameters; The quality monitoring and predictive analysis module is used to build a food cold chain logistics quality monitoring model and perform predictive analysis and data preprocessing.

Images