Food data processing method and device based on big data and storage medium

Abstract

The invention relates to the technical field of food data processing, in particular to a food data processing method and device based on big data and a storage medium, and the method comprises the steps: collecting environment data, food data and spatio-temporal data in a food cold-chain transportation process; carrying out heterogeneous fusion on the collected data to construct a space-time matrix and a food incidence matrix; performing dynamic feature extraction and environment coupling weight distribution on the space-time matrix to obtain a temperature feature, a microorganism feature and an environment coupling feature; splicing the temperature features, the microorganism features and the environment coupling weight into a feature vector, and judging a data state and performing early warning according to the feature vector; the food state after food cold chain transportation is collected; and updating the analysis process of the data state according to the food state, the environmental data and the spatio-temporal data. According to the invention, accurate monitoring processing of food data in food cold-chain transportation is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of food data processing, and in particular to a food data processing method, device and storage medium based on big data. Background Art

[0002] Food data monitoring and processing uses big data technology to clean, integrate and analyze data in all aspects of food production, circulation and consumption, identify safety risks through data mining, machine learning and other algorithms, improve supervision efficiency, ensure food safety, optimize production processes, provide a scientific basis for policy making, and protect public health.

[0003] Food data processing in existing technologies mostly involves safety testing and abnormal analysis, lacks in-depth analysis of specific environments or special conditions, lacks real-time monitoring and prediction of the impact of food temperature fluctuations on quality in cold chain logistics, and lacks real-time monitoring and risk warning of microbial growth dynamics during food processing. Summary of the invention

[0004] The object of the present invention is to provide a food data processing method, device and storage medium based on big data to solve at least one of the problems existing in the prior art.

[0005] To achieve the above object, the present invention adopts the following technical solutions: A food data processing method based on big data includes: Collect environmental data, food data, and spatiotemporal data during food cold chain transportation; Heterogeneous fusion of the collected data is performed to construct a spatiotemporal matrix and a food association matrix; Dynamic feature extraction and environmental coupling weight allocation are performed on the spatiotemporal matrix to obtain temperature features, microbial features and environmental coupling features; The temperature characteristics, microbial characteristics and environmental coupling weights are spliced ​​into feature vectors, and the data status is judged and an early warning is issued based on the feature vectors; Collect the status of food after cold chain transportation; The analytical process of updating data status based on food status, environmental data and spatiotemporal data.

[0006] Furthermore, the collection time of environmental data and spatiotemporal data is aligned, and the collection time of environmental data and spatiotemporal data is numbered in chronological order to obtain the collection number t. The environmental data and spatiotemporal data of the same collection time are heterogeneously fused to construct the spatiotemporal matrix E(t), and E(t)=[T(t),H(t),C(t),L(t)] is set, where T(t) represents temperature, H(t) represents humidity, C(t) represents carbon dioxide concentration, and L(t) represents altitude.

[0007] Furthermore, based on the duration of the preset analysis cycle, the temperature in the spatiotemporal matrix within 1 hour is dynamically extracted to obtain the temperature feature, and the expression of the temperature feature is: , where F(t) represents the temperature feature, v represents the first extraction parameter, v∈N and v≤V, V represents the second extraction parameter, V=3600 / t1, t1 represents the duration of the preset analysis cycle, and λ represents the temperature decay factor.

[0008] Furthermore, based on the initial microbial content, the temperature in the spatiotemporal matrix and the food association matrix are dynamically extracted to obtain microbial features. The expression of the microbial features is: G(t)=G0×e G1(i)×[T(t)-G2(i)] / G2(i) , where G(t) represents the microbial characteristics and G0 represents the initial microbial content.

[0009] Furthermore, environmental coupling weights are assigned to the carbon dioxide concentration and altitude in the space-time matrix to obtain environmental coupling characteristics. The expression of the environmental coupling characteristics is: W(t)=β1×(1-L(t) / Lmax)+β2×C(t) / Cmax, where β1 represents the altitude weight, β2 represents the carbon dioxide weight, β1+β2=1, Lmax represents the preset maximum tolerable altitude, and Cmax represents the preset critical carbon dioxide concentration.

[0010] Furthermore, the temperature characteristics, microbial characteristics and environmental coupling weights are concatenated to obtain a feature vector X(t), set X(t)=[F(t), G(t), W(t)], and the dot product of feature vectors with adjacent collection numbers is used as the risk feature value, and the expression of the risk feature value is R(t)=X(t)·X(t-1); The risk characteristic values ​​in the food cold chain transportation process are extracted, the mean and standard deviation of the extracted risk characteristic values ​​are calculated, and the characteristic value tolerance interval is formulated based on the mean and standard deviation of the extracted risk characteristic values. The characteristic value tolerance interval is set to [μ(R(t))-3×σ(R(t)),μ(R(t))+3×σ(R(t))]. If the risk characteristic value of the current analysis period does not fall within the characteristic value tolerance interval, the data status is judged to be abnormal; otherwise, the data status is judged to be normal. An early warning is issued based on the data status and risk characteristic value. When the data status is abnormal, if the risk characteristic value is less than the first characteristic threshold, a severe warning is issued; if the risk characteristic value is greater than or equal to the first characteristic threshold and less than the second characteristic threshold, a moderate warning is issued; if the risk characteristic value is greater than or equal to the second characteristic threshold, a mild warning is issued; when the data status is normal, if the risk characteristic value is less than the first characteristic threshold, a moderate warning is issued; if the risk characteristic value is greater than or equal to the first characteristic threshold, no warning is issued.

[0011] Furthermore, the risk characteristic values ​​are divided into two categories according to the food status: normal characteristic values ​​and spoiled characteristic values. The number of normal characteristic values ​​that do not belong to the characteristic value tolerance interval is counted as the first type of abnormal number, and the number of spoiled characteristic values ​​that do not belong to the characteristic value tolerance interval is counted as the second type of abnormal number. The state correlation coefficient is analyzed. The expression of the state correlation coefficient is S=σ(R1(t)) Nr1 / NR1 / σ(R2(t)) Nr2 / NR2 , where S represents the state correlation coefficient, R1(t) represents the normal characteristic value, Nr1 represents the number of type I anomalies, NR1 represents the number of normal characteristic values, R2(t) represents the deteriorated characteristic value, Nr2 represents the number of type II anomalies, and NR2 represents the number of deteriorated characteristic values; The transport time is taken as the independent variable and the product of humidity and carbon dioxide concentration is taken as the dependent variable for regression analysis to analyze the regression equation of the product of humidity and carbon dioxide concentration on the transport time and calculate the determination coefficient of the regression equation. If the determination coefficient is greater than the determination threshold, the analysis process of the state correlation coefficient is not updated. Otherwise, the expression of the state correlation coefficient is updated to S=r×σ(R1(t)) Nr1 / NR1 / σ(R2(t)) Nr2 / NR2 , where r represents the coefficient of determination.

[0012] Furthermore, the state correlation coefficient is compared with the correlation threshold. When the state correlation coefficient is greater than the correlation threshold, the feature tolerance interval is updated to [μ(R(t))-(3-S)×σ(R(t)),μ(R(t))+(3+S)×σ(R(t))] to update the analysis process of the data state. Otherwise, the analysis process of the data state is not updated.

[0013] On the other hand, the present invention also provides a food data processing device based on big data, comprising: The collection module is used to collect environmental data, food data and spatiotemporal data during the food cold chain transportation process and the food status after the food cold chain transportation; The fusion module is used to perform heterogeneous fusion of the collected data to construct a spatiotemporal matrix and a food association matrix; The analysis module is used to extract dynamic features and assign environmental coupling weights to the spatiotemporal matrix to obtain temperature features, microbial features, and environmental coupling features; The early warning module is used to combine temperature characteristics, microbial characteristics and environmental coupling weights into a feature vector, and is also used to judge the data status and issue an early warning based on the feature vector; The update module is used to update the analysis process of data status based on food status, environmental data and spatiotemporal data.

[0014] On the other hand, the present invention also provides a storage medium storing instructions, which, when executed on a computer, enables the computer to execute any of the food data processing methods based on big data as described above.

[0015] The beneficial effects of the present invention are as follows: through multi-source data fusion, dynamic feature extraction, environmental coupling weight allocation and feedback-driven comprehensive analysis, accurate prediction and graded control of food spoilage risks can be achieved in complex cold chain scenarios, solving the limitations of static threshold criteria in traditional methods, and realizing active risk prediction and adaptive regulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 This is a flow chart of the food data processing method based on big data in this embodiment.

[0018] Figure 2 Flow chart of the feature analysis method of the space-time matrix in this embodiment.

[0019] Figure 3 Flow chart of the data status updating method of this embodiment.

[0020] Figure 4 It is a structural schematic diagram of the food data processing device based on big data in this embodiment. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0022] It should be understood that in the description of the embodiments of the present invention, the meaning of multiple (or multiple) is more than two, greater than, less than, and exceeding are understood to exclude the number itself, and above, below, and within are understood to include the number itself. If there is a description of "first", "second", etc., it is only used for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0023] See also Figure 1 As shown, it is a food data processing method based on big data in this embodiment, including: Step S1, collecting environmental data, food data and spatiotemporal data during the cold chain transportation of food, the environmental data including temperature, humidity and carbon dioxide concentration, the temperature is in degrees Celsius, the humidity is relative humidity, which is a percentage value, the carbon dioxide concentration is in ppm, the food data includes food type, initial microbial content, microbial growth rate and critical temperature for deterioration, the critical temperature for deterioration is in degrees Celsius, the microbial growth rate is a specific growth rate, the spatiotemporal data includes transportation time and altitude, the transportation time is in hours, the altitude is in meters, the collection frequency of the environmental data and spatiotemporal data is once per preset analysis period, in this embodiment, the duration of the preset analysis period is set to 3 seconds, in this embodiment, the setting of the preset analysis period is not specifically limited, and can be freely set by technicians in this field, such as 1 second, 5 seconds, etc., the environmental data and spatiotemporal data are collected by sensors, and the food data are collected by food microbiological testing methods and uploaded.

[0024] Specifically, in step S1 described in this embodiment, by collecting temperature, humidity, carbon dioxide concentration, altitude, transportation time and food microbiological data to cover all dimensions of environment, time, space and biological properties, multi-dimensional data fusion is achieved to avoid misjudgment caused by analyzing a single parameter.

[0025] Please continue reading Figure 1 As shown, the food data processing method based on big data also includes: Step S2, heterogeneously fuse the collected data to construct a spatiotemporal matrix and a food association matrix.

[0026] Specifically, in step S2 described in this embodiment, the collection time of the environmental data and the spatiotemporal data is aligned, and the collection time of the environmental data and the spatiotemporal data is numbered in chronological order to obtain the collection number t, and the environmental data and the spatiotemporal data of the same collection time are heterogeneously fused to construct a spatiotemporal matrix E(t), and E(t)=[T(t), H(t), C(t), L(t)] is set, where T(t) represents temperature, H(t) represents humidity, C(t) represents carbon dioxide concentration, and L(t) represents altitude.

[0027] Specifically, in step S2 described in this embodiment, the food types are associated with their corresponding microbial growth rates and the critical temperature of deterioration of the food association matrix to construct a food association matrix, E(i), and E(i)=[G1(i),G2(i)] is set, where i represents the food type, G1(i) represents the microbial growth rate, and G2(i) represents the critical temperature of deterioration.

[0028] Specifically, in step S2 described in this embodiment, the environmental data and the spatiotemporal data are aligned by timestamps to construct a spatiotemporal matrix and a food association matrix, thereby eliminating feature deviations caused by data time misalignment, supporting efficient feature extraction and weight allocation, and improving data processing efficiency.

[0029] Please continue reading Figure 1 As shown, the food data processing method based on big data also includes: Step S3, extracting dynamic features and assigning environmental coupling weights to the spatiotemporal matrix and to obtain temperature features, microbial features and environmental coupling features.

[0030] See also Figure 2 As shown, it is a characteristic analysis method of the space-time matrix, including: Step S31, extracting dynamic features of the temperature in the spatiotemporal matrix to obtain temperature features.

[0031] Specifically, in step S31 of this embodiment, dynamic feature extraction is performed on the temperature in the spatiotemporal matrix within 1 hour based on the duration of the preset analysis cycle to obtain a temperature feature, and the expression of the temperature feature is: , where F(t) represents the temperature feature, v represents the first extraction parameter, v∈N and v≤V, V represents the second extraction parameter, V=3600 / t1, t1 represents the duration of the preset analysis cycle, λ represents the temperature decay factor, 0<λ≤0.5. It can be understood that the value of the temperature decay factor is not specifically limited in this embodiment, and those skilled in the art can freely set it as long as it satisfies the analysis of the temperature feature. The optimal value of the temperature decay factor is: λ=0.2.

[0032] Specifically, in step S31 of this embodiment, by analyzing the characteristics of the temperature data in the spatiotemporal matrix, a temperature decay factor is used to give a higher weight to recent temperature fluctuations, so as to more sensitively capture sudden temperature anomalies.

[0033] Please continue reading Figure 2 As shown, the feature analysis method of the space-time matrix also includes: Step S32, performing dynamic feature extraction on the temperature in the spatiotemporal matrix and the food association matrix to obtain microbial features.

[0034] Specifically, in step S32 of this embodiment, the temperature in the spatiotemporal matrix and the food association matrix are dynamically extracted based on the initial microbial content to obtain microbial features. The expression of the microbial features is: G(t)=G0×e G1(i)×[T(t)-G2(i)] / G2(i) , where G(t) represents the microbial characteristics and G0 represents the initial microbial content.

[0035] Please continue reading Figure 2 As shown, the feature analysis method of the space-time matrix also includes: Step S33, assigning environmental coupling weights to the carbon dioxide concentration and altitude in the spatiotemporal matrix to obtain environmental coupling features.

[0036] Specifically, in step S33 described in this embodiment, environmental coupling weights are assigned to the carbon dioxide concentration and altitude in the space-time matrix to obtain environmental coupling characteristics, and the expression of the environmental coupling characteristics is: W(t)=β1×(1-L(t) / Lmax)+β2×C(t) / Cmax, where β1 represents the altitude weight, β2 represents the carbon dioxide weight, β1+β2=1, Lmax represents the preset maximum tolerable altitude, and Cmax represents the preset carbon dioxide critical concentration.

[0037] Specifically, in this embodiment, the altitude weight is set to 0.6, the carbon dioxide weight is set to 0.4, the preset maximum tolerable altitude is set to 3000, and the preset carbon dioxide critical concentration is set to 1000. It can be understood that in this embodiment, the altitude weight, carbon dioxide weight, preset maximum tolerable altitude and preset carbon dioxide critical concentration are not specifically limited, and those skilled in the art can set them freely. It is related to the types of microorganisms in food. The preset maximum tolerable altitude indicates the upper limit of the altitude at which microorganisms in food can survive. Exceeding this value indicates that the inhibitory effect of altitude on microorganisms is in a saturated state. The preset carbon dioxide critical concentration is the concentration threshold at which carbon dioxide significantly inhibits microorganisms.

[0038] Specifically, in step S33 of this embodiment, environmental coupling weights are allocated to the carbon dioxide concentration and altitude in the space-time matrix to quantify the synergistic effect of hypoxia and modified atmosphere packaging in plateau transportation and enhance risk sensitivity.

[0039] Please continue reading Figure 1 As shown, the food data processing method based on big data also includes: Step S4, concatenate the temperature characteristics, microbial characteristics and environmental coupling weights into a feature vector, judge the data status and issue an early warning based on the feature vector.

[0040] Specifically, in step S4 described in this embodiment, the temperature characteristics, microbial characteristics and environmental coupling weights are spliced ​​to obtain a feature vector X(t), X(t)=[F(t), G(t), W(t)] is set, and the dot product of the feature vectors with adjacent collection numbers is used as the risk feature value. The expression of the risk feature value is R(t)=X(t)·X(t-1), where R(t) represents the risk feature value.

[0041] It can be understood that the present embodiment does not specifically limit the analysis method of the risk characteristic value, and those skilled in the art can freely set it. For example, the characteristic vector can also be input into the LSTM network model to realize the prediction of food data, judge the similarity between the predicted data and the current data, so as to analyze the risk characteristic value.

[0042] Specifically, in step S4 described in this embodiment, the risk characteristic values ​​of the food cold chain transportation process are extracted, the average value and standard deviation of the extracted risk characteristic values ​​are calculated, and the characteristic value tolerance interval is formulated based on the average value and standard deviation of the extracted risk characteristic values. The characteristic value tolerance interval is set to [μ(R(t))-3×σ(R(t)),μ(R(t))+3×σ(R(t))]. If the risk characteristic value of the current analysis period does not fall within the characteristic value tolerance interval, the data state is judged to be abnormal; otherwise, the data state is judged to be normal; wherein μ() represents the average value of the data in the calculation brackets, and σ() represents the standard deviation of the data in the calculation brackets.

[0043] Specifically, in step S4 described in this embodiment, an early warning is issued based on the data status and the risk characteristic value. When the data status is abnormal, if the risk characteristic value is less than the first characteristic threshold, a severe early warning is issued; if the risk characteristic value is greater than or equal to the first characteristic threshold and less than the second characteristic threshold, a moderate early warning is issued; if the risk characteristic value is greater than or equal to the second characteristic threshold, a mild early warning is issued; when the data status is normal, if the risk characteristic value is less than the first characteristic threshold, a moderate early warning is issued; if the risk characteristic value is greater than or equal to the first characteristic threshold, no early warning is issued.

[0044] Specifically, in this embodiment, the first characteristic threshold is set to 0.2, and the second characteristic threshold is set to 0.9. In this embodiment, the values ​​of the first characteristic threshold and the second characteristic threshold are not specifically limited, and those skilled in the art can set them freely as long as they meet the warning requirement. The value range of the first characteristic threshold should satisfy (0, 0.3], and the value range of the second characteristic threshold should satisfy [0.8, 1].

[0045] Specifically, in step S4 of this embodiment, a risk characteristic value is calculated by constructing a characteristic vector, and a tolerance interval is dynamically set in combination with the 3σ principle. The dynamic tolerance interval is used to adapt to changes in data distribution and reduce misjudgments caused by data drift.

[0046] Please continue reading Figure 1 As shown, the food data processing method based on big data also includes: Step S5, collecting the food status after the food cold chain transportation, the food status includes normal and spoiled, and the food status is collected by user interactive input.

[0047] Please continue reading Figure 1As shown, the food data processing method based on big data also includes: Step S6, the analysis process of updating the data status according to the food status, environmental data and spatiotemporal data.

[0048] See also Figure 3 As shown, it is a method for updating data status, including: Step S61, analyzing the state correlation coefficient based on the risk characteristic value and the food state.

[0049] Specifically, in step S61 of this embodiment, the risk characteristic values ​​are divided into two categories according to the food state: normal characteristic values ​​and spoiled characteristic values. The number of normal characteristic values ​​that do not belong to the characteristic value tolerance interval is counted as the first type of abnormal number, and the number of spoiled characteristic values ​​that do not belong to the characteristic value tolerance interval is counted as the second type of abnormal number, and the state correlation coefficient is analyzed. The expression of the state correlation coefficient is S=σ(R1(t)) Nr1 / NR1 / σ(R2(t)) Nr2 / NR2 , where S represents the state correlation coefficient, R1(t) represents the normal eigenvalue, Nr1 represents the number of type I anomalies, NR1 represents the number of normal eigenvalues, R2(t) represents the deteriorated eigenvalue, Nr2 represents the number of type II anomalies, and NR2 represents the number of deteriorated eigenvalues.

[0050] Specifically, in step S61 of this embodiment, the reliability of model prediction is quantified based on the statistical association between abnormal feature values ​​and actual food status to distinguish occasional noise from real deterioration.

[0051] Please continue reading Figure 3 As shown, the data status updating method further includes: Step S62, updating the analysis process of the data state according to the state correlation coefficient.

[0052] Specifically, in step S62 described in this embodiment, the state correlation coefficient is compared with the correlation threshold. When the state correlation coefficient is greater than the correlation threshold, the feature tolerance interval is updated to [μ(R(t))-(3-S)×σ(R(t)),μ(R(t))+(3+S)×σ(R(t))] to update the analysis process of the data state. Otherwise, the analysis process of the data state is not updated.

[0053] Specifically, in this embodiment, the correlation threshold is set to 0.7. In this embodiment, the value of the correlation threshold is not specifically limited, and those skilled in the art can freely set it. It only needs to satisfy the update of the analysis process of the data state. The value of the correlation threshold should satisfy [0.6, 1).

[0054] Specifically, in step S62 of this embodiment, the characteristic tolerance interval is dynamically updated by comparing and analyzing the state association coefficient, so that the characteristic tolerance interval is related to the association characteristics of food data between different food states.

[0055] Please continue reading Figure 3 As shown, the data status updating method further includes: Step S63, the analysis process of updating the state correlation coefficient according to the environmental data and the transportation duration.

[0056] Specifically, in step S63 described in this embodiment, the transport time is used as an independent variable, and the product of humidity and carbon dioxide concentration is used as a dependent variable for regression analysis to analyze the regression equation of the product of humidity and carbon dioxide concentration on the transport time, and calculate the determination coefficient of the regression equation. If the determination coefficient is greater than the determination threshold, the analysis process of the state correlation coefficient is not updated. Otherwise, the expression of the state correlation coefficient is updated to S=r×σ(R1(t)) Nr1 / NR1 / σ(R2(t)) Nr2 / NR2 , where r represents the coefficient of determination.

[0057] Specifically, in this embodiment, the decision threshold is set to 0.8. In this embodiment, there is no specific limitation on the value of the decision threshold, and those skilled in the art can freely set it. It only needs to satisfy the analysis of the state correlation coefficient. The value of the decision threshold should satisfy [0.7, 0.9].

[0058] Specifically, in step S63 of this embodiment, by modeling the synergistic effect of humidity and carbon dioxide concentration in long-distance transportation, the expression of the state correlation coefficient is dynamically adjusted to solve the problem of model failure under sudden environmental changes.

[0059] See also Figure 4 As shown, it is a food data processing device based on big data in this embodiment, including: The collection module is used to collect environmental data, food data and spatiotemporal data during the food cold chain transportation process and the food status after the food cold chain transportation; A fusion module, used for performing heterogeneous fusion of the collected data to construct a spatiotemporal matrix and a food association matrix, the fusion module being connected to the collection module; An analysis module is used to extract dynamic features and allocate environmental coupling weights to the spatiotemporal matrix to obtain temperature features, microbial features and environmental coupling features, and the analysis module is connected to the fusion module; An early warning module is used to combine temperature characteristics, microbial characteristics and environmental coupling weights into a feature vector, and is also used to judge the data status and issue an early warning based on the feature vector. The early warning module is connected to the analysis module; The updating module is used to update the analysis process of the data status according to the food status, environmental data and spatiotemporal data, and the updating module is connected with the early warning module.

[0060] The embodiment of the present application also provides a computer-readable storage medium storing instructions, which, when executed on a computer, enables the computer to execute the food data processing method based on big data as described in the above method embodiment.

[0061] It will be appreciated by those skilled in the art that all or some of the steps and systems in the disclosed method above may be implemented as software, firmware, hardware and appropriate combinations thereof. Some physical components or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor or a microprocessor, or may be implemented as hardware, or may be implemented as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or a non-transitory medium) and a communication medium (or a temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as a computer-readable program, a data structure, a program module or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassette, magnetic tape, disk storage or other magnetic storage device, or any other medium that may be used to store desired information and may be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media generally contain computer-readable programs, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0062] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not limitations on the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is impossible to list all the embodiments here. All obvious changes or modifications derived from the technical solution of the present invention are still within the protection scope of the present invention.

Claims

1. A food data processing method based on big data, characterized in that: include: Collect environmental data, food data, and spatiotemporal data during food cold chain transportation; Heterogeneous fusion of the collected data is performed to construct a spatiotemporal matrix and a food association matrix; Dynamic feature extraction and environmental coupling weight allocation are performed on the spatiotemporal matrix to obtain temperature features, microbial features and environmental coupling features; The temperature characteristics, microbial characteristics and environmental coupling weights are spliced ​​into feature vectors, and the data status is judged and an early warning is issued based on the feature vectors; Collect the status of food after cold chain transportation; The analytical process of updating data status based on food status, environmental data and spatiotemporal data.

2. The food data processing method based on big data according to claim 1, characterized in that: The collection time of environmental data and spatiotemporal data is aligned, and the collection time of environmental data and spatiotemporal data is numbered in chronological order to obtain the collection number t. The environmental data and spatiotemporal data of the same collection time are heterogeneously fused to construct the spatiotemporal matrix E(t), and E(t)=[T(t),H(t),C(t),L(t)] is set, where T(t) represents temperature, H(t) represents humidity, C(t) represents carbon dioxide concentration, and L(t) represents altitude.

3. The food data processing method based on big data according to claim 2 is characterized in that: Based on the duration of the preset analysis cycle, the temperature in the spatiotemporal matrix within 1 hour is dynamically extracted to obtain the temperature feature. The expression of the temperature feature is: , where F(t) represents the temperature feature, v represents the first extraction parameter, v∈N and v≤V, V represents the second extraction parameter, V=3600 / t1, t1 represents the duration of the preset analysis cycle, and λ represents the temperature decay factor.

4. The food data processing method based on big data according to claim 3 is characterized in that: Based on the initial microbial content, the temperature in the spatiotemporal matrix and the food association matrix are dynamically extracted to obtain microbial features. The expression of the microbial features is: G(t)=G0×e G1(i)×[T(t)-G2(i)] / G2(i) , where G(t) represents the microbial characteristics and G0 represents the initial microbial content.

5. The food data processing method based on big data according to claim 4 is characterized in that: Environmental coupling weights are assigned to the carbon dioxide concentration and altitude in the space-time matrix to obtain environmental coupling characteristics. The expression of the environmental coupling characteristics is: W(t)=β1×(1-L(t) / Lmax)+β2×C(t) / Cmax, where β1 represents the altitude weight, β2 represents the carbon dioxide weight, β1+β2=1, Lmax represents the preset maximum tolerable altitude, and Cmax represents the preset critical carbon dioxide concentration.

6. The food data processing method based on big data according to claim 5 is characterized in that: The temperature characteristics, microbial characteristics and environmental coupling weights are spliced ​​to obtain the feature vector X(t), set X(t)=[F(t), G(t), W(t)], and the dot product of the feature vectors with adjacent collection numbers is used as the risk feature value. The expression of the risk feature value is R(t)=X(t)·X(t-1); Extract the risk characteristic values ​​in the food cold chain transportation process, calculate the average value and standard deviation of the extracted risk characteristic values, and formulate the characteristic value tolerance interval based on the average value and standard deviation of the extracted risk characteristic values. Set the characteristic value tolerance interval to [μ(R(t))-3×σ(R(t)),μ(R(t))+3×σ(R(t))]. If the risk characteristic value of the current analysis period does not fall within the characteristic value tolerance interval, the data status is determined to be abnormal. Otherwise, the data status is determined to be normal; Issue an early warning based on the data status and risk feature value. When the data status is abnormal, if the risk feature value is less than the first feature threshold, a severe early warning is issued. If the risk feature value is greater than or equal to the first feature threshold and less than the second feature threshold, a moderate early warning is issued. If the risk feature value is greater than or equal to the second feature threshold, a mild early warning is issued. When the data status is normal, if the risk characteristic value is less than the first characteristic threshold, a moderate warning is issued; if the risk characteristic value is greater than or equal to the first characteristic threshold, no warning is issued.

7. The food data processing method based on big data according to claim 6 is characterized in that: The risk characteristic values ​​are divided into two categories according to the food status: normal characteristic values ​​and spoiled characteristic values. The number of normal characteristic values ​​that do not belong to the characteristic value tolerance interval is counted as the first type of abnormal number, and the number of spoiled characteristic values ​​that do not belong to the characteristic value tolerance interval is counted as the second type of abnormal number. The state correlation coefficient is analyzed. The expression of the state correlation coefficient is S=σ(R1(t)) Nr1 / NR1 / σ(R2(t)) Nr2 / NR2 , where S represents the state correlation coefficient, R1(t) represents the normal characteristic value, Nr1 represents the number of type I anomalies, NR1 represents the number of normal characteristic values, R2(t) represents the deteriorated characteristic value, Nr2 represents the number of type II anomalies, and NR2 represents the number of deteriorated characteristic values; The transport time is taken as the independent variable and the product of humidity and carbon dioxide concentration is taken as the dependent variable for regression analysis to analyze the regression equation of the product of humidity and carbon dioxide concentration on the transport time and calculate the determination coefficient of the regression equation. If the determination coefficient is greater than the determination threshold, the analysis process of the state correlation coefficient is not updated. Otherwise, the expression of the state correlation coefficient is updated to S=r×σ(R1(t)) Nr1 / NR1 / σ(R2(t)) Nr2 / NR2 , where r represents the coefficient of determination.

8. The food data processing method based on big data according to claim 7 is characterized in that: The state correlation coefficient is compared with the correlation threshold. When the state correlation coefficient is greater than the correlation threshold, the feature tolerance interval is updated to [μ(R(t))-(3-S)×σ(R(t)),μ(R(t))+(3+S)×σ(R(t))] to update the analysis process of the data state. Otherwise, the analysis process of the data state is not updated.

9. A food data processing device based on big data, characterized in that: include: The collection module is used to collect environmental data, food data and spatiotemporal data during the food cold chain transportation process and the food status after the food cold chain transportation; The fusion module is used to perform heterogeneous fusion of the collected data to construct a spatiotemporal matrix and a food association matrix; The analysis module is used to extract dynamic features and assign environmental coupling weights to the spatiotemporal matrix to obtain temperature features, microbial features, and environmental coupling features; The early warning module is used to combine temperature characteristics, microbial characteristics and environmental coupling weights into a feature vector, and is also used to judge the data status and issue an early warning based on the feature vector; The update module is used to update the analysis process of data status based on food status, environmental data and spatiotemporal data.

10. A storage medium, characterized in that: Instructions are stored, which, when executed on a computer, enable the computer to execute the food data processing method based on big data as described in any one of claims 1 to 8.

Images