Fresh mail temperature monitoring method, system and device and medium

Abstract

The invention provides a fresh food mail temperature monitoring method, system and device and a medium. The method comprises the steps that temperature curves, mail circulation details and transportation storage information of internal parts in the transportation process of fresh food mails and deterioration time of various types of fresh food at different temperatures are collected; preprocessing each fresh mail for the collected data to obtain transportation data and classification labels of each fresh mail containing time series; training the neural network model until the neural network model reaches convergence, and obtaining a trained mail temperature prediction model for predicting the temperature of the mail internals at future moments; according to the method, a set of low-cost single-grade temperature prediction model for the fresh product packaging box internal parts is established, the blank of temperature monitoring is filled by predicting the internal part temperature, and when the predicted temperature is abnormal, measures are taken in time to effectively reduce quality decay.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and particularly to a method, device, equipment and medium for monitoring the temperature of fresh food mails. Background Art

[0002] With the development of the express delivery industry and the improvement of people's living standards, the demand for sending fresh products has increased sharply. At present, there are two transportation methods: cold chain transportation and non-cold chain transportation. The cold chain has full-process refrigeration equipment. Non-cold chain transportation effectively maintains the quality of perishable foods, ensures quality and safety, and reduces process losses by shortening the transportation time limit. The core of both methods lies in temperature. For fresh products transported by non-cold chain, being in their suitable temperature environment can ensure that they are not easily spoiled. However, due to the lack of effective temperature monitoring measures, the spoilage of fresh agricultural products will cause the risk of claims for express delivery companies and even the loss of major customers; for users who purchase fresh agricultural products, there is currently only single express delivery time and location information without temperature monitoring.

[0003] With the development of information and communication technologies, temperature monitoring during transportation has become one of the important research topics. At present, traditional temperature monitoring systems mainly use remote monitoring technologies such as Wireless Sensor Network (WSN) and Radio Frequency Identification (RFID) to communicate between the underlying sensors and the upper-layer Internet. For example, wireless sensors such as Zigbee and RFID tags are applied to the real-time monitoring of cold chain logistics to achieve online monitoring and non-contact information transmission.

[0004] However, a single sensor can only obtain historical environmental information at local points (such as inside the carriage), but only monitors the real-time temperature of the environment such as the carriage rather than the actual temperature of the goods, which will cause many fresh products to deteriorate during transportation, resulting in economic and customer losses.

[0005] In view of this, there is an urgent need to provide a method for monitoring the contents of fresh food mails and realizing the prediction of fresh food spoilage according to the temperature monitoring situation. Summary of the Invention

[0006] To overcome the problems existing in the related technologies, the present disclosure provides a method, device, equipment and medium for monitoring the temperature of fresh food mails to solve the technical problems in the related technologies.

[0007] One or more embodiments of this specification provide a method for monitoring the temperature of fresh food mails, including the following steps:

[0008] Data collection: Collect the temperature curve of the contents inside each fresh food mail during transportation, the mail transfer details, the transportation and storage information, and the spoilage time of various types of fresh food at different temperatures. Among them, the temperature curve includes the initial temperature at the time of delivery, the temperature of the contents inside the mail at multiple time periods during transportation, and the temperature information of the contents inside the mail at the time of delivery; the mail transfer details and transportation and storage information include the express information, environmental information, refrigerant information, and content information of each fresh food mail; the temperature of the contents inside each fresh food mail is collected by a temperature sensing device set inside the mail.

[0009] Data preprocessing: Preprocess the collected data for each fresh food mail to obtain the transportation data including time series and classification labels for each fresh food mail. Among them, the classification label is spoiled or unspoiled, and information about the abnormal temperature time point, spoilage time, and location of the spoiled mail; and

[0010] Model training: Input the processed data as training data into the neural network model to be trained, and perform supervised iterative training based on the output prediction results and classification labels of the neural network model to be trained, and update the model parameters of the neural network model until the neural network model converges, obtaining a trained mail temperature prediction model for predicting the temperature of the contents inside the mail at future times.

[0011] One or more embodiments of this specification provide a fresh food mail temperature monitoring system, including:

[0012] A cloud platform, a temperature monitoring terminal, and a mail information collection terminal that are communicatively connected to each other;

[0013] The temperature monitoring terminal receives the temperature data collected by the temperature sensing device set inside the mail, and synchronizes the collected data to the temperature monitoring terminal and the mail information collection terminal;

[0014] The mail information collection terminal is used for mail information entry and management, including the temperature curve of the contents inside each fresh food mail during transportation, the mail transfer details, the transportation and storage information, and the spoilage time of various types of fresh food at different temperatures, and sends the mail information to the cloud platform; among them,

[0015] The temperature curve includes the initial temperature at the time of delivery, the temperature of the contents inside the mail at multiple time periods during transportation, and the temperature information of the contents inside the mail at the time of delivery; the mail transfer details and transportation and storage information include the express information, environmental information, refrigerant information, and content information of each fresh food mail;

[0016] The cloud platform includes:

[0017] A data preprocessing module is used to preprocess the received mail information for each fresh food mail, and obtain the transportation data and classification labels including time series for each fresh food mail. The classification labels are spoiled or unspoiled, and the abnormal temperature time points, spoilage time and location of the spoiled mails.

[0018] A model training module is used to input the processed data as training data into the neural network model to be trained, and perform supervised iterative training according to the output prediction results and classification labels of the neural network model to be trained, and update the model parameters of the neural network model until the neural network model converges, so as to obtain a pre-trained mail temperature prediction model for predicting the temperature of the mail contents at a future time.

[0019] One or more embodiments of this specification provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the fresh food mail temperature monitoring method as described above is implemented.

[0020] One or more embodiments of this specification provide a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the fresh food mail temperature monitoring method as described above is implemented.

[0021] The fresh food mail temperature monitoring method, system, device and medium provided by this disclosure have the advantage that by collecting the temperature curves, mail transfer details and transportation and storage information of each fresh food mail during the transportation process, and considering the factors such as the inner packaging situation, refrigerant selection, and refrigerant quantity of the fresh food mail that have a direct impact on the fresh food temperature, the transportation data including time series for each fresh food mail is obtained. And according to the classification labels of the mail-related spoilage information, after data preprocessing, it is input into the model for training, so that the model can learn the relationship between the temperature, packaging and other transportation conditions of each fresh food in the corresponding delivery environment and whether it spoils, and the change of the temperature of the inner contents over time under the corresponding transportation conditions. Compared with the prior art that evaluates the spoilage of each mail based on a small amount of data or data with little correlation, it has higher accuracy. The disclosure establishes a set of low-cost single-item-level temperature prediction models for the inner contents of the packaging boxes of fresh food products. By predicting the temperature of the inner contents, the blank of temperature monitoring is filled, and measures can be taken in time when the predicted temperature is abnormal to effectively reduce quality spoilage. Description of the Drawings

[0022] To more clearly illustrate the technical solutions in one or more embodiments of this specification or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments recorded in this specification. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0023] Figure 1 Flowchart of a fresh food mail temperature monitoring method provided for one or more embodiments of this specification;

[0024] Figure 2 Full process block diagram of the fresh food mail temperature monitoring method provided for one or more embodiments of this specification;

[0025] Figure 3 Process block diagram of processing data through feature engineering provided for one or more embodiments of this specification;

[0026] Figure 4 Model network structure diagram of the initial mail temperature prediction model provided for one or more embodiments of this specification;

[0027] Figure 5 Schematic block diagram of a fresh food mail temperature monitoring system provided for one or more embodiments of this specification;

[0028] Figure 6 Schematic diagram of the architecture of the fresh food mail temperature monitoring system provided for one or more embodiments of this specification;

[0029] Figure 7 Interface diagram of the mail binding module provided for one or more embodiments of this specification;

[0030] Figure 8 Diagram showing the display of order details information provided for one or more embodiments of this specification;

[0031] Figure 9 Diagram showing the display of the mail temperature control curve provided for one or more embodiments of this specification; and

[0032] Figure 10 Schematic diagram of the structure of a computer device provided for one or more embodiments of this specification. Detailed implementation manners

[0033] In order to enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the following will clearly and completely describe the technical solutions in one or more embodiments of this specification in conjunction with the accompanying drawings in one or more embodiments of this specification. Obviously, the described embodiments are only a part of the embodiments of this specification, rather than all the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0034] The following will make a detailed description of the present invention in combination with the specific implementation manners and the accompanying drawings of the specification.

[0035] Method Embodiment

[0036] According to an embodiment of the present invention, a method for monitoring the temperature of fresh food mails is provided. As Figure 1 shown, it is a flowchart of the method for monitoring the temperature of fresh food mails provided in this embodiment. The method for monitoring the temperature of fresh food mails according to the embodiment of the present invention includes:

[0037] Step S1, data collection; collect the temperature curve of the inner contents during the transportation of each fresh food mail, the mail transfer details, the transportation and storage information, and the deterioration time of various types of fresh food at different temperatures.

[0038] The temperature curve includes the initial temperature at the time of delivery, the temperature of the inner contents of the mail at multiple time periods during transportation, and the temperature information of the inner contents of the mail at the time of delivery; the mail transfer details and transportation and storage information include the express information, environmental information, refrigerant information, and inner content information of each fresh food mail. Among them, the refrigerant information includes refrigerant selection, refrigerant quantity, packaging method, placement position, etc.; the inner content information includes the inner content category and whether it deteriorates during transportation. Among them, the deterioration information includes abnormal temperature time, deterioration time, and deterioration type (such as frostbite, crushing, spoilage, etc.); the express information includes the direction, transportation method, start transportation time, and expected delivery time; the environmental information includes the temperature of the transportation vehicle, the weather during transportation, etc. The temperature of the inner contents of each fresh food mail can be collected by a temperature sensing device set in the mail.

[0039] Step S2, data preprocessing; perform preprocessing such as data cleaning and standardization on the data collected in step S1 for each fresh food mail to obtain the transportation data and classification labels including time series for each fresh food mail, where the classification labels are deteriorated or not deteriorated, and the temperature abnormal time points, deterioration times, and locations of the deteriorated mails.

[0040] S3. Model training: Use the processed data as training data and input it into the neural network model to be trained. Perform supervised iterative training based on the output prediction results and classification labels of the neural network model to be trained, and update the model parameters of the neural network model until the neural network model converges, obtaining a trained mail temperature prediction model for predicting the temperature of the contents of the mail at future times.

[0041] The fresh food mail temperature monitoring method provided in this embodiment collects the temperature curves, mail transfer details, and transportation and storage information during the transportation of each fresh food mail. Among them, factors directly affecting the fresh food temperature, such as the inner packaging situation, refrigerant selection, and refrigerant quantity of the fresh food mail, are considered to obtain the transportation data including time series for each fresh food mail. And according to the classification label of the mail-related spoilage information, after data preprocessing, it is input into the model for training, enabling the model to learn the relationship between the temperature, packaging, and other transportation conditions of each fresh food in the corresponding delivery environment and whether it spoils, as well as the change of the temperature of the contents of the mail over time under the corresponding transportation conditions. Compared with the prior art that evaluates the spoilage of each mail based on a small amount of data or data with little relevance, it has higher accuracy. The method in this embodiment establishes a low-cost single-item-level temperature prediction model for the contents of the packaging box of fresh food products. By predicting the temperature of the contents, it fills the gap in temperature monitoring and takes timely measures when the predicted temperature is abnormal to effectively reduce quality spoilage.

[0042] This embodiment further includes a temperature and spoilage prediction step, which is specifically as follows:

[0043] Step S4. Temperature and spoilage prediction: Based on the collected fresh food mail transfer details, transportation and storage information, and the temperature information of the contents of the fresh food mail collected within a preset time period.

[0044] This embodiment uses the trained model to predict the temperature of the contents of the mail for the fresh food during transportation at future times, and predicts the pre-spoilage time of the fresh food when the temperature is abnormal, which can enable timely measures to be taken before the abnormality occurs, effectively reducing quality spoilage, reducing the customer complaint rate, reducing the compensation amount, and ensuring product quality.

[0045] In this embodiment, since fresh food products have relatively strict requirements for storage temperature conditions, the length of the time interval from when the fresh food is picked up to the start of transportation is also a key factor affecting the quality of fresh food. If the fresh food is not stored under the specified conditions for a long time after being picked up, it will also affect the quality of fresh food. Therefore, the fresh food mail information in this embodiment also includes the packing time and the time interval from the packing time. Considering these two factors will result in higher accuracy of the prediction results.

[0046] In order to enable the model to more accurately predict the time of temperature anomaly and the deterioration time when the temperature is abnormal, this embodiment also includes using the temperature anomaly time obtained by model prediction and the temperature anomaly time point of the mail contents collected by the temperature sensing device as parameters of the model to adaptively adjust the model, so as to realize online adaptive optimization training of the model, so that the model predicts the temperature anomaly time point with higher accuracy, thereby predicting the deterioration time of fresh food more accurately.

[0047] In this embodiment, since the data collected and analyzed in step S1 are multi-dimensional and have complex relationships, in order to extract more key and informative features, remove redundant data, and better determine the relationship between features, the data obtained by the feature engineering team is processed, and whether the fresh food deteriorates during transportation is determined based on the processed data to establish deterioration label data. Finally, the model obtained by training the processed data can realize the temperature prediction of the fresh food in the future and the pre-deterioration time of the fresh food under abnormal temperature. Therefore, step S2 specifically includes the following steps:

[0048] Step S21, performing feature engineering processing on the data obtained in step S1;

[0049] For details, please refer to Figure 2 and Figure 3 As shown, there are respectively a full-process flowchart of the fresh mail temperature monitoring method provided in this embodiment and a flowchart of processing data through feature engineering, wherein the specific process of processing data through feature engineering is as follows:

[0050] (1) Detect and process missing temperature values, abnormal temperature sent back by sensors, and other data;

[0051] (2) Standardize and normalize dimensionless features, such as temperature data;

[0052] (3) Binarization and dummy coding of quantitative features, such as packaging method and packaging time;

[0053] (4) Feature selection stage: Use filtering, wrapping and other feature selection methods to identify and select the most critical and informative features for model prediction from the features obtained through the above processing. This step is crucial because it not only helps improve the performance of the model and reduce the risk of overfitting, but also reduces the complexity of the model and improves training efficiency.

[0054] (5) Feature dimensionality reduction stage: Finally, principal component analysis (PCA) or linear discriminant analysis is used to achieve feature dimensionality reduction, eliminating redundant features such as refrigerant temperature, and retaining the features that best represent the essence of the data and are most valuable for the prediction task.

[0055] The features of each fresh food email finally selected through the above feature engineering include environmental temperature, refrigerant quantity, placement location, whether there is an aluminum foil thermal insulation bag, packing time, time interval from the packing time, normal / anomalous temperature time, spoilage time, and spoilage type, etc. These features in multiple time periods constitute a time series, thus extracting the time series features that reflect the temperature change of the inner contents and affect the temperature of the inner contents, and performing embedded representation on the categorical features so as to be able to input into the model together with the numerical features.

[0056] Step S22, establish classification labels for the feature data processed by feature engineering according to whether fresh food spoilage occurs.

[0057] In this embodiment, considering that the data exhibits obvious time series dynamic characteristics, models specifically designed for time series analysis can be adopted, such as autoregressive integrated moving average model (ARIMA), seasonal autoregressive integrated moving average model (SARIMA), and recursive neural network (RNN), long short-term memory network (LSTM), gated recurrent unit (GRU), etc. In this embodiment, since deep learning models such as LSTM and GRU show the following significant advantages compared with the traditional ARIMA model:

[0058] (1) Handling non-linearity and complex patterns: As deep learning models, LSTM and GRU can capture complex non-linear relationships in time series, while the ARIMA model is mainly applicable to linear time series data.

[0059] (2) Long-term dependence relationships: The gating mechanisms of LSTM and GRU can effectively solve the problem of gradient disappearance, so as to be able to capture and utilize long-term dependence relationships, which is very important for time series analysis.

[0060] (3) Flexibility and adaptability: LSTM and GRU models can adapt to different data features and prediction tasks by adjusting the network structure and parameters, while the parameter estimation process of the ARIMA model is relatively fixed and has limited adaptability to data.

[0061] (4) Robustness: LSTM and GRU show better robustness in the face of noise and outliers in data, while the ARIMA model is more sensitive to these interferences.

[0062] (5) Generalization ability: Deep learning models have proven their powerful generalization ability in multiple fields and tasks and are more effective in dealing with diverse time series data.

[0063] (6) End-to-end learning: LSTM and GRU models support end-to-end learning processes and can directly learn feature representations from raw data, while the ARIMA model requires manual feature engineering and parameter adjustment.

[0064] Based on the analysis of the collected temperature data in the embodiments, in order to process its long-term dependencies, non-linear relationships, and adapt to complex sequence data, it is more suitable to select a deep learning model for construction. Through the analysis of the model structures of LSTM and GRU, and the determination of the training speed and training resources, GRU has fewer training parameters and lower computational complexity compared to LSTM, which can improve the training efficiency. At the same time, there is no significant difference in the prediction accuracy. Therefore, the GRU model is finally selected as the initial mail temperature prediction model.

[0065] In one embodiment, referring to Figure 4 As shown, it is the model network structure diagram of the initial mail temperature prediction model provided in this embodiment. The initial mail temperature prediction model includes an input layer, a hidden layer, and an output layer. The input layer receives multiple features of time series data and transmits them as inputs to the corresponding GRU units respectively. The GRU units are connected in series according to time steps to achieve sequence connection, so that each GRU unit receives the output from the previous time step and the input of the current time step, and determines which information should be retained or forgotten through the update gate and the reset gate, so as to capture the long-term dependencies in the time series data; the output layer is the temperature of the mail contents at the current time step, and the output layer is responsible for generating the final prediction result according to the output of the hidden layer, including the temperature value of the contents at the future t moment, whether the temperature is abnormal, and the fresh produce spoilage time in the case of temperature abnormality.

[0066] In this embodiment, the initial mail temperature prediction model uses the weighted cross-entropy loss function as the loss function during the training of the GRU prediction model, uses the Adam adaptive optimization algorithm to train the model, and adopts the L2 regularization and EarlyStopping strategies to prevent overfitting.

[0067] In a specific embodiment, 6 GRU units are set in the input layer to realize using the data of the previous 6 time steps of the current feature data as input features.

[0068] In this embodiment, according to the prediction results, corresponding measures can also be taken when the temperature is abnormal to reduce the compensation for fresh agricultural product mails caused by temperature abnormalities. Display the data of the whole process of fresh agricultural products to assist business marketing.

[0069] This embodiment can also output a more effective and accurate packaging configuration plan for different environmental conditions, transportation directions, and fresh product categories, reasonably configure the types and quantities of packaging materials and refrigerants, and reduce the refrigerant cost. For example, during the prediction stage, the internal temperature prediction curve of the mail can be obtained by changing the input of different packaging configuration plans, so as to select the best packaging configuration plan according to information such as the fresh produce transportation direction, duration, and spoilage temperature. There is no doubt that the process of training and using the model is the same as the above temperature prediction process, except that the training data and output results used are different, which will not be elaborated here.

[0070] System embodiment

[0071] According to an embodiment of the present invention, a fresh food mail temperature monitoring system is provided. As Figure 5 shown, it is a schematic block diagram of the fresh food mail temperature monitoring system provided in this embodiment. The fresh food mail temperature monitoring system according to an embodiment of the present invention includes:

[0072] A cloud platform 10, a temperature monitoring terminal 20, and a mail information collection terminal 30 that are communicatively connected to each other;

[0073] The temperature monitoring terminal 20 receives the temperature data collected by the temperature sensing device arranged in the mail contents, and synchronizes the collected data to the temperature monitoring terminal 20 and the mail information collection terminal 30;

[0074] The mail information collection terminal 30 is used for mail information entry and management, including the temperature curve of the contents during the transportation of each fresh food mail, the mail transfer details, the transportation and storage information, and the deterioration time of various types of fresh food at different temperatures, and sends the mail information to the cloud platform 10; among them,

[0075] The temperature curve includes the initial temperature at the time of delivery, the temperature of the mail contents at multiple time periods during transportation, and the temperature information of the mail contents at the time of delivery; the mail transfer details and transportation and storage information include the express information, environmental information, refrigerant information, and contents information of each fresh food mail. Among them, the refrigerant information includes refrigerant selection, refrigerant quantity, packaging method, placement location, etc.; the contents information includes the contents category and whether it deteriorates during transportation. Among them, the deterioration information includes abnormal temperature time, deterioration time, and deterioration type (such as frostbite, crushing, spoilage, etc.); the express information includes the direction, transportation method, start transportation time, and expected delivery time; the environmental information includes the temperature of the transportation vehicle, the weather during transportation, etc.

[0076] The cloud platform 10 is used for analyzing and processing the data reported by various devices, specifically including a data preprocessing module, a model training module, and a prediction module;

[0077] The data preprocessing module is used for preprocessing the received mail information, performing data cleaning and standardization on each fresh food mail, etc., to obtain the transportation data and classification labels including time series of each fresh food mail, where the classification labels are deteriorated or not deteriorated, and the temperature abnormal time points, deterioration time, and locations of the deteriorated mails.

[0078] A model training module is used to input the processed data as training data into a neural network model to be trained, and perform supervised iterative training based on the output prediction results and classification labels of the neural network model to be trained, update the model parameters of the neural network model until the neural network model converges, and obtain a pre-trained mail temperature prediction model for predicting the temperature of the mail contents at a future time.

[0079] A prediction module is used to predict the temperature of the contents at a future time using the pre-trained and optimized temperature prediction model based on the fresh food mail transfer details, transportation and storage information sent by the temperature monitoring terminal 20 and the mail information collection terminal 30, and determine temperature anomalies according to a preset temperature anomaly threshold, and predict the fresh food spoilage time when the predicted temperature is abnormal or when the temperature sensing device collects abnormal temperature.

[0080] The fresh food mail temperature monitoring system provided in this embodiment collects the temperature curves, mail transfer details, and transportation and storage information of each fresh food mail during the transportation process through the temperature monitoring terminal 20 and the mail information collection terminal 30. Among them, factors such as the inner packaging situation, refrigerant selection, and refrigerant quantity of the fresh food mail, which have a direct impact on the fresh food temperature, are considered, and the transportation data including time series of each fresh food mail is obtained. And according to the mail-associated spoilage information classification label, after the cloud platform 10 performs data preprocessing, it is input into the model for training, so that the model learns the relationship between the temperature, packaging and other transportation conditions of each fresh food in the corresponding delivery environment and whether it spoils, and the change of the temperature of the contents over time under the corresponding transportation conditions. Compared with the prior art that evaluates the spoilage of each mail based on a small amount of data or data with little relevance, it has higher accuracy; then, through the trained model, the temperature of the mail contents at a future time during the transportation of fresh food is predicted, and the pre-spoilage time of fresh food is predicted when the temperature is abnormal. The system in this embodiment establishes a low-cost single-item-level temperature prediction model for the contents of the packaging box of fresh food products. By predicting the temperature of the contents, it fills the gap in temperature monitoring, takes measures in time before anomalies occur, effectively reduces quality spoilage, reduces the customer complaint rate, reduces the compensation amount, and ensures product quality.

[0081] In this embodiment, the mail information collection terminal 30 can be a mobile APP terminal, which is used to realize the management of pickup and delivery orders and the binding of the temperature collection device and the mail.

[0082] In this embodiment, since fresh food products have relatively strict requirements for storage temperature conditions, the length of the time interval from fresh food pickup to the start of transportation is also a key factor affecting the quality of fresh food. If the fresh food is not stored under the specified conditions for a long time after pickup, it will also affect the quality of fresh food. Therefore, the fresh food mail information in this embodiment also includes the packing time collected by the mail information collection terminal 30 and the time interval from the packing time. Considering these two factors will result in higher accuracy of the prediction results.

[0083] In this embodiment, the data preprocessing module is configured to implement preprocessing through feature engineering, and establish classification labels for the feature data after feature engineering processing according to whether fresh food deterioration occurs.

[0084] In one embodiment, features of each fresh food mail are obtained through feature engineering screening, including environmental temperature, refrigerant quantity, placement location, whether there is an aluminum foil insulation bag, packing time, time interval from the packing time, normal / abnormal temperature time, deterioration time, and deterioration type, etc.; it also includes performing embedded representation on categorical features so that they can be input into the model together with numerical features.

[0085] The system of this embodiment refers to Figure 6 As shown, it is a schematic diagram of the architecture of the fresh food mail temperature monitoring system provided in this embodiment. This system enables different types of devices to communicate with each other and access the Internet through the Internet of Things gateway, that is, the bridge between the device and the cloud platform, which mainly undertakes the functions of protocol conversion and edge computing. The Internet of Things perception devices use different communication technologies in different scenarios. Common communication technologies include: BLE, ZigBee, LoRa, Wi-Fi, RS485, RFID, etc. Among them, devices that do not have an IP address by themselves cannot be directly connected to the Internet and need to implement protocol conversion through a gateway device that supports different communication technologies, so as to indirectly realize the networking of the device and communicate with the cloud platform. The Internet of Things gateway of this project is a Bluetooth gateway for label data collection. This device also undertakes the role of edge computing and has functions such as gateway configuration, local identity authentication, security management, data storage, data encryption transmission, and data comparison, which not only ensures the safe and reliable transmission of data but also provides support for remote maintenance.

[0086] The temperature sensing device can communicate with the temperature monitoring terminal 20 at regular intervals through low-power Bluetooth and the Bluetooth gateway.

[0087] In this embodiment, the mail information collection terminal 30 is provided with a mail binding module, a pickup and delivery order query module, a mail delivery information collection module, and an abnormal mail query module;

[0088] The mail binding module is used to implement the binding of the mail with the temperature sensing device (temperature control label). The binding information of the temperature sensing device (temperature control label) includes information such as the label number, binding time, and initial binding temperature of the temperature sensing device.

[0089] For reference Figure 7As shown in the figure, it is the interface diagram of the mail binding module provided by the embodiment. After the mail is bound to the temperature control label through the mail binding module, the mail information collection terminal 30 determines whether the temperature is abnormal according to the temperature information fed back by the temperature control label and various preset fresh food temperature abnormality thresholds, and issues an alarm; when abnormal temperature is detected, it also feeds back to the cloud platform 10, and the cloud platform 10 will predict the fresh food deterioration time through the temperature prediction model for the predicted temperature abnormality or the temperature abnormality information fed back by the mail information collection terminal 30.

[0090] In a specific embodiment, the mail binding module can identify the barcode on the waybill of the mail by scanning code and realize the binding of the mail information and the mail order number. After the system queries that the mail is a valid waybill without label binding, it will display the waybill information on the sender details page.

[0091] The mail delivery information collection module is used for entering mail information. Among them, the mail information may include mail transfer details information, transportation and storage information, and the deterioration time information of fresh food at different temperatures; the mail transfer details and transportation and storage information include express information, environmental information, refrigerant information, and inner part information of each fresh food mail.

[0092] The pickup and delivery order query module is used for querying the transportation information of the mail order, which may include waybill information, temperature control label information with normal temperature, and the mail transfer details list. The temperature control label information with normal temperature is displayed with a green background, and the temperature control label information with abnormal temperature is displayed with a red background. Among them, the temperature control label information includes label number, current temperature, collection location, and collection time; specifically, refer to Figure 8 As shown in the figure, it is the order details information display diagram provided by this embodiment.

[0093] In an embodiment, the pickup and delivery order query module further includes a temperature control curve query sub-module, which is used to generate a temperature change curve graph from the time when the mail is bound to the temperature control label to the current query time according to the temperature information collected by the temperature sensing device. Specifically, refer to Figure 9 As shown in the figure, it is the mail temperature control curve display diagram provided by this embodiment.

[0094] The abnormal mail query module is used for querying and displaying the data list of temperature abnormal mails, which may include waybill information, temperature control label information with abnormal temperature, the mail transfer details list, temperature control curves and other contents.

[0095] The embodiment of the present invention is the device embodiment corresponding to the above method embodiment. The specific operations of the processing steps of each module can be understood with reference to the description of the method embodiment, and will not be elaborated here.

[0096] Such as Figure 10As shown, the present invention also provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the fresh mail temperature monitoring method in the above embodiments.

[0097] The present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the fresh mail temperature monitoring method in the above embodiments, or when the computer program is executed by a processor, it implements the fresh mail temperature monitoring method in the above embodiments.

[0098] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0099] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other, and the key point of each embodiment is to illustrate the differences from other embodiments. In particular, for the device or system embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and for the relevant parts, reference can be made to the description of the method embodiments. The device and system embodiments described above are only illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative work.

[0100] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features. These modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and the content not detailedly described in the specification of the present invention belongs to the well-known technology of those skilled in the art.

Claims

1. A method for monitoring the temperature of fresh mail, characterized in that: The following steps are involved: Data collection: collect the temperature curve of the contents of each fresh mail during transportation, mail flow details, transportation and storage information, and the deterioration time of each type of fresh food at different temperatures; the temperature curve includes the initial temperature at delivery, the temperature of the mail contents at multiple time periods during transportation, and the temperature information of the mail contents at delivery; the mail flow details and transportation and storage information include the express information, environmental information, refrigerant information, and contents information of each fresh mail; the temperature of the contents of each fresh mail is collected by the temperature sensing device installed in the mail; Data preprocessing: preprocess the collected data for each fresh mail, and obtain the transportation data and classification labels of each fresh mail including time series, where the classification labels are spoiled or not spoiled, and information about the time point of abnormal temperature, spoilage time and location of spoiled mail; Model training: The processed data is input as training data to the neural network model to be trained, and supervised iterative training is performed according to the output prediction results and classification labels of the neural network model to be trained, and the model parameters of the neural network model are updated until the neural network model reaches convergence, and the trained mail temperature prediction model is obtained for the temperature prediction of mail contents in the future.

2. The method for monitoring the temperature of fresh mail according to claim 1, characterized in that: The refrigerant information includes refrigerant selection, refrigerant quantity, packaging method, and placement location; the internal content information includes the internal content category and whether it has deteriorated during transportation, wherein the deterioration information includes abnormal temperature time, deterioration time and deterioration type; the express delivery information includes route direction, transportation method, start transportation time and expected delivery time.

3. The method for monitoring the temperature of fresh mail according to claim 2, characterized in that: The following steps are also included: Temperature and deterioration prediction: Based on the collected fresh mail flow details, transportation and storage information, and the temperature information of the fresh mail contents collected within a preset time period, the pre-trained and optimized temperature prediction model is used to predict the temperature of the contents at future times, and the temperature anomaly is determined based on the preset temperature anomaly thresholds for various types of fresh food. When a temperature anomaly is predicted or the temperature sensing device detects an abnormal temperature, the time of fresh food deterioration is predicted.

4. The method for monitoring the temperature of fresh mail according to claim 1, characterized in that: The data preprocessing is performed through feature engineering methods, including data cleaning, feature construction and feature scaling.

5. The method for monitoring the temperature of fresh mail according to claim 1, characterized in that: The neural network model includes an input layer, a hidden layer and an output layer; The input layer receives multiple features of the time series data and passes them to the corresponding GRU units as inputs, and each GRU unit is connected in series according to the time step to achieve sequence connection, so that each GRU unit receives the output from the previous time step and the input of the current time step; The output layer is responsible for generating the final prediction results based on the output of the hidden layer, including the temperature value of the internal parts at time t in the future, whether the temperature is abnormal, and the time when the fresh food will deteriorate when the temperature is abnormal.

6. The method for monitoring the temperature of fresh mail according to claim 1, characterized in that: It also includes model adaptive optimization steps, as follows: The temperature anomaly time obtained through model prediction and the temperature anomaly time point of the mail contents collected by the temperature sensing device are used as parameters of the model for adaptive adjustment of the model to achieve online adaptive optimization training of the model.

7. A fresh mail temperature monitoring system, characterized in that: include: The cloud platform, temperature monitoring terminal and email information collection terminal are interconnected; The temperature monitoring terminal receives the temperature data collected by the temperature sensing device installed in the mail, and synchronizes the collected data to the temperature monitoring terminal and the mail information collection terminal; The mail information collection terminal is used to input and manage mail information, including the temperature curve of the contents of each fresh mail during transportation, mail flow details, transportation and storage information, and the deterioration time of each category of fresh food at different temperatures, and the mail information is sent to the cloud platform; The temperature curve includes the initial temperature at the time of delivery, the temperature of the mail contents at multiple periods during transportation, and the temperature information of the mail contents at the time of delivery; the mail flow details and transportation and storage information include the express information, environmental information, refrigerant information, and contents information of each fresh mail; The cloud platform includes: The data preprocessing module is used to perform data preprocessing on the received mail information for each fresh mail, and obtain the transportation data and classification labels of each fresh mail including the time series, wherein the classification labels are spoiled or not spoiled, and the time point of abnormal temperature, spoilage time and location of spoiled mail; The model training module is used to input the processed data as training data to the neural network model to be trained, perform supervised iterative training according to the output prediction results and classification labels of the neural network model to be trained, update the model parameters of the neural network model until the neural network model reaches convergence, and obtain the pre-trained mail temperature prediction model for predicting the temperature of mail contents in the future.

8. The fresh food mail temperature monitoring system according to claim 7, characterized in that: The cloud platform also includes The prediction module is used to predict the temperature of the contents of fresh mail at future times based on the fresh mail flow details, transportation and storage information sent by the temperature monitoring terminal and the mail information collection terminal, and the temperature information of the contents of fresh mail collected within a preset period of time, using a pre-trained and optimized temperature prediction model, and determine the temperature anomaly according to the preset temperature anomaly threshold, and predict the time of fresh food deterioration when the temperature anomaly is predicted or the temperature sensing device collects abnormal temperature.

9. The fresh food mail temperature monitoring system according to claim 7, characterized in that: The email information collection terminal is provided with an email binding module; An email binding module is used to bind emails to temperature sensing devices. The temperature sensing device binding information includes the label number of the temperature sensing device, the binding time, and the initial binding temperature. After the email is bound to the temperature control tag, the email information collection terminal determines whether it is an abnormal temperature and issues an alarm based on the temperature information fed back by the temperature control tag and the preset temperature abnormality thresholds for various types of fresh food. When an abnormal temperature is detected, it also feeds back to the cloud platform, which will use the temperature prediction model to predict the temperature abnormality or the temperature abnormality information fed back by the email information collection terminal to predict the time when the fresh food will deteriorate.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the fresh mail temperature monitoring method according to any one of claims 1 to 6 is implemented.

11. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the fresh mail temperature monitoring method according to any one of claims 1 to 6 is implemented.

Images