Intelligent cold-chain logistics transfer method and platform capable of remote monitoring

Abstract

The application provides a remote monitoring intelligent cold chain logistics transfer method and platform, and belongs to the technical field of cold chain transportation. The method comprises the following steps: collecting multi-dimensional physical data and cold chain environment parameter data of fresh goods, obtaining attributes, transportation requirements, and a transfer stage of the fresh goods; determining a state monitoring threshold and an environment control target according to the transfer stage, the attributes, and the transportation requirements of the fresh goods; performing feature extraction and fusion on the multi-dimensional physical data and the cold chain environment parameter data to obtain fused feature data; performing linkage analysis and comprehensive judgment on the fused feature data based on the transfer stage, the state monitoring threshold, the environment control target, and a preset association rule to obtain a state category of the fresh goods and a stage compliance evaluation, and generating a remote control signal recognizable by a remote monitoring end; and sending the remote control signal to a corresponding cold chain transfer unit and starting a local control action. The application improves the efficiency and reliability of cold chain logistics transfer.

Description

Technical Field

[0001] This application relates to the field of cold chain transportation technology, and in particular to an intelligent cold chain logistics transfer method and platform that can be remotely monitored. Background Technology

[0002] With the booming development of fresh food e-commerce and consumption upgrades, my country's cold chain logistics scale continues to expand, with explosive growth in demand for cross-regional, long-distance, and multi-category transshipment. Fresh produce is extremely sensitive to environmental factors such as temperature and humidity; even slight fluctuations can lead to spoilage. Current technologies rely on single environmental sensors to collect data and monitor through one-way comparisons of fixed thresholds, failing to incorporate information about the physical state of the goods themselves. Furthermore, they cannot adapt to the differentiated needs of different product categories during the delivery stage, further leading to invalid alarms or missed detections. Simultaneously, existing control measures rely on multi-stage manual intervention, resulting in long response times and difficulty in quickly handling anomalies in complex transshipment scenarios. Summary of the Invention

[0003] To address the aforementioned technical issues, this application provides a remotely monitored intelligent cold chain logistics transfer method and platform.

[0004] A first aspect of this application provides a remotely monitorable intelligent cold chain logistics transfer method, comprising: Collect multi-dimensional physical data and cold chain environment parameter data of fresh goods to be transferred, and receive transfer task information from the logistics management platform. The transfer task information includes the attributes, transportation requirements, and transfer stage of the fresh goods to be transferred. Based on the transshipment stage, the attributes of the fresh goods to be transshipped, and the transportation requirements, determine the appropriate status monitoring threshold and environmental control targets; Feature extraction and fusion are performed on the multi-dimensional physical data and the cold chain environment parameter data to obtain fused feature data; Based on the transfer stage, the status monitoring threshold, the environmental control target, and the preset association rules, the fused feature data is analyzed and comprehensively judged to obtain the status category and stage compliance assessment of the fresh goods to be transferred. Based on the state category and the stage compliance assessment, a remote control signal that can be recognized by the remote monitoring terminal is generated. The remote control signal is sent to the cold chain transfer unit corresponding to the current transfer stage through the remote monitoring terminal; in response to the remote control signal, each cold chain transfer unit initiates local control actions and feeds back the execution results to the remote monitoring terminal.

[0005] A second aspect of this application provides a remotely monitored intelligent cold chain logistics transfer platform, comprising: The information collection module is used to collect multi-dimensional physical data and cold chain environment parameter data of the fresh goods to be transferred, and to receive transfer task information from the logistics management platform. The transfer task information includes the attributes, transportation requirements and transfer stage of the fresh goods to be transferred. The strategy generation module is used to determine appropriate status monitoring thresholds and environmental control targets based on the transshipment stage, the attributes of the fresh goods to be transshipped, and the transportation requirements. The data fusion module is used to extract and fuse features from the multi-dimensional physical data and the cold chain environment parameter data to obtain fused feature data. The intelligent analysis module is used to perform linked analysis and comprehensive judgment on the fused feature data based on the transfer stage, the status monitoring threshold, the environmental control target and the preset association rules, so as to obtain the status category and stage compliance assessment of the fresh goods to be transferred. The instruction generation module is used to generate remote control signals that can be recognized by the remote monitoring terminal based on the state category and the stage compliance assessment. The linkage execution module is used to send the remote control signal to the cold chain transfer unit corresponding to the current transfer stage through the remote monitoring terminal; in response to the remote control signal, each cold chain transfer unit starts local control action and feeds back the execution result to the remote monitoring terminal.

[0006] A third aspect of this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of the above-described remotely monitored intelligent cold chain logistics transfer method.

[0007] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described remotely monitorable intelligent cold chain logistics transfer method.

[0008] The beneficial effects of the remotely monitored intelligent cold chain logistics transfer method and platform provided in this application are as follows: This application enhances the targeting of monitoring and control by collecting physical data of fresh goods and cold chain environmental parameters, and matching monitoring thresholds and control targets based on transfer task information, thus avoiding control deviations caused by single data or uniform standards. Furthermore, through feature extraction and fusion, it achieves correlation analysis between cargo status and environmental parameters, and, combined with comprehensive judgment based on preset correlation rules, improves the accuracy and timeliness of identifying cargo status categories and stage compliance. Simultaneously, it generates control signals based on the remote monitoring terminal and sends them to the corresponding cold chain transfer unit, realizing real-time control across different locations. In addition, it receives execution feedback, breaking the geographical limitations and passive monitoring mode of traditional cold chain transfer. This application achieves intelligent and visualized control of the entire stage of cold chain logistics transfer, improving the efficiency and reliability of cold chain logistics transfer. Attached Figure Description

[0009] Figure 1 A schematic flowchart of a remotely monitorable intelligent cold chain logistics transfer method provided in an embodiment of this application; Figure 2 A structural block diagram of a remotely monitorable intelligent cold chain logistics transfer platform provided in an embodiment of this application; Figure 3 This is a schematic block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0010] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0011] To make the purpose, technical solution, and advantages of this application clearer, the following will be described in conjunction with the appendix. Figure 1-3 The following is an explanation using specific examples.

[0012] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating a remotely monitorable intelligent cold chain logistics transfer method according to an embodiment of this application. The method includes: S101: Collect multi-dimensional physical data and cold chain environment parameter data of fresh goods to be transferred, and receive transfer task information from the logistics management platform. The transfer task information includes the attributes of the fresh goods to be transferred, transportation requirements, and transfer stage.

[0013] In this embodiment, the multi-dimensional physical data includes real-time temperature, humidity, texture hardness, freshness-related parameters, and packaging integrity data; cold chain environmental parameter data includes real-time temperature, humidity, gas composition (oxygen and carbon dioxide concentration), air pressure, and cold chain equipment operating status parameters (such as refrigeration unit power and fan speed) within the cold chain transfer unit (refrigerated truck, container, cold storage).

[0014] The transfer task information received from the logistics management platform includes the attributes of the fresh goods to be transferred, transportation requirements, and the transfer stage. The attributes of the fresh goods to be transferred include category (seafood, fruits and vegetables, dairy products), origin, initial quality grade, shelf life, and whether it is a special category (perishable, quick-frozen, requiring constant temperature preservation). Transportation requirements include the target temperature range throughout the journey, humidity control range, allowable temperature fluctuation range, transportation time requirements, and whether light protection / shockproofing is required. The transfer stage includes sea / land / air transport.

[0015] S102: Determine appropriate status monitoring thresholds and environmental control targets based on the transshipment stage, the attributes of the fresh goods to be transshipped, and transportation requirements.

[0016] In this embodiment, the status monitoring threshold and environmental control target are determined using a logic of scenario adaptation, cargo adaptation, and requirement adaptation. Specifically, firstly, basic control benchmarks are determined based on the attributes of the fresh goods to be transported. For example, if it is a storable product (root vegetables, frozen meat), the status monitoring threshold range can be appropriately relaxed. Secondly, the transportation requirements further refine the control standards. For example, if the transportation requirement is a constant temperature of 0-2℃ throughout the journey, then this range is used as the environmental control target.

[0017] Different transit stages necessitate differentiated adaptations between condition monitoring thresholds and environmental control objectives. For example, in the sea transport stage, due to long transit cycles and significant environmental fluctuations, the condition monitoring thresholds incorporate a cumulative exceedance duration dimension (single temperature exceedance less than or equal to 10 minutes, 24-hour cumulative exceedance less than or equal to 30 minutes). Environmental control objectives prioritize fluctuation resistance, while also including environmental control objectives related to equipment corrosion protection in marine salt spray environments. In the land transport stage, characterized by frequent starts and stops and complex road conditions, condition monitoring thresholds are set to enhance tolerance for instantaneous fluctuations, with environmental control objectives focusing on rapid temperature recovery compensation. In the air transport stage, affected by air pressure changes and limited space, condition monitoring thresholds incorporate air pressure adaptability indicators, with environmental control objectives of precise temperature control and rapid adaptation. Furthermore, based on flight connection windows, temporary control objectives are set for the ground transit stage, such as pre-adjusting the cargo temperature to the air cargo hold's suitable range within 20 minutes.

[0018] S103: Perform feature extraction and fusion of multi-dimensional physical data and cold chain environmental parameter data to obtain fused feature data.

[0019] In this embodiment, for multi-dimensional physical data, quality-related features are extracted, including the time-series variation features of temperature change rate, humidity tolerance, texture hardness decay trend, and freshness-related indicators (breathing intensity, pH value). At the same time, abrupt changes in packaging integrity (parameter jumps caused by seal damage) are captured. For cold chain environmental parameter data, the steady-state / transient features of temperature distribution uniformity, humidity fluctuation frequency, gas component (oxygen, carbon dioxide) concentration gradient, and refrigeration unit operating status are extracted, as well as the abnormal fluctuation features of environmental parameters corresponding to external environmental disturbances (salt spray in sea transportation, bumps in land transportation).

[0020] During the feature extraction process, multi-dimensional physical data and cold chain environmental parameter data are transformed into structured feature vectors through methods such as temporal feature decomposition (separation of trend term, periodic term, and random term), statistical feature calculation (mean, variance, peak value), and mutation point detection.

[0021] In this embodiment, a correlation mapping rule is first established based on the attributes of the transit stage and the fresh goods to be transited. For example, the rate of change of goods temperature is time-series aligned with the amplitude of ambient temperature fluctuations, and the freshness index of the fresh goods to be transited is correlated with the concentration of gas components. Then, an attention mechanism fusion method is used to integrate the two types of features. Specifically, the weights of the correlated features (ambient temperature and goods temperature) are used as the main features, and redundant features are filtered and removed using a feature selection algorithm. Simultaneously, the fusion strategy is optimized for the scenario characteristics of different transit stages, including: strengthening the fusion of long-term trends of environmental parameters and goods condition decay characteristics in the sea transport stage; highlighting the correlation fusion of instantaneous fluctuations of environmental parameters and goods shock resistance characteristics in the land transport stage; and focusing on the synergistic representation of air pressure changes and goods adaptability characteristics in the air transport stage. The final generated fused feature data is then presented.

[0022] S104: Based on the transit stage, status monitoring threshold, environmental control target and preset association rules, perform linkage analysis and comprehensive judgment on the fused feature data to obtain the status category and stage compliance assessment of the fresh goods to be transited.

[0023] In this embodiment, based on the transshipment stage, a set of typified features required for the current transshipment stage is extracted from the fused feature data. Based on this set of typified features, corresponding parameters from the fused feature data that correspond to the status monitoring threshold and environmental control targets of the current transshipment stage are extracted to obtain target parameters. Subsequently, the target parameters are compared with the status monitoring threshold and environmental control targets of the current transshipment stage to obtain comparison results. This generates the degree of matching and risk pattern between the fresh goods to be transshipped and the cold chain environmental parameter data in the current transshipment stage, resulting in a status category and stage compliance assessment. The status category of the fresh goods to be transshipped includes the classification and labeling results of the current quality and safety status of the goods, used to distinguish the safety level of the fresh goods to be transshipped. The stage compliance assessment is used to determine whether the transshipment operation in this stage meets the standards and can smoothly connect to the next stage.

[0024] S105: Based on the status category and stage compliance assessment, generate remote control signals that can be recognized by the remote monitoring terminal.

[0025] In this embodiment, based on the state category, the priority level and control mode of environmental control required for the current transfer stage are determined, and a multi-instruction coordinated control strategy is generated according to the degree of matching, risk mode and main causes in the stage compliance assessment and traceability report.

[0026] Conflict detection is performed on the pre-control strategy and the multi-command coordinated control strategy. If no conflict is detected, the two are merged into an enhanced remote control signal; if a conflict is detected, arbitration is performed based on the urgency of the state category, the characteristic strength of the risk mode, and the connection requirements of the next transfer stage to generate a remote control signal.

[0027] S106: The remote control signal is sent to the cold chain transfer unit corresponding to the current transfer stage through the remote monitoring terminal; in response to the remote control signal, each cold chain transfer unit starts local control action and feeds back the execution result to the remote monitoring terminal.

[0028] In this embodiment, the remote monitoring terminal acts as a hub for command relay, route distribution, and node matching. Upon receiving a remote control signal, it performs addressing and targeted transmission based on the cold chain transfer unit identifier corresponding to the current transfer stage. Specifically, the remote monitoring terminal first reads the stage code, equipment number, and task ID carried in the signal, and matches it with the cold chain carriers in operation at that stage in the logistics management platform, including corresponding hardware units such as refrigerated shipping containers, refrigerated land vehicles, air transport container equipment, and transit cold storage. Then, it transmits the standardized control signal to the local control module of the corresponding cold chain transfer unit through a wireless communication network, while simultaneously logging the transmission process, retaining the signal transmission time, target object, and original content.

[0029] After receiving a remote control signal matching its own node, the local control module of each cold chain transfer unit parses the remote control signal, verifies its permissions, and determines the legality of the command. Upon successful verification, it initiates the corresponding local control action based on the control parameters, execution mode, and priority carried in the remote control signal. Specifically, differentiated operations are performed based on the equipment characteristics and cargo requirements of different transfer stages: land transport units can adjust the refrigeration unit power, airflow speed, and internal airflow structure; sea transport units adjust the internal temperature and humidity, gas composition, and defogging / dehumidification mechanisms; air transport units coordinate with air pressure changes for rapid temperature stabilization compensation and sealing protection; and transit cold storage units perform pre-cooling, constant temperature maintenance, or gradient cooling.

[0030] As can be seen from the above, this application enhances the targeting of monitoring and control by collecting physical data of fresh goods and cold chain environmental parameters, and matching monitoring thresholds and control targets based on transfer task information, thus avoiding control deviations caused by single data or uniform standards. Furthermore, through feature extraction and fusion, the correlation analysis between cargo status and environmental parameters is achieved, and combined with comprehensive judgment based on preset correlation rules, the accuracy and timeliness of identifying cargo status categories and stage compliance can be improved. Simultaneously, control signals are generated based on remote monitoring terminals and sent to corresponding cold chain transfer units, realizing real-time control across different locations. In addition, receiving execution feedback breaks through the geographical limitations and passive monitoring mode of traditional cold chain transfer. This application achieves intelligent and visualized control of the entire stage of cold chain logistics transfer, improving the efficiency and reliability of cold chain logistics transfer.

[0031] In one embodiment of this application, based on the transit stage, status monitoring threshold, environmental control target, and preset association rules, the fused feature data is analyzed and comprehensively judged to obtain the status category and stage compliance assessment of fresh goods, including: Based on the transit stage, extract the set of typified features required for the current transit stage from the fused feature data; Based on the typified feature set, the corresponding parameters of the current transfer stage status monitoring threshold and environmental control target are extracted from the fused feature data to obtain the target parameters; The target parameters are compared with the current state monitoring thresholds and environmental control targets of the current transfer phase to obtain the comparison results; Based on the comparison results, the matching degree and risk pattern of fresh goods to be transferred and cold chain environment parameter data in the current transfer stage are generated, and the status category and stage compliance assessment are obtained.

[0032] In this embodiment, the current transit stage is used as the selection criterion to perform adaptive feature extraction on the fused feature data, resulting in a set of categorized features required for the current transit stage. The fused feature data includes related information such as the physical state of the goods, the cold chain environment, and equipment operation. Since the control priorities differ across transit stages, it is necessary to eliminate redundant features irrelevant to the current stage based on the specific characteristics of sea, land, and air transport scenarios, retaining only features that represent the safety and environmental compliance of the goods at the current stage.

[0033] Based on the categorized feature set, quantitative indicators corresponding to the current stage's state monitoring thresholds and environmental control targets are further extracted, forming target parameters for comparison. This process transforms structured features into physical quantities that can be compared with state monitoring thresholds and environmental control targets according to a pre-defined parameter mapping relationship. For example, the actual temperature of the current area is extracted from temperature distribution features, the rate of temperature change is extracted from fluctuation features, and oxygen content and carbon dioxide concentration are extracted from gas composition features. These target parameters correspond one-to-one with state monitoring thresholds and environmental control targets. The parameter mapping relationship is established based on the control priorities of different transshipment stages (sea, land, and air transport), as well as the corresponding state monitoring thresholds and environmental control targets for each stage, establishing corresponding association rules between structured features in the fused feature data and directly comparable physical quantities.

[0034] In this embodiment, the target parameters are analyzed item by item and compared with the status monitoring thresholds and environmental control target values ​​of the current transit stage according to the time series, to obtain the comparison results. The comparison content includes whether the parameter is within the upper and lower limits of the status monitoring threshold, the magnitude and duration of deviation from the threshold, and the difference from the environmental control target, etc. At the same time, normal deviation, warning deviation, and over-limit deviation are distinguished, forming a structured comparison result including numerical deviation, deviation duration, and time series distribution. Among them, different comparison rules are adopted for different scenarios such as long-cycle maritime transport, variable disturbances in land transport, and high time-constraint air transport. For example, maritime transport adds the judgment of cumulative over-limit duration, land transport strengthens the verification of instantaneous peak value, and air transport focuses on the judgment of recovery speed in a short period of time.

[0035] Based on the comparison results, the matching degree and risk pattern of fresh goods to be transferred and cold chain environment parameter data in the current transfer stage are generated, and the status category and stage compliance assessment are obtained.

[0036] As can be seen from the above, this embodiment extracts a typified feature set guided by the transfer stage, which can eliminate redundant data and effectively improve data processing efficiency and analysis targeting. By comparing the target parameters with the staged status monitoring thresholds and environmental control targets item by item, an intuitive structured result is effectively formed. At the same time, based on the comparison results, the matching degree is calculated, risk patterns are identified, and status categories and stage compliance assessments are output. This can achieve efficient transformation from multi-source fusion data to standardized control conclusions, which not only improves the accuracy and intelligence level of judging the status and environmental compliance of fresh goods to be transferred, but also provides a reliable and concrete decision-making basis for subsequent remote control.

[0037] In one embodiment of this application, based on the comparison results, the degree of matching and risk pattern between the fresh goods to be transferred and the cold chain environment parameter data at this stage are generated, and the status category and stage compliance assessment are obtained, including: Based on the time series data of the comparison results, a correlation network between environmental parameters and cargo status parameters is established. Calculate the information transmission strength from environmental parameters to cargo status in the correlation network, and generate a causal influence spectrum; The degree of matching is obtained by calculating the matching degree between the causal impact spectrum and the ideal causal model of the current transit stage; Based on the causal influence spectrum, a pre-trained anomaly detection model is used to identify whether there are pre-defined abnormal correlation structural features. When abnormal correlation structure features are identified, the corresponding risk pattern is identified based on a preset pattern library; Based on the degree of matching, the characteristic strength of the risk pattern, and the absolute deviation level of the key cargo status parameters in the comparison results, the overall condition of the fresh goods to be transferred is evaluated, and the evaluation results are obtained. Based on the assessment results, the status category is determined, and a phase compliance assessment is generated based on the degree of matching.

[0038] In this embodiment, based on the continuous time series data formed by the comparison results, a correlation network between environmental parameters and cargo status parameters is constructed. This correlation network uses various environmental parameters (temperature, humidity, gas concentration, air pressure, vibration) and cargo status parameters (freshness index, packaging integrity, deterioration rate) as nodes, and uses the correlation and lag response relationship of different parameters in time series as connecting edges to comprehensively depict the dynamic correlation between environmental changes and cargo status changes.

[0039] Based on the interconnected network, the information transmission strength from environmental parameter nodes to cargo status parameter nodes is calculated. Using the quantification method of transmission entropy, the contribution weight and transmission delay of each type of environmental change corresponding to the corresponding state change are determined, thereby generating a causal impact spectrum. The contribution weight is the value of transmission entropy; the larger the value, the stronger the influence of X on Y. The transmission delay is the l-value (historical order of X) that maximizes the transmission entropy, representing the time delay from environmental parameter change to cargo status change. This causal impact spectrum not only identifies the dominant environmental factors but also includes the strength of each factor's influence, its effect delay, and coupling relationship, allowing for the identification of the core environmental triggers causing cargo status changes. Subsequently, this causal impact spectrum is matched dimension-by-dimensionally with a pre-defined ideal causal model for the current transshipment stage to obtain a matching degree value. The matching degree represents the degree of fit between the actual operating state and the current stage's standard compliance state. The formula for calculating the information transmission strength from environmental parameter (X) to cargo status (Y) is as follows: The transfer entropy from environmental parameter X to cargo state Y is used to quantify the information intensity transferred from X to Y, representing the driving effect of X on Y. Let Y be the k-th order historical state of cargo at time n, that is, the set of states of Y over k consecutive times before time n. , ,..., ); Let X be the l-th order historical state of the environmental parameter X at time n, that is, the set of states of X for l consecutive times before time n. , ,..., ); Let p( be the next state of cargo Y at time n+1); ) is a probability function, which can be obtained through frequency statistics of time series data; Given the k-th order history of Y and the l-th order history of X, let Y take the next time step. The conditional probability; Given the k-th order history of Y, what is the next time step for Y? The conditional probability.

[0040] A pre-trained anomaly detection model is used to perform a global scan of the causal impact spectrum to identify any abnormal correlation structures that deviate from the ideal pattern. These abnormal correlation structures include anomalously enhanced transmission paths, abnormal parameter coupling relationships, and abrupt information transmission peaks. When the anomaly detection model determines the presence of such features, template matching is performed between these features and a pre-defined risk pattern library. Based on the feature morphology, impact intensity, involved parameter types, and temporal distribution, a corresponding risk pattern is matched. This pre-defined risk pattern library is a structured template library built based on anomaly scenarios across multiple cold chain transport stages. Examples of risk types strongly correlated with the transport stage include continuous overheating leading to deterioration, localized temperature unevenness causing state differentiation, vibration-induced temperature changes accelerating losses, and sudden pressure changes causing packaging failure. The feature intensity and impact range of the corresponding risk pattern are then output.

[0041] The hierarchical architecture of the anomaly detection model includes a temporal feature encoding layer, an association structure modeling layer, and an anomaly discrimination layer. The temporal feature encoding layer is used to extract the temporal evolution features of information transmission intensity and transmission delay in the causal influence spectrum, and to capture abrupt peaks and trend shifts. The association structure modeling layer is used to model the association network structure of environment-cargo parameters and identify abnormal enhancement paths and abnormal coupling relationships. The anomaly discrimination layer is used to fuse temporal features and structural features to output the anomaly probability and anomaly type.

[0042] The temporal feature encoding layer consists of a one-dimensional CNN and an LSTM. The one-dimensional CNN has a kernel size of 3-5, a kernel count of 32-64, a stride of 1, and uses ReLU as the activation function, combined with max pooling to enhance peak features. The LSTM has hidden layer dimensions of 64-128, a time step of 10-30 for a 10-30 minute temporal window, and a dropout coefficient of 0.2-0.3. The association structure modeling layer includes a graph convolutional network, which transforms the association network of the causal influence spectrum into an adjacency matrix. Node features include parameter type and baseline propagation entropy, and edge features... To optimize information transmission strength and latency, the graph convolutional network is configured with 2-3 layers and a hidden layer dimension of 64. The LeakyReLU activation function is selected, and the adjacency matrix is ​​normalized. The anomaly detection layer consists of two fully connected layers with the number of neurons ranging from 128 to the number of risk pattern categories. The FocalLoss loss function is used with weight coefficients α=0.25 and γ=2, outputting the probability values ​​of each anomaly type. The anomaly detection threshold is set to 0.7 by default and can be adjusted according to the transportation stage, such as lowering it to 0.6 to improve sensitivity during air transport.

[0043] Based on the degree of matching, the characteristic strength of the risk pattern, and the absolute deviation level of key cargo status parameters in the comparison results, the overall status of the fresh goods to be transshipped is assessed, and the assessment results are obtained. The assessment results are used as the status category. At the same time, a stage compliance assessment for the current transshipment stage is generated based on the degree of matching between the causal influence spectrum and the ideal model. For example, the causal influence spectrum is matched with the ideal causal model of the current transshipment stage dimension by dimension. By quantifying the fit between the actual correlation structure and the benchmark structure in terms of driving factor strength and transmission delay deviation, the degree of matching is obtained. Based on the local deviation points identified in the matching process, a stage compliance assessment report is generated, including the overall fit level, key deviation dimensions, and compliance risk warnings.

[0044] As can be seen from the above, this embodiment, by constructing a correlation network between environmental parameters and cargo status parameters based on time-series data, quantifying the intensity of information transmission, and generating a causal impact spectrum, overcomes the limitations of traditional correlation analysis. It improves the accuracy of locating the true impact path and intensity of environmental changes on cargo status from the data correlation level to the causal transmission level. Furthermore, by calculating the matching degree between the causal impact spectrum and the ideal causal patterns at each transit stage, the degree of conformity between the current operating conditions and standard control is obtained, improving the objectivity and accuracy of stage compliance determination. Simultaneously, by using a pre-trained anomaly detection model to identify anomaly correlation structure features and matching risk patterns according to a preset pattern library, implicit and coupled anomaly patterns can be discovered in advance, achieving earlier and more accurate risk identification.

[0045] In one embodiment of this application, the overall condition of the fresh goods to be transshipped is evaluated based on the degree of matching, the characteristic strength of the risk pattern, and the absolute deviation level of key cargo status parameters in the comparison results, to obtain the evaluation results, including: The deviation of the matching degree from the preset benchmark matching degree is calculated to obtain the first evaluation component; Extract the feature intensity of the risk pattern to obtain the second evaluation component; The absolute deviation levels of the cargo status parameters in the comparison results are normalized and weighted to obtain the third evaluation component. Based on the current evaluation weight configuration of the current transit stage, the first evaluation component, the second evaluation component, and the third evaluation component are weighted and integrated to generate a comprehensive evaluation score. The comprehensive evaluation score is input into the preset evaluation decision model to obtain the evaluation result.

[0046] In this embodiment, firstly, a first evaluation component is obtained by calculating the difference between the matching degree and a preset benchmark matching degree. This first evaluation component is used to quantitatively characterize the degree of deviation between the actual environment-cargo causal transmission relationship and the standard ideal model. A smaller deviation value indicates that the overall operational status is more in line with regulatory control requirements, while a larger deviation value indicates an overall compliance deviation. The benchmark matching degree is a matching degree threshold used to characterize the ideal state of the environment-cargo causal relationship, based on the scenario characteristics of different transit stages, the causal influence spectrum characteristics of historical compliant transit data, and calibrated through statistical analysis and expert experience.

[0047] Secondly, the characteristic intensity values ​​corresponding to the risk pattern in the causal impact spectrum are extracted as the second evaluation component. The characteristic intensity is determined by the amplitude, duration, and number of affected parameters of the abnormal correlation structure, representing the severity level of the abnormal problem. The absolute deviation levels of the target parameters related to the cargo status from the status monitoring thresholds are normalized, and weighted coefficients are set according to the cargo category and preservation sensitivity level before summing to obtain the third evaluation component.

[0048] After calculating the three assessment components, a weighted fusion is performed based on the characteristics of the transshipment stage to generate a comprehensive assessment score that is stage-adaptive. For example, based on the different control focuses of sea, land, and air transport scenarios, the corresponding stage-specific assessment weight configurations are invoked, assigning differentiated weights to the first, second, and third assessment components: long-cycle sea transport increases the weight of the first assessment component, which represents long-term trends; land transport, with its frequent fluctuations, strengthens the weight of the second assessment component, which reflects the degree of sudden risk; and air transport, with its high timeliness and high sensitivity, increases the weight of the third assessment component, which reflects the real-time status of the goods. The three independent indicators are then integrated into a single continuous value through weighted summation to obtain the comprehensive assessment score.

[0049] Finally, the comprehensive evaluation score is input into the preset evaluation decision model to obtain the evaluation result.

[0050] The assessment and decision-making model is an intelligent judgment model that outputs a conclusion on the conformity of cargo status categories and stages based on the comprehensive assessment score of fresh goods. Its function is to establish a mapping relationship between quantitative scores and qualitative assessment conclusions. The model is constructed with the control requirements of different transshipment stages (sea, land, and air transport) as its guide, and is trained based on historical transshipment data. The inputs are the comprehensive assessment scores of the deviation value of the matching degree, the strength of risk pattern characteristics, and the absolute deviation level of cargo status parameters. The output is the assessment conclusion, including status categories such as normal, minor warning, moderate abnormality, and severe risk. The model adopts a hybrid architecture of staged threshold judgment and machine learning classification. On the one hand, the model has built-in comprehensive assessment score threshold ranges for different transshipment stages; on the other hand, the model learns the correlation between scores and assessment conclusions in historical data through a gradient boosting tree algorithm, achieving accurate judgment of fuzzy score ranges. Simultaneously, the model supports dynamic iterative optimization, allowing new transshipment case data to be included in the training set, continuously adjusting the judgment thresholds and classification weights to ensure the accuracy and applicability of assessments under different cargo categories and scenarios. The number of decision trees in the gradient boosting tree is set to 100-200, and the depth of a single tree is controlled at 3-5 layers to avoid overfitting. The learning rate is set to 0.1-0.2, and the log loss function is selected to adapt to the classification task.

[0051] As can be seen from the above, this embodiment obtains a comprehensive evaluation score by weighted summation of the deviation value of the matching degree, the intensity of risk pattern characteristics, and the absolute deviation level of parameters, followed by normalization, thus avoiding the one-sidedness of a single indicator evaluation. Furthermore, the weighted fusion based on the evaluation weights configured at the current transshipment stage can adapt to the different control focuses of sea, land, and air transport scenarios, further improving the scenario-specific relevance and rationality of the comprehensive evaluation score. Finally, the comprehensive evaluation score is input into the evaluation decision model to output the evaluation result, ensuring both the standardization and reproducibility of the evaluation process and improving the efficiency and accuracy of the judgment.

[0052] In one embodiment of this application, a remotely monitored intelligent cold chain logistics transfer method further includes: Adjustable parameters of the quality indicator assessment decision model based on currently collected cold chain environmental parameter data include: Calculate the signal-to-noise ratio and fluctuation amplitude of cold chain environmental parameter data within a preset time window; The comparison is based on the signal-to-noise ratio and the preset quality threshold, and the fluctuation amplitude and the preset stability threshold. When the signal-to-noise ratio is less than the quality threshold, or the fluctuation amplitude is greater than the stability threshold, the environmental parameter data quality is determined to have deteriorated, and the parameter adjustment mechanism is executed.

[0053] In this embodiment, the adjustable parameters of the evaluation decision model are adjusted based on the quality indicators of the currently collected cold chain environmental parameter data. Specifically, firstly, within a preset time window, the cold chain environmental parameter data is statistically processed, and the signal-to-noise ratio (SNR) and fluctuation amplitude of the corresponding time-series data are calculated. The SNR represents the clarity of the data acquisition and the level of anti-interference, while the fluctuation amplitude is obtained by calculating the range, variance, or variance of the parameters within the window, used to characterize the stationarity and abrupt changes of the data sequence. The preset time window is a fixed duration interval defined by historical data statistical analysis and business requirements, based on the environmental fluctuation characteristics of different cold chain transport stages, data sampling frequency, and anomaly response time requirements. The SNR is calculated based on the ratio of the signal mean to the noise standard deviation of the cold chain environmental parameter data within the preset time window. The signal mean is the arithmetic mean of the parameters within the window, and the noise is the difference between the original parameter value and the signal mean. The fluctuation amplitude is calculated by dividing the sum of the squares of the differences between each data point of the cold chain environmental parameter and the mean of the parameter within the preset time window by the total number of data points.

[0054] After calculating the signal-to-noise ratio (SNR) and fluctuation amplitude, each value is compared with preset quality and stability thresholds. The quality threshold is set differently based on various cold chain environmental parameter types, sensor accuracy levels, and characteristics of the transit stage, defining the minimum SNR requirement for valid data. The stability threshold is set based on the normal disturbance range for different scenarios (sea, land, and air transport), distinguishing between physiologically reasonable fluctuations and abnormally drastic jumps. Specifically, the quality threshold is determined by testing the sensor's minimum effective SNR under different operating conditions, based on different cold chain environmental parameter types, sensor accuracy levels, and transit stage characteristics. The stability threshold is based on the normal environmental disturbance range of different transit stages, statistically analyzing the distribution range of the range, variance, or variation of environmental parameters in historical compliant transit data, and defining critical values ​​according to anomaly response time requirements.

[0055] Specifically, when either the signal-to-noise ratio is less than the quality threshold or the fluctuation amplitude is greater than the stability threshold, it is determined that the cold chain environment parameter data for the current period has quality degradation problems such as excessive noise, severe shaking, and distortion. It is no longer suitable to use the parameters of the original evaluation decision model for evaluation and calculation, and a parameter adjustment mechanism is executed.

[0056] As can be seen from the above, this embodiment calculates the signal-to-noise ratio and fluctuation amplitude of cold chain environmental parameter data based on a preset time window, and realizes the quantitative judgment of data quality according to preset quality thresholds and stability thresholds. This can improve the accuracy of identifying data quality degradation problems caused by sensor interference, drift, transmission jitter or operating condition disturbances, and avoid the lag and subjectivity caused by manual judgment. Furthermore, the parallel comparison of signal-to-noise ratio and fluctuation amplitude improves the comprehensiveness and accuracy of quality judgment.

[0057] In one embodiment of this application, when the signal-to-noise ratio is less than a quality threshold, or the fluctuation amplitude is greater than a stability threshold, it is determined that the environmental parameter data quality has deteriorated, and a parameter adjustment mechanism is executed, including: Based on the cause category of the decline in environmental parameter data quality, select the corresponding adjustment strategy. The cause categories include quality decline caused by signal-to-noise ratio and quality decline caused by fluctuation amplitude. When the cause category is signal-to-noise ratio-induced quality degradation, a preset first step length is used to increase the smoothing intensity of the fused feature data, and a preset second step length is used to reduce the weight of the absolute deviation level of the cargo state parameters in the third evaluation component. When the cause category is quality degradation due to fluctuation amplitude, the sliding analysis window of the time series data in the comparison results is extended by the preset third step length, and the anti-interference correction term of the causal influence spectrum analysis is added accordingly. Based on the selected adjustment strategy, update the adjustable parameters of the data preprocessing module or weighted fusion module of the evaluation decision model.

[0058] In this embodiment, the triggering sources of environmental parameter data quality degradation are classified, resulting in causal categories including quality degradation due to insufficient signal-to-noise ratio and quality degradation due to excessive fluctuation amplitude. Differentiated model adjustment strategies are selected for the underlying data problem corresponding to different causal factors.

[0059] Specifically, when the cause of quality degradation is determined to be insufficient signal-to-noise ratio (SNR), it indicates that there is significant noise interference in the current environmental parameters, and the effective signal is submerged by clutter. In this case, a dedicated adjustment strategy for the corresponding SNR degradation is activated. For example, the smoothing processing intensity of the fused feature data is increased by a preset first step length, and random noise is suppressed through weighted smoothing, temporal filtering, and noise reduction fitting to enhance the expression of effective features. At the same time, the weight of the absolute deviation level of the cargo state parameters in the third evaluation component is reduced by a preset second step length to reduce the interference of noise data on the calculation of cargo degradation amplitude. When the cause of quality degradation is determined to be excessive fluctuation amplitude, it indicates that there are short-term drastic changes, abnormal peaks, or non-physiological disturbances in the environmental parameters, and the temporal stability of the data is disrupted. In this case, the adjustment strategy for the corresponding fluctuation amplitude is switched. For example, by pre-setting a third step to extend the sliding analysis window of the time series data on which the comparison results depend, the statistical interval is expanded to smooth out local distortions caused by instantaneous disturbances and improve the global stability of time series analysis. At the same time, corresponding anti-interference correction terms are added in the causal influence spectrum calculation stage. For instance, based on the time series distribution characteristics of environmental parameters within the extended sliding analysis window, the mean, variance, and abnormal peak amplitude of the parameters within the window are first extracted. Then, the multiple of the fluctuation amplitude exceeding the stability threshold and the duration of the disturbance are used as correction factors and weighted multiplied with the original calculation results of the transmission entropy. At the same time, extreme outliers within the window that deviate from the mean by more than three standard deviations are removed to eliminate interference with information transmission strength and transmission delay. The first step is the adjustment increment of the smoothing coefficient of the fused feature data, with the preset logic being a linear mapping value of the signal-to-noise ratio degradation degree; the second step is the reduction of the weight of the absolute deviation of the cargo status parameter from the level, with the preset logic being the product of the signal-to-noise ratio degradation degree and the cargo category sensitivity coefficient; the third step is the extension increment of the sliding analysis window of the time series data, with the preset logic being a segmented threshold mapping value of the fluctuation amplitude exceeding the standard multiple.

[0060] After selecting an adjustment strategy that matches the causal category, the adjustable parameters of the corresponding functional modules in the evaluation decision model are updated in a targeted manner according to the strategy content. For signal-to-noise ratio degradation, the smoothing coefficient and filtering strength of the data preprocessing module are mainly corrected. For fluctuation amplitude degradation, the sliding window length of the time series analysis module, the correction coefficient of the causal calculation module, and the component weight configuration of the weighted fusion module are mainly updated.

[0061] As can be seen from the above, this embodiment avoids the problems of redundancy or insufficient effect caused by general adjustment by matching the corresponding evaluation decision model adjustment strategy based on two different cause categories: insufficient signal-to-noise ratio and excessive fluctuation amplitude. The adjustable parameters of the corresponding module of the model are updated in a targeted manner based on the selected adjustment strategy, which not only improves the anti-interference ability, robustness and judgment accuracy of the evaluation decision model when the quality deteriorates, but also ensures the pertinence, controllability and reproducibility of the evaluation decision model correction.

[0062] In one embodiment of this application, a remotely monitored intelligent cold chain logistics transfer method further includes: The status category, stage compliance assessment, risk model, comprehensive assessment score and related timestamp information are structured and encapsulated to generate a status snapshot data package of fresh goods to be transferred at the current transfer stage. Based on state snapshot data packets and historical state snapshot data packets, a full-link state evolution map of fresh goods to be transferred is constructed along the time dimension. The entire link state evolution map is analyzed to identify key nodes and related environmental events of state deterioration, and the main causes of state deterioration are traced based on the transmission path of risk patterns to obtain the tracing results. Based on the traceability results, a traceability report is generated and pushed to the remote monitoring terminal for display, and a pre-control strategy is generated.

[0063] In this embodiment, information such as status category, stage compliance assessment, risk mode, comprehensive assessment score and corresponding timestamp, transit stage identifier, and cargo category number are first processed according to a preset communication and storage protocol through field arrangement, format verification, and encoding encapsulation to obtain a status snapshot data packet of the fresh goods to be transited at the current transit stage. Each status snapshot data packet serves as a complete and indivisible digital slice of the fresh goods' status at that moment, including both classification results and assessment results along with associated contextual information. The preset communication and storage protocol is determined based on the transmission bandwidth, data security, terminal compatibility, and platform storage specifications of cold chain remote monitoring.

[0064] Using the current single-frame state snapshot data packet as the real-time node, and based on the historical state snapshot data packets generated by the same batch of fresh goods in this stage and previous transfer stages, the data packets are continuously spliced ​​and aligned along the time axis to construct a full-link state evolution map covering the entire transfer process. This full-link state evolution map uses time as the core main line, state snapshots as nodes, and parameter changes, risk transmission, and environmental events as related edges, presenting the continuous state change trajectory of fresh goods from origination and transfer to the current node.

[0065] Based on the end-to-end state evolution map, a time-series in-depth analysis is conducted. Through trend fitting, the occurrence nodes, degradation rates, and degradation degrees of state deterioration are located, and corresponding environmental parameter fluctuations, equipment actions, and transfer scenario switching before and after the node are matched synchronously. According to the transmission path and diffusion characteristics of the risk model presented in the evolution map, the source causes are traced back from the degradation results. It is distinguished whether the cause is due to initial environmental deviation, staged control lag, local equipment anomaly, external disturbance intrusion, or transmission cumulative effect. Finally, a structured tracing result is formed, including degradation nodes, triggering events, transmission chains, core causes, and degree of impact.

[0066] Based on the traceability results, a traceability report is generated, which includes a description of the deterioration, the occurrence node of the deterioration in the closed state, the traceability path, the cause analysis, and the impact assessment. The report is then pushed to the remote monitoring terminal through the communication link and visualized in the form of charts, curves, and text descriptions.

[0067] As can be seen from the above, this embodiment achieves standardized and lightweight storage and transmission of fresh goods status information by generating status snapshot data packets; based on the constructed end-to-end status evolution map, discrete time-series data can be integrated into a continuous time-series trajectory, completely restoring the status change process of the entire cargo transfer cycle; by locating key deterioration nodes through map analysis, matching environmental events, and tracing the root cause along the risk transmission path, the efficiency and accuracy of problem investigation can be improved; finally, the generated traceability report is pushed and displayed, and pre-control strategies are output, improving the transparency, traceability capability, and proactive prevention and control level of the entire cold chain process supervision.

[0068] In one embodiment of this application, a remote control signal recognizable by the remote monitoring terminal is generated based on the state category and stage compliance assessment, including: Based on the status category, determine the priority level and control mode of environmental regulation required for the current transfer phase; Based on the degree of matching, risk patterns, and main causes in the phased compliance assessment and retrospective report, a multi-instruction coordinated control strategy is generated. Conflict detection is performed on pre-control strategies and multi-instruction coordinated control strategies. If no conflict is detected, the two signals will be merged into an enhanced remote control signal; If a conflict is detected between the two, arbitration is conducted based on the urgency of the status category, the characteristic strength of the risk mode, and the connection requirements of the next transfer stage to generate the final remote control signal.

[0069] In this embodiment, the priority level and control mode of environmental control are determined according to the status category of the current transit stage, which is suitable for the current safety status of the goods. The status category is a direct representation of the overall risk level of the goods, which determines the control response speed, execution intensity and resource allocation order. Different levels correspond to differentiated priority levels from routine maintenance, early warning fine-tuning, rapid correction to emergency handling, and are matched with corresponding control modes such as constant temperature stability, gradient compensation, amplitude limit correction and mandatory protection.

[0070] Based on the established regulatory tone, a multi-instruction coordinated regulation strategy is generated according to the matching degree and risk patterns in the phased compliance assessment, as well as the main causes of deterioration in the traceability report. This multi-instruction coordinated regulation strategy generates a combination of time-coordinated and parameter-linked instructions for multiple regulation objects such as temperature, humidity, gas composition, ventilation, and cooling power. It not only provides compensatory regulation for compliance gaps in the matching degree, but also blocks the transmission path of risk patterns, and conducts source-corrective regulation based on the core causes in the traceability report.

[0071] Further conflict detection is performed on the pre-control strategy and the multi-instruction coordinated control strategy. The consistency of the two strategies in terms of control objectives, parameter settings, execution sequence, and execution intensity is compared to determine whether there are situations such as opposite objectives, mutually exclusive parameters, contradictory timing, or conflicting execution objects.

[0072] If no conflict is detected, the preventive instructions of the pre-control strategy and the real-time correction instructions of the coordinated control strategy will be integrated and time-sequenced to form an enhanced remote control signal that includes real-time correction and forward protection. If a conflict is detected, a multi-level arbitration mechanism will be activated, which will be weighted based on three core indicators: the urgency of the status category, the characteristic strength of the risk mode, and the working condition connection requirements of the next transfer stage, and finally generate a remote control signal.

[0073] From the above, it can be concluded that this embodiment determines the appropriate environmental control priority and control mode based on the state category, realizing the rational allocation of control resources and the precise matching of response levels; and by combining the degree of matching, risk mode and deterioration causes to generate a multi-instruction collaborative control strategy, it can carry out multi-parameter linkage correction from the root cause of the anomaly, improving the targeting and eradication effect of the control; at the same time, by performing conflict detection on the pre-control strategy and the collaborative control strategy, it not only ensures the timeliness and safety of anomaly handling, but also achieves a smooth connection across the transfer stage.

[0074] In one embodiment of this application, a remotely monitored intelligent cold chain logistics transfer method further includes: The transshipment phase includes sea, land, and air transport; Before switching from land transport to air transport, obtain flight information for the fresh goods to be transferred, including pre-allocated parking spaces and estimated ground transfer operation time. Based on the geographical location of the parking position, the current time, and real-time meteorological data, predict the temperature change trend of fresh goods to be transferred when exposed to the apron environment during the ground transfer operation. Based on the temperature change trend, the current status category of the fresh goods to be transferred, and the transportation requirements of the next transfer stage, the target temperature and cooling rate required by the ground cold chain vehicle or freight station cold storage for receiving the goods are calculated. Before and after the aircraft lands, connection control commands, including the target temperature and cooling rate, will be sent in advance as connection signals to the designated ground connection unit. After receiving the connection control command, the ground connection unit adjusts its own equipment operating status.

[0075] In this embodiment, the transshipment stage includes three scenarios: sea transport, land transport, and air transport.

[0076] Specifically, before the land-to-air transfer is completed, the flight information corresponding to the fresh goods to be transferred is obtained through the air logistics interface and ground dispatch platform. The flight information includes not only the flight number and take-off and landing time, but also the coordinates of the pre-assigned parking position, the ground transfer route, and the estimated total ground transfer operation time. Based on the geographical location information such as the latitude and longitude of the pre-assigned parking position, the current environmental baseline value, and the meteorological data such as apron temperature, solar radiation intensity, wind speed, and precipitation transmitted in real time from the meteorological platform, and by integrating the thermal conductivity characteristics of the fresh goods packaging, the exposed area, and the ground transfer operation time, the temperature change trend of the fresh goods in the open apron environment without closed cold chain protection is predicted, and the total temperature rise of the fresh goods during the apron exposure period is predicted.

[0077] Based on temperature change trends, the current status of the cargo, and the temperature control requirements within the aircraft cabin during air transport, this study uses coupled calculations with multiple constraints to determine the temperature control indicators required for ground-based refrigerated vehicles or cold storage facilities used for cargo transfer. These indicators include the target temperature to be met immediately upon receiving the cargo and the cooling rate required to offset the temperature rise from ground exposure. The formula for calculating the target temperature is as follows: The ideal temperature after compensation for temperature rise due to apron exposure and correction for sensitivity coefficient; These are the standard temperature control settings specified for the air cargo hold phase, i.e., the temperature control requirements for transport within the cabin during the air cargo phase. To predict the total temperature rise of fresh goods during the tarmac exposure period: It is the sensitivity coefficient to deterioration of goods (a dimensionless correction coefficient determined by the current state category and the category of goods).

[0078] The formula for calculating the cooling rate is: The base cooling rate is theoretically the rate required to offset the temperature rise caused by exposure to the apron. This represents the estimated time for ground transfer operations.

[0079] During the scheduling window before aircraft landing, parameters such as target temperature and cooling rate are encapsulated into standardized connection control commands and sent in advance to the designated ground cold chain connection unit via a dedicated communication link in the form of connection signals. The connection signals include the corresponding cold chain vehicle on-board controller and the cargo station cold storage central control system, allowing sufficient time for equipment adjustment.

[0080] In this embodiment, after receiving the connection control command, the ground connection unit analyzes the target temperature and cooling rate configuration, and starts the refrigeration unit in advance, adjusts the air supply mode and cooling power according to the connection control command, and gradually adjusts the internal environmental parameters of the carriage or cold storage to the target range.

[0081] As can be seen from the above, this embodiment obtains flight information, including parking positions and ground transfer time, before land-to-air transport, and predicts the temperature change trend of cargo apron exposure based on geographical location and real-time meteorological data. This allows for the quantification of temperature change risks in the connection process, enabling advance perception of temperature control risks. By sending connection control commands in advance, ground refrigerated vehicles and cargo terminal cold storage can adjust their equipment status in advance, ensuring the continuous and stable cold chain throughout the land-to-air transfer process. At the same time, it improves the closed-loop management capability of the entire cold chain process and the preservation effect of fresh goods.

[0082] Corresponding to the remotely monitorable intelligent cold chain logistics transfer method in the above embodiments, Figure 2 This is a structural block diagram of a remotely monitorable intelligent cold chain logistics transfer platform provided in one embodiment of this application. For ease of explanation, only the parts relevant to the embodiment of this application are shown. References Figure 2 The remotely monitored intelligent cold chain logistics transfer platform 20 includes: an information collection module 21, a strategy generation module 22, a data fusion module 23, an intelligent analysis module 24, an instruction generation module 25, and a linkage execution module 26.

[0083] Among them, the information collection module 21 is used to collect multi-dimensional physical data and cold chain environment parameter data of the fresh goods to be transferred, and to receive transfer task information from the logistics management platform. The transfer task information includes the attributes of the fresh goods to be transferred, transportation requirements, and transfer stage. The strategy generation module 22 is used to determine the appropriate status monitoring threshold and environmental control target based on the transfer stage, the attributes of the fresh goods to be transferred, and the transportation requirements. Data fusion module 23 is used to extract and fuse features from multi-dimensional physical data and cold chain environmental parameter data to obtain fused feature data; The intelligent analysis module 24 is used to perform linked analysis and comprehensive judgment on the fused feature data based on the transfer stage, status monitoring threshold, environmental control target and preset association rules, so as to obtain the status category and stage compliance assessment of the fresh goods to be transferred. The instruction generation module 25 is used to generate remote control signals that can be recognized by the remote monitoring terminal based on the status category and stage compliance assessment. The linkage execution module 26 is used to send remote control signals to the cold chain transfer unit corresponding to the current transfer stage through the remote monitoring terminal; in response to the remote control signal, each cold chain transfer unit starts local control action and feeds back the execution result to the remote monitoring terminal.

[0084] See Figure 3 , Figure 3 This is a schematic block diagram of an electronic device provided according to an embodiment of this application. Figure 3 The electronic device 300 in this embodiment may include one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memories 304 store computer programs, including program instructions. The processors 301 execute the program instructions stored in the memories 304. Specifically, the processors 301 are configured to invoke the program instructions to perform the functions of the modules in the aforementioned device embodiments, for example... Figure 2 The functions of the information acquisition module 21, strategy generation module 22, data fusion module 23, intelligent analysis module 24, instruction generation module 25, and linkage execution module 26 are shown.

[0085] It should be understood that, in the embodiments of this application, the processor 301 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0086] Input device 302 may include a touchpad, a fingerprint sensor (for collecting the user's fingerprint information and fingerprint orientation information), a microphone, etc., and output device 303 may include a display (LCD, etc.), a speaker, etc.

[0087] The memory 304 may include read-only memory and random access memory, and provides instructions and data to the processor 301. A portion of the memory 304 may also include non-volatile random access memory. For example, the memory 304 may also store device type information.

[0088] In specific implementations, the processor 301, input device 302, and output device 303 described in the embodiments of this application can execute the implementation methods described in any embodiment of the remotely monitored intelligent cold chain logistics transfer method provided in the embodiments of this application, or they can execute the implementation methods of the electronic devices described in the embodiments of this application, which will not be repeated here.

[0089] In another embodiment of this application, a computer-readable storage medium is provided. This computer-readable storage medium stores a computer program, which includes program instructions. When executed by a processor, the program instructions implement all or part of the processes in the methods described above. Alternatively, the computer program can instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include any entity or device capable of carrying computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0090] The computer-readable storage medium can be an internal storage unit of the electronic device in any of the foregoing embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the electronic device. Furthermore, the computer-readable storage medium can include both internal and external storage units of the electronic device. The computer-readable storage medium is used to store computer programs and other programs and data required by the electronic device. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0091] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0092] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the electronic devices and units described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0093] In the several embodiments provided in this application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces or units, or it may be an electrical, mechanical, or other form of connection.

[0094] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments of this application, depending on actual needs.

[0095] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0096] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A remotely monitored intelligent cold chain logistics transfer method, characterized in that, include: Collect multi-dimensional physical data and cold chain environment parameter data of fresh goods to be transferred, and receive transfer task information from the logistics management platform. The transfer task information includes the attributes, transportation requirements, and transfer stage of the fresh goods to be transferred. Based on the transshipment stage, the attributes of the fresh goods to be transshipped, and the transportation requirements, determine the appropriate status monitoring threshold and environmental control targets; Feature extraction and fusion are performed on the multi-dimensional physical data and the cold chain environment parameter data to obtain fused feature data; Based on the transshipment stage, the status monitoring threshold, the environmental control target, and preset association rules, the fused feature data is subjected to linkage analysis and comprehensive judgment to obtain the status category and stage compliance assessment of the fresh goods to be transshipped; wherein, the linkage analysis and comprehensive judgment includes: using a preset assessment decision model to assess the overall condition of the fresh goods to be transshipped, and determining the status category and stage compliance assessment based on the assessment results; and, during the assessment process using the assessment decision model, adjusting the adjustable parameters of the assessment decision model based on the quality indicators of the currently collected cold chain environmental parameter data, including: Calculate the signal-to-noise ratio (SNR) and fluctuation amplitude of cold chain environmental parameter data within a preset time window; compare the SNR with a preset quality threshold and the fluctuation amplitude with a preset stability threshold; when the SNR is less than the quality threshold or the fluctuation amplitude is greater than the stability threshold, determine that the environmental parameter data quality has deteriorated, and execute a parameter adjustment mechanism, including: Based on the system cause category, the smoothing processing intensity, weight allocation, sliding analysis window, and anti-interference correction term are adaptively adjusted, and the adjustable parameters of the data preprocessing module or weighted fusion module of the evaluation decision model are updated synchronously. Based on the state category and the stage compliance assessment, a remote control signal that can be recognized by the remote monitoring terminal is generated. The remote control signal is sent to the cold chain transfer unit corresponding to the current transfer stage through the remote monitoring terminal; in response to the remote control signal, each cold chain transfer unit initiates local control actions and feeds back the execution results to the remote monitoring terminal.

2. The intelligent cold chain logistics transfer method with remote monitoring according to claim 1, characterized in that, The process of performing a linked analysis and comprehensive judgment on the fused feature data based on the transshipment stage, the status monitoring threshold, the environmental control target, and preset association rules to obtain the status category and stage compliance assessment of fresh goods includes: Based on the aforementioned transfer stage, extract the set of typified features required for the current transfer stage from the fused feature data; Based on the typified feature set, the corresponding parameters of the current transfer stage status monitoring threshold and environmental control target are extracted from the fused feature data to obtain the target parameters; The target parameters are compared with the current state monitoring threshold and environmental control target of the current transfer stage to obtain the comparison results; Based on the comparison results, the matching degree and risk pattern of the fresh goods to be transferred and the cold chain environment parameter data in the current transfer stage are generated, and the status category and the stage compliance assessment are obtained.

3. The intelligent cold chain logistics transfer method with remote monitoring according to claim 2, characterized in that, Based on the comparison results, the matching degree and risk pattern of the fresh goods to be transferred and the cold chain environment parameter data at this stage are generated, and the status category and the stage compliance assessment are obtained, including: Based on the time series data of the comparison results, a correlation network between environmental parameters and cargo status parameters is established. Calculate the information transmission strength from environmental parameters to cargo status in the association network, and generate a causal influence spectrum; The degree of matching is obtained by calculating the matching degree between the causal influence spectrum and the ideal causal pattern of the current transit stage; Based on the causal influence spectrum, a pre-trained anomaly detection model is used to identify whether there are preset abnormal correlation structural features. When the abnormal correlation structure features are identified, the corresponding risk pattern is identified based on a preset pattern library; Based on the matching degree, the characteristic strength of the risk pattern, and the absolute deviation level of the key cargo status parameters in the comparison results, the overall condition of the fresh goods to be transferred is evaluated to obtain the evaluation result. Based on the evaluation results, the state category is determined, and based on the degree of matching, the phase compliance evaluation is generated.

4. The intelligent cold chain logistics transfer method with remote monitoring according to claim 3, characterized in that, The overall condition of the fresh goods to be transferred is evaluated based on the matching degree, the characteristic strength of the risk pattern, and the absolute deviation level of the key cargo status parameters in the comparison results, to obtain the evaluation results, including: The deviation between the matching degree and the preset benchmark matching degree is calculated to obtain the first evaluation component; Extract the feature intensity of the risk pattern to obtain the second evaluation component; The absolute deviation levels of the cargo status parameters in the comparison results are normalized and weighted to obtain the third evaluation component. Based on the evaluation weight configuration of the current transit stage, the first evaluation component, the second evaluation component, and the third evaluation component are weighted and fused to generate a comprehensive evaluation score; The comprehensive evaluation score is input into a preset evaluation decision model to obtain the evaluation result.

5. The intelligent cold chain logistics transfer method with remote monitoring according to claim 4, characterized in that, When the signal-to-noise ratio is less than the quality threshold, or the fluctuation amplitude is greater than the stability threshold, it is determined that the environmental parameter data quality has deteriorated, and a parameter adjustment mechanism is executed, including: Based on the cause category of the decline in environmental parameter data quality, select the corresponding adjustment strategy. The cause category includes the quality decline caused by the signal-to-noise ratio and the quality decline caused by the fluctuation amplitude. When the cause category is the quality degradation caused by the signal-to-noise ratio, a preset first step length is used to increase the smoothing intensity of the fused feature data, and a preset second step length is used to reduce the weight of the absolute deviation level of the cargo status parameter in the third evaluation component. When the cause category is the quality degradation due to the fluctuation amplitude, a preset third step length is used to extend the sliding analysis window of the time series data in the comparison results, and an anti-interference correction term for the causal influence spectrum is added accordingly. The first step length is the adjustment increment of the smoothing coefficient of the fused feature data, with the preset logic being a linear mapping value of the signal-to-noise ratio degradation degree. The second step length is the reduction of the absolute deviation level weight of the cargo state parameters, with the preset logic being the product of the signal-to-noise ratio degradation degree and the cargo category sensitivity coefficient. The third step length is the extension increment of the time series data sliding analysis window, with the preset logic being a segmented threshold mapping value of the fluctuation amplitude exceeding the standard multiple. Based on the selected adjustment strategy, the adjustable parameters of the data preprocessing module or weighted fusion module of the evaluation decision model are updated.

6. The intelligent cold chain logistics transfer method with remote monitoring according to claim 4, characterized in that, Also includes: The status category, the stage compliance assessment, the risk mode, the comprehensive assessment score, and related timestamp information are structured and encapsulated to generate a status snapshot data package of the fresh goods to be transferred in the current transfer stage. Based on the state snapshot data packet and the historical state snapshot data packet, a full-link state evolution map of the fresh goods to be transferred is constructed along the time dimension. The entire link state evolution map is analyzed to identify key nodes and associated environmental events of state deterioration, and the main causes of state deterioration are traced based on the transmission path of the risk mode to obtain the tracing results. Based on the traceability results, a traceability report is generated and pushed to the remote monitoring terminal for display, and a pre-control strategy is generated.

7. The intelligent cold chain logistics transfer method with remote monitoring according to claim 6, characterized in that, The process of generating a remote control signal recognizable by the remote monitoring terminal based on the state category and the stage compliance assessment includes: Based on the aforementioned state categories, determine the priority level and control mode of environmental regulation required for the current transfer phase; Based on the degree of matching in the phased compliance assessment, the risk pattern, and the main causes in the retrospective report, a multi-instruction coordinated control strategy is generated. Conflict detection is performed between the pre-control strategy and the multi-instruction collaborative control strategy; If no conflict is detected, the two signals will be merged into an enhanced remote control signal; If a conflict is detected between the two, arbitration is conducted based on the urgency of the state category, the characteristic strength of the risk mode, and the connection requirements of the next transfer stage to generate the final remote control signal.

8. The intelligent cold chain logistics transfer method with remote monitoring according to claim 7, characterized in that, Also includes: The transshipment phase includes sea transport, land transport, and air transport; Before the land transport is converted to the air transport, the flight information of the fresh goods to be transferred is obtained, including the pre-allocated parking positions and the estimated ground transfer operation time. Based on the geographical location of the parking position, the current time, and real-time meteorological data, predict the temperature change trend of the fresh goods to be transferred when exposed to the apron environment during the ground transfer operation. Based on the temperature change trend, the current status category of the fresh goods to be transferred, and the transportation requirements of the next transfer stage, the target temperature and cooling rate required for the ground cold chain vehicle or freight station cold storage to receive the goods are calculated. Before and after the aircraft lands, the connection control command, including the target temperature and cooling rate, will be sent to the designated ground connection unit in advance as a connection signal. Upon receiving the connection control command, the ground connection unit adjusts its own equipment operating status.

9. A remotely monitored intelligent cold chain logistics transfer platform system, characterized in that, include: The information collection module is used to collect multi-dimensional physical data and cold chain environment parameter data of the fresh goods to be transferred, and to receive transfer task information from the logistics management platform. The transfer task information includes the attributes, transportation requirements and transfer stage of the fresh goods to be transferred. The strategy generation module is used to determine appropriate status monitoring thresholds and environmental control targets based on the transshipment stage, the attributes of the fresh goods to be transshipped, and the transportation requirements. The data fusion module is used to extract and fuse features from the multi-dimensional physical data and the cold chain environment parameter data to obtain fused feature data. The intelligent analysis module is used to perform linked analysis and comprehensive judgment on the fused feature data based on the transshipment stage, the status monitoring threshold, the environmental control target, and preset association rules, to obtain the status category and stage compliance assessment of the fresh goods to be transshipped; wherein, the linked analysis and comprehensive judgment includes: using a preset evaluation decision model to evaluate the overall status of the fresh goods to be transshipped, and determining the status category and stage compliance assessment based on the evaluation results; and, during the evaluation process using the evaluation decision model, adjusting the adjustable parameters of the evaluation decision model based on the quality indicators of the currently collected cold chain environmental parameter data, including: Calculate the signal-to-noise ratio (SNR) and fluctuation amplitude of cold chain environmental parameter data within a preset time window; compare the SNR with a preset quality threshold and the fluctuation amplitude with a preset stability threshold; when the SNR is less than the quality threshold or the fluctuation amplitude is greater than the stability threshold, determine that the environmental parameter data quality has deteriorated, and execute a parameter adjustment mechanism, including: Based on the system cause category, the smoothing processing intensity, weight allocation, sliding analysis window, and anti-interference correction term are adaptively adjusted, and the adjustable parameters of the data preprocessing module or weighted fusion module of the evaluation decision model are updated synchronously. The instruction generation module is used to generate remote control signals that can be recognized by the remote monitoring terminal based on the state category and the stage compliance assessment. The linkage execution module is used to send the remote control signal to the cold chain transfer unit corresponding to the current transfer stage through the remote monitoring terminal; in response to the remote control signal, each cold chain transfer unit starts local control action and feeds back the execution result to the remote monitoring terminal.

Images