Online monitoring method and system for refrigerated container based on digital twinning

Abstract

The invention discloses a digital twinning-based refrigerated container online monitoring method and system, and the method comprises the following steps: collecting refrigerated container data, and generating a refrigerated container basic information base; continuously monitoring data, and constructing digital twin bodies of the refrigerated container; configuring a heat flow gating fusion unit and an access control disturbance temperature field time sequence deduction model in the digital twin of the refrigerated container, and generating a three-dimensional temperature field prediction result and a temperature control risk assessment result; generating an alarm processing record and a to-be-patrolled list according to alarm rule configuration and role alarm pushing; and making a patrol plan and pushing the patrol plan to a mobile terminal application system to complete field data verification and abnormal condition reporting. According to the invention, a digital twin-driven refrigerated container intelligent monitoring system is constructed, temperature control risk prediction and closed-loop patrol update are realized, and the system has the advantages of high-precision early warning, low false alarm rate and remarkable improvement of operation and maintenance efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of equipment operation monitoring, in particular to a cold storage container online monitoring method and system based on digital twinning. BACKGROUND

[0002] The cold storage container needs to maintain a stable temperature control environment during the process of port area, yard and long-distance transportation, and its operation safety depends on the continuous monitoring of temperature, power, equipment running state and access control state. The existing monitoring method is mostly based on periodic collection of single sensor data and manual patrol record, and the state is obtained through temperature probe, controller log and manual inspection table, and the monitoring system generally adopts rule triggering mode to alarm for over-temperature, insufficient power or equipment failure. However, such monitoring means usually cannot model the multi-source monitoring data uniformly, and also lack the ability to visualize the temperature distribution and heat flow change in the container, resulting in obvious delay in response of the monitoring system to complex environmental disturbance, container door opening and other scenes, and the system cannot provide predictive risk identification.

[0003] In the prior art, the monitoring system of the cold storage container generally lacks a structured description of the dynamic behavior of the container, cannot establish a temperature field model and an operation condition model evolving with time, and cannot carry out combined analysis based on access control state, external environmental conditions and other factors. The lack of digital twinning mechanism makes it impossible to effectively associate the monitoring data, and the reported alarms are usually based on single-point threshold judgment, resulting in frequent false alarms and missed alarms. At the same time, the existing patrol management process still relies on manual route planning, and there is no feedback loop between alarm records and patrol process, which cannot continuously correct the digital model. Therefore, the existing technology cannot meet the high reliability requirements of the cold storage container operation in real-time monitoring, temperature control risk prediction, accurate alarm pushing and patrol closed-loop management.

[0004] Therefore, how to provide a cold storage container online monitoring method and system based on digital twinning is a problem that those skilled in the art need to solve. SUMMARY

[0005] One object of the present application is to provide a cold storage container online monitoring method and system based on digital twinning, which constructs a digital twinning driven intelligent monitoring system of the cold storage container, realizes temperature control risk prediction and closed-loop patrol update, and has the advantages of high-precision early warning, low false alarm rate and significant improvement of operation and maintenance efficiency.

[0006] According to the cold storage container online monitoring method based on digital twinning of the present application, the following steps are included:

[0007] The container number, the company to which the container belongs, the type of goods, the preset temperature control range, the time in place and the location in place of the cold storage container are collected, an in-place file is generated, and a cold storage container basic information database is formed;

[0008] Based on the basic information database of refrigerated containers, continuously monitor the data, construct a digital twin of refrigerated containers, and establish the three-dimensional temperature field digital twin status and operating condition status.

[0009] A heat flow gating fusion unit and a time-series model for door access disturbance temperature field are configured in the digital twin of a refrigerated container. Door area scenario-specific training and loss design are performed on the door access status and external environmental conditions to generate three-dimensional temperature field prediction results and temperature control risk assessment results.

[0010] The temperature control risk assessment results are linked with the on-site records. Alarms are generated according to alarm rules and role-based alarm push, such as temperature over-limit alarms, low power alarms, and equipment failure alarms. Alarm processing records and a list of items to be inspected are formed.

[0011] Based on the list of items to be inspected, an inspection plan is developed and pushed to the mobile application system. On-site data verification and abnormal situation reporting are completed, and inspection records and abnormal handling results are generated and updated.

[0012] Optionally, the generation of the refrigerated container basic information database specifically includes:

[0013] Collect refrigerated container entry instructions, record container number, company, cargo type, preset temperature control range, entry time, on-site location and contact information, and generate a set of original basic information data for refrigerated containers;

[0014] The original basic information data set of refrigerated containers is subjected to field integrity verification, content legality verification, and container number uniqueness verification. It is then converted into a standardized record set of basic information. Records that fail the verification are marked as abnormal records to be processed, and an on-site status identifier field is generated.

[0015] The standardized record set of basic information is aggregated using the container number as the primary key. All standardized records of the same container number within one on-site period are integrated into an on-site file, and the corresponding on-site period number and on-site duration fields are generated.

[0016] A basic information database for refrigerated containers was built based on on-site records;

[0017] By linking the refrigerated container basic information database with the terminal layout data, spatial mapping records composed of container number, on-site location, on-site duration, and on-site status identifier are written into the refrigerated container basic information database, forming an on-site archive spatial mapping set.

[0018] Optionally, the generation of the three-dimensional temperature field digital twin state and operating condition state specifically includes:

[0019] Search the on-site archives and create a corresponding digital twin of the refrigerated container for each on-site archive, forming a set of digital twins of the refrigerated containers corresponding to the on-site refrigerated containers;

[0020] Configure monitoring data access channels for the digital twin set of refrigerated containers, perform timestamp completion, container number matching, field format standardization and outlier identification, and convert them into a standardized set of monitoring data records;

[0021] The standardized monitoring data record set is associated with the refrigerated container digital twin set according to the container number to obtain the three-dimensional temperature field digital twin status and operating condition status that are dynamically updated over time.

[0022] Establish a state time series for the three-dimensional temperature field digital twin state and operating condition state in chronological order;

[0023] The latest three-dimensional temperature field digital twin status and operating condition status are linked and written back to the on-site archives to form a set of status index records that can be retrieved by container number, on-site location, or company.

[0024] Optionally, the generation of the temperature control risk assessment result specifically includes:

[0025] Retrieve on-site records from the refrigerated container basic information database, and read the three-dimensional temperature field digital twin state time series and operating condition state time series from the refrigerated container digital twin set to generate a door area scene training sample set;

[0026] In a digital twin of a refrigerated container, a heat flow gating fusion unit is instantiated to generate a heat flow gating feature sequence, which is then associated with the corresponding three-dimensional temperature field digital twin state time slice to form a heat flow gating input sequence.

[0027] In the digital twin of a refrigerated container, a time-series model for the temperature field of access control disturbance is configured. The training sample set of the door area scene is divided into multiple door area scene subsets according to cargo type, preset temperature control range and external environmental conditions. Training data sets for the time-series model of the temperature field of access control disturbance are constructed respectively.

[0028] For each subset of door area scenarios, a door area scenario-specific training and loss design scheme is set in the access control disturbance temperature field time series extrapolation model to obtain the converged access control disturbance temperature field time series extrapolation model parameters, which are then deployed to the corresponding refrigerated container digital twins.

[0029] The current access control status, external environmental conditions and the latest three-dimensional temperature field digital twin status are obtained from online monitoring. The heat flow gating fusion unit generates a heat flow gating feature sequence and inputs it into the access control disturbance temperature field time series extrapolation model to obtain the three-dimensional temperature field prediction result time series.

[0030] Based on the time series of three-dimensional temperature field prediction results and the preset temperature control range in the refrigerated container basic information database, the predicted temperature distribution of each future time interval is analyzed on a time-by-time basis to generate temperature control risk assessment results, which are then written into the digital twin of the refrigerated container and linked to the corresponding on-site file.

[0031] Optionally, the generation of the alarm processing record and the list of items to be inspected specifically includes:

[0032] Read the temperature control risk assessment results and retrieve the corresponding on-site files to generate a risk association record set, which serves as input for alarm determination;

[0033] Each risk-related record in the set is processed, the temperature risk score is calculated, the temperature exceedance alarm level and the temperature exceedance alarm triggering identifier are determined according to the interval, and a set of candidate records for temperature exceedance alarms is generated.

[0034] By combining the operating status in the digital twin of the refrigerated container, a threshold judgment is made on the power status field to generate a power risk index, which is then matched with the power alarm conditions to generate a set of candidate records for low power alarm.

[0035] Perform fault mode matching on the equipment operating status field to generate equipment fault risk indicators, and match them with equipment fault alarm conditions to generate a set of equipment fault alarm candidate records;

[0036] The candidate records for temperature over-limit alarms, low power alarms, and equipment malfunction alarms are merged into a total set of alarm candidate records. This set of alarm records is then compiled and supplemented with a summary of temperature control risk assessment results and a summary of on-site records.

[0037] Read the alarm record set to generate a role-based alarm push task set, and push it to the monitoring terminal and mobile application system. Generate an initial alarm processing record for each alarm record, recording the alarm generation time, target role, push channel and initial processing status.

[0038] During the alarm handling process, the system receives alarm confirmation operations, processing progress information, and processing result information returned by the monitoring terminal and mobile application system, updates the status of the alarm handling records, and forms a set of alarm handling records with handling status identifiers.

[0039] Based on the set of alarm handling records marked as pending inspection, a pending inspection list is generated according to box number, location, alarm type, and alarm level, and the on-site files are associated.

[0040] Optionally, the generation of the inspection records and anomaly handling results specifically includes:

[0041] Read the list to be inspected and retrieve the on-site files in the refrigerated container basic information database. Combine the container number, on-site location, alarm type and alarm level to generate a set of inspection tasks. Form an inspection plan according to the regional order and time order and write it into the mobile application system to obtain the inspection plan distribution results.

[0042] Load the on-site archives and digital twin summary information of the refrigerated container corresponding to the inspection task into the mobile application system, collect temperature, power indication, equipment operation indication and access control status on site, compare with the status and operating condition summary information of the three-dimensional temperature field digital twin, and generate the initial data of the inspection record.

[0043] When abnormal temperature, abnormal power, or equipment failure is detected during the inspection, the mobile application system records the abnormality type, on-site reading, abnormality description, and image information to form an abnormality reporting record. This record is then linked with the initial data of the inspection record and the on-site archives to obtain an initial set of abnormality handling results.

[0044] The initial data of the inspection records and the initial set of anomaly handling results are fed back to the digital twin of the refrigerated container and the basic information database of the refrigerated container. The status of the three-dimensional temperature field digital twin and the operating condition status are corrected and updated. The inspection records and anomaly handling result summaries are added to the on-site archives, and the alarm handling record set is updated to generate a new list to be inspected.

[0045] According to an embodiment of the present invention, an online monitoring system for refrigerated containers based on digital twins includes:

[0046] The basic information management module is used to collect information such as refrigerated container number, company, cargo type, preset temperature control range, on-site time and on-site location, generate on-site files and form a basic information database for refrigerated containers.

[0047] The digital twin monitoring module is used to access temperature, operating status, power consumption, and access control status monitoring data based on the refrigerated container basic information database, construct a digital twin of the refrigerated container, and establish a three-dimensional temperature field digital twin status and operating condition status.

[0048] The door area scenario analysis module is used to configure the heat flow gate control fusion unit and the door access disturbance temperature field time series extrapolation model in the digital twin of the refrigerated container. It performs door area scenario-specific training and loss design on the door access status and external environmental conditions, and generates three-dimensional temperature field prediction results and temperature control risk assessment results.

[0049] The alarm management module is used to associate temperature control risk assessment results with on-site records, configure alarm rules and role-based alarm push to generate alarms for excessive temperature, low power and equipment failure, and form alarm handling records and a list of items to be inspected.

[0050] The inspection management and mobile application module is used to formulate inspection plans based on the list of items to be inspected and push them to the mobile application system, perform on-site data verification and abnormal situation reporting, and generate inspection records and abnormal handling results to update the refrigerated container basic information database and refrigerated container digital twin.

[0051] The beneficial effects of this invention are:

[0052] This invention constructs a complete digital twin monitoring system covering a refrigerated container's basic information database, a three-dimensional temperature field digital twin status, operational status, a heat flow gating fusion unit, and a temporal extrapolation model of the temperature field under access control disturbances. This system enables precise, dynamic, and predictive monitoring of the entire lifecycle of refrigerated containers. Compared to existing technologies that rely solely on static threshold judgments based on single-point sensor data, this invention leverages the coupling relationship between access control status, external environmental conditions, and internal heat flow changes to generate three-dimensional temperature field prediction results and temperature control risk assessment results. This allows for early identification of temperature control anomalies, significantly reducing false alarms and missed alarms. Furthermore, through the linkage of the alarm management module, inspection management module, and mobile application module, a closed-loop update system is achieved, encompassing alarm recording, role-based alarm push notifications, inspection plan formulation, on-site data verification, and anomaly handling result feedback. This enables the refrigerated container's digital twin to continuously self-correct and maintain high confidence, providing verifiable, traceable, and predictable decision-making support for yard operations and maintenance. Consequently, this significantly improves the real-time performance, accuracy, and operational efficiency of cold chain monitoring. Attached Figure Description

[0053] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0054] Figure 1 This is a flowchart of an online monitoring method for refrigerated containers based on digital twins proposed in this invention;

[0055] Figure 2 This is a schematic diagram of the digital twin of a refrigerated container and the three-dimensional temperature field analysis, which is part of the online monitoring method for refrigerated containers based on digital twins proposed in this invention.

[0056] Figure 3 This is a schematic diagram of the temperature control risk alarm and patrol closed-loop processing of an online monitoring method for refrigerated containers based on digital twins proposed in this invention. Detailed Implementation

[0057] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0058] refer toFigures 1-3 A method for online monitoring of refrigerated containers based on digital twins includes the following steps:

[0059] Collect the container number, company, cargo type, preset temperature control range, on-site time and location of the refrigerated container, generate on-site files and form a basic information database of refrigerated containers;

[0060] Based on the basic information database of refrigerated containers, continuously monitor the data, construct a digital twin of refrigerated containers, and establish the three-dimensional temperature field digital twin status and operating condition status.

[0061] A heat flow gating fusion unit and a time-series model for door access disturbance temperature field are configured in the digital twin of a refrigerated container. Door area scenario-specific training and loss design are performed on the door access status and external environmental conditions to generate three-dimensional temperature field prediction results and temperature control risk assessment results.

[0062] The temperature control risk assessment results are linked with the on-site records. Alarms are generated according to alarm rules and role-based alarm push, such as temperature over-limit alarms, low power alarms, and equipment failure alarms. Alarm processing records and a list of items to be inspected are formed.

[0063] Based on the list of items to be inspected, an inspection plan is developed and pushed to the mobile application system. On-site data verification and abnormal situation reporting are completed, and inspection records and abnormal handling results are generated and updated.

[0064] In this embodiment, the generation of the refrigerated container basic information database specifically includes:

[0065] Collect refrigerated container entry instructions, record container number, company, cargo type, preset temperature control range, entry time, on-site location and contact information, and generate a set of original basic information data for refrigerated containers;

[0066] The original basic information data set of refrigerated containers is subjected to field integrity verification, content legality verification, and container number uniqueness verification. It is then converted into a standardized record set of basic information. Records that fail the verification are marked as abnormal records to be processed, and an on-site status identifier field is generated.

[0067] The standardized record set of basic information is aggregated using the container number as the primary key. All standardized records of the same container number within one on-site period are integrated into an on-site file, and the corresponding on-site period number and on-site duration fields are generated.

[0068] The on-site file includes the container number, the company to which it belongs, the type of goods, the preset temperature control range, the location on-site, the entry time, the on-site status indicator, and the contact information;

[0069] A basic information database for refrigerated containers was built based on on-site records;

[0070] The refrigerated container basic information database establishes a retrieval structure according to the container number index, the on-site location index, and the company index. On-site records are written into the refrigerated container basic information database, and the entry time, on-site duration, and on-site status identifier are used as time-series attribute fields of the refrigerated container basic information database, providing stable and traceable basic information input for building a digital twin of refrigerated containers.

[0071] By linking the refrigerated container basic information database with the terminal layout data, spatial mapping records composed of container number, on-site location, on-site duration, and on-site status identifier are written into the refrigerated container basic information database, forming an on-site archive spatial mapping set.

[0072] In this embodiment, the generation of the three-dimensional temperature field digital twin state and the operating condition state specifically includes:

[0073] Search the on-site archives and create a corresponding digital twin of the refrigerated container for each on-site archive, forming a set of digital twins of the refrigerated containers corresponding to the on-site refrigerated containers;

[0074] In the digital twin of the refrigerated container, initialize static attribute fields such as container number, company, cargo type, preset temperature control range, on-site location and on-site status identifier, and reserve dynamic attribute fields for the three-dimensional temperature field digital twin status and operating condition status.

[0075] Configure monitoring data access channels for the digital twin set of refrigerated containers, perform timestamp completion, container number matching, field format standardization and outlier identification, and convert them into a standardized set of monitoring data records;

[0076] Among them, monitoring data is continuously received from temperature acquisition devices, operation status acquisition devices, power consumption acquisition devices and access control status acquisition devices. The monitoring data in the standardized record set of the monitoring data includes temperature, operation status, power consumption and access control status.

[0077] The standardized monitoring data record set is associated with the refrigerated container digital twin set according to the container number to obtain the three-dimensional temperature field digital twin status and operating condition status that are dynamically updated over time.

[0078] Among them, each monitoring data is standardized and recorded in the corresponding refrigerated container digital twin to update the three-dimensional temperature field digital twin status and operating condition status. The temperature field is written into the spatial node corresponding to the three-dimensional temperature field digital twin status, and the operating status field, power field, and access control status field are written into the equipment operating status, power status, and access control status subfields in the operating condition status.

[0079] Establish a state time series for the three-dimensional temperature field digital twin state and operating condition state in chronological order;

[0080] Specifically, the updated state of the three-dimensional temperature field digital twin and the operating condition state are written into the corresponding state time series to form a state time series and an operating condition state time series of the three-dimensional temperature field digital twin composed of continuous state segments, and the corresponding monitoring data source information is recorded in the state time series.

[0081] The latest three-dimensional temperature field digital twin status and operating condition status are linked and written back to the on-site archives to form a set of status index records that can be retrieved by container number, on-site location, or company.

[0082] In this embodiment, the generation of the temperature control risk assessment result specifically includes:

[0083] Retrieve on-site records from the refrigerated container basic information database, and read the three-dimensional temperature field digital twin state time series and operating condition state time series from the refrigerated container digital twin set to generate a door area scene training sample set;

[0084] The door area scene training sample set is generated by time alignment and scene classification of the access control status time series and the external environmental condition time series based on the box number, on-site location, preset temperature control range and cargo type in the on-site file. It includes access control status, external environmental conditions, three-dimensional temperature field digital twin status time slices and operating condition status time slices.

[0085] In a digital twin of a refrigerated container, a heat flow gating fusion unit is instantiated to generate a heat flow gating feature sequence, which is then associated with the corresponding three-dimensional temperature field digital twin state time slice to form a heat flow gating input sequence.

[0086] Specifically, the access control status time series and the external environmental condition time series are used as the door zone disturbance input, and the three-dimensional temperature field digital twin status time series and the operating condition status time series are used as the basic heat flow input. The door zone disturbance feature sequence and the basic heat flow feature sequence are constructed inside the heat flow gating fusion unit. The door zone disturbance feature sequence and the basic heat flow feature sequence are gating weighted and fused to obtain the heat flow gating feature sequence, which characterizes the impact of door zone disturbance on the heat flow inside the box.

[0087] In the digital twin of a refrigerated container, a time-series model for the temperature field of access control disturbance is configured. The training sample set of the door area scene is divided into multiple door area scene subsets according to cargo type, preset temperature control range and external environmental conditions. Training data sets for the time-series model of the temperature field of access control disturbance are constructed respectively.

[0088] The access control disturbance temperature field time series extrapolation model uses the heat flow gating input sequence and the three-dimensional temperature field digital twin state time slice of the previous time period as input, and the three-dimensional temperature field digital twin state time slice of the next time period as the supervision target.

[0089] For each subset of door area scenarios, a door area scenario-specific training and loss design scheme is set in the access control disturbance temperature field time series extrapolation model to obtain the converged access control disturbance temperature field time series extrapolation model parameters, which are then deployed to the corresponding refrigerated container digital twins.

[0090] The specific training and loss design scheme for the access control area takes the degree of temperature deviation from the preset temperature control range, the amplitude of temperature fluctuation, and the time required for the temperature to recover from the change in access control status as training objectives, and performs iterative training in the access control disturbance temperature field time series extrapolation model.

[0091] The current access control status, external environmental conditions and the latest three-dimensional temperature field digital twin status are obtained from online monitoring. The heat flow gating fusion unit generates a heat flow gating feature sequence and inputs it into the access control disturbance temperature field time series extrapolation model to obtain the three-dimensional temperature field prediction result time series.

[0092] Based on the time series of three-dimensional temperature field prediction results and the preset temperature control range in the refrigerated container basic information database, the predicted temperature distribution of each time interval in the future is analyzed on a time-by-time basis to generate temperature control risk assessment results, which are written into the digital twin of the refrigerated container and linked to the corresponding on-site file.

[0093] Among them, the predicted temperature distribution of each time interval in the future is statistically analyzed to predict the spatial range, duration and recovery trend of the temperature exceeding the preset temperature control range. The time series of the three-dimensional temperature field prediction results corresponding to the same on-site file are aggregated to generate a set of temperature control risk indicators that include the degree of temperature exceeding the limit, the duration of exceeding the limit and the recovery ability. The temperature control risk assessment results are then obtained.

[0094] In this embodiment, the generation of the alarm processing record and the list of items to be inspected specifically includes:

[0095] Read the temperature control risk assessment results and retrieve the corresponding on-site files to generate a risk association record set, which serves as input for alarm determination;

[0096] The risk association record set includes the container number, on-site location, preset temperature control range, temperature control risk assessment result, and on-site status identifier;

[0097] Each risk-related record in the set is processed, the temperature risk score is calculated, the temperature exceedance alarm level and the temperature exceedance alarm triggering identifier are determined according to the interval, and a set of candidate records for temperature exceedance alarms is generated.

[0098] The temperature risk score is obtained by comparing the degree of temperature exceeding the limit, the duration of exceeding the limit, and the recovery ability in the temperature control risk assessment results with the preset temperature control range and the corresponding alarm level boundary.

[0099] By combining the operating status in the digital twin of the refrigerated container, a threshold judgment is made on the power status field to generate a power risk index, which is then matched with the power alarm conditions to generate a set of candidate records for low power alarm.

[0100] Perform fault mode matching on the equipment operating status field to generate equipment fault risk indicators, and match them with equipment fault alarm conditions to generate a set of equipment fault alarm candidate records;

[0101] The candidate records for temperature over-limit alarms, low power alarms, and equipment malfunction alarms are merged into a total set of alarm candidate records. This set of alarm records is then compiled and supplemented with a summary of temperature control risk assessment results and a summary of on-site records.

[0102] The alarm record set is obtained by deduplicating and merging the total set of alarm candidate records according to container number, location, alarm type and alarm level, and merging multiple alarms of the same type for the same refrigerated container within the same time window, including temperature over-limit alarms, low power alarms and equipment failure alarms.

[0103] Read the alarm record set to generate a role-based alarm push task set, and push it to the monitoring terminal and mobile application system. Generate an initial alarm processing record for each alarm record, recording the alarm generation time, target role, push channel and initial processing status.

[0104] The set of role-based alarm push tasks is mapped to different roles such as on-site staff, maintenance staff, and management staff based on the company, location, and preset role subscription relationships in the on-site file;

[0105] During the alarm handling process, the system receives alarm confirmation operations, processing progress information, and processing result information returned by the monitoring terminal and mobile application system, updates the status of the alarm handling records, and forms a set of alarm handling records with handling status identifiers.

[0106] Among them, the alarm handling record records the alarm confirmation time, on-site handling measures and closed-loop completion time. Alarm handling records that have been handled are marked as closed, and alarm handling records that have not been confirmed or handled are marked as pending inspection.

[0107] Based on the set of alarm handling records marked as pending inspection, a pending inspection list is generated according to box number, location, alarm type, and alarm level, and the on-site files are associated.

[0108] In this embodiment, the generation of the inspection record and the anomaly handling result specifically includes:

[0109] Read the list to be inspected and retrieve the on-site files in the refrigerated container basic information database. Combine the container number, on-site location, alarm type and alarm level to generate a set of inspection tasks. Form an inspection plan according to the regional order and time order and write it into the mobile application system to obtain the inspection plan distribution results.

[0110] Load the on-site archives and digital twin summary information of the refrigerated container corresponding to the inspection task into the mobile application system, collect temperature, power indication, equipment operation indication and access control status on site, compare with the status and operating condition summary information of the three-dimensional temperature field digital twin, and generate the initial data of the inspection record.

[0111] When abnormal temperature, abnormal power, or equipment failure is detected during the inspection, the mobile application system records the abnormality type, on-site reading, abnormality description, and image information to form an abnormality reporting record. This record is then linked with the initial data of the inspection record and the on-site archives to obtain an initial set of abnormality handling results.

[0112] The initial data of the inspection records and the initial set of anomaly handling results are fed back to the digital twin of the refrigerated container and the basic information database of the refrigerated container. The status of the three-dimensional temperature field digital twin and the operating condition status are corrected and updated. The inspection records and anomaly handling result summaries are added to the on-site archives, and the alarm handling record set is updated to generate a new list to be inspected.

[0113] A digital twin-based online monitoring system for refrigerated containers includes:

[0114] The basic information management module is used to collect information such as refrigerated container number, company, cargo type, preset temperature control range, on-site time and on-site location, generate on-site files and form a basic information database for refrigerated containers.

[0115] The digital twin monitoring module is used to access temperature, operating status, power consumption, and access control status monitoring data based on the refrigerated container basic information database, construct a digital twin of the refrigerated container, and establish a three-dimensional temperature field digital twin status and operating condition status.

[0116] The door area scenario analysis module is used to configure the heat flow gate control fusion unit and the door access disturbance temperature field time series extrapolation model in the digital twin of the refrigerated container. It performs door area scenario-specific training and loss design on the door access status and external environmental conditions, and generates three-dimensional temperature field prediction results and temperature control risk assessment results.

[0117] The alarm management module is used to associate temperature control risk assessment results with on-site records, configure alarm rules and role-based alarm push to generate alarms for excessive temperature, low power and equipment failure, and form alarm handling records and a list of items to be inspected.

[0118] The inspection management and mobile application module is used to formulate inspection plans based on the list of items to be inspected and push them to the mobile application system, perform on-site data verification and abnormal situation reporting, and generate inspection records and abnormal handling results to update the refrigerated container basic information database and refrigerated container digital twin.

[0119] Example 1:

[0120] To verify the feasibility of this invention in practice, it was applied to a scenario of centralized storage and supervision of refrigerated containers at a coastal port. In this scenario, there are a large number of refrigerated containers, they are densely stacked, and the types of goods are complex. During frequent entry and exit, loading and unloading, and inspection of the containers, temperature fluctuations are easily caused by access control openings, changes in the external environment, or abnormal equipment operation. Traditional monitoring methods mainly rely on single-point temperature alarms and manual inspections, which have problems such as delayed detection, a large number of false alarms, and difficulty in tracing back inspection feedback, making it difficult to meet the needs of high-intensity cold chain supervision.

[0121] In practical application, the first step is to register the information of refrigerated containers entering the terminal, collecting information such as container number, company, cargo type, preset temperature control range, and location, forming an on-site file and writing it into the refrigerated container basic information database. Subsequently, by accessing monitoring data such as temperature, operating status, power consumption, and access control status, a corresponding digital twin of each refrigerated container is built, continuously updating the three-dimensional temperature field digital twin status and operating condition status. During container operation, when a door is opened or external environmental conditions change, the system, based on the heat flow gating fusion unit, performs a fusion analysis of access control status, external environmental conditions, and changes in internal heat flow. It also uses an access control disturbance temperature field time-series extrapolation model to predict future temperature distribution, generating three-dimensional temperature field prediction results and temperature control risk assessment results in advance. This process allows managers to know the risk trend before the temperature actually exceeds the limit, rather than passively responding only after an anomaly occurs.

[0122] After operating for a period of time, continuous comparison of system predictions with actual inspection records revealed that the system can proactively identify potential temperature control risks during periods of frequent access control operations and drastic changes in the external environment, reducing passive temperature alarms. After linking the temperature control risk assessment results with on-site records, the system generates alarms for excessive temperature, low battery, and equipment malfunction according to alarm rules, and promptly sends this information to on-duty and maintenance personnel via role-based alarm push notifications. Compared to the previous method of manually screening alarms, alarm information is more focused on the risks that truly require attention, and alarm handling records are clearer and more traceable.

[0123] In the inspection and management phase, the system automatically generates inspection plans based on the list of items to be inspected and pushes them to the mobile application system. Upon arrival at the site, inspection personnel verify the status against the 3D temperature field digital twin and operational conditions using the mobile app. Any anomalies are reported directly, and on-site information is saved. Inspection records and anomaly handling results are fed back in real-time to the refrigerated container digital twin and the basic information database for use in correcting subsequent temperature field predictions and risk assessments. Continuous comparison of the same station's operation at different times reveals a significant improvement in inspection efficiency, a marked reduction in duplicate and ineffective inspections, and more timely responses to temperature control anomalies. Overall, the safety and management efficiency of refrigerated container operations are effectively improved.

[0124] Table 1. Performance comparison between digital twin-based online monitoring method for refrigerated containers and traditional methods.

[0125] Comparison index Traditional monitoring method The present application Average discovery time of temperature anomaly (relative to anomaly occurrence) 32 minutes after anomaly occurrence 18 minutes before anomaly occurrence Average duration of temperature anomaly 96 minutes 41 minutes Number of temperature overruns in single-box presence period 7 times 2 times False alarm ratio triggered by access control opening 38% 9% Total number of temperature control related alarms 1240 512 Proportion of alarms actually requiring on-site handling 46% 81% Average number of daily patrol tasks 64 39 Average processing time of single patrol 22 minutes 14 minutes Proportion of repeated and invalid tasks found by patrol 31% 6% Average duration of temperature anomaly closed loop completion 5.2 hours 2.1 hours Completeness rate of traceability of refrigerated container operating status 72% 98%

[0126] As shown in Table 1, after adopting the digital twin online monitoring method of this invention, the detection time of temperature anomalies has been advanced from "32 minutes after the anomaly occurs" in the traditional method to "18 minutes before the anomaly occurs." This indicates that the system no longer relies on single-point threshold alarms, but achieves proactive identification through three-dimensional temperature field prediction results and temperature control risk assessment. This change directly reduces the average duration of temperature anomalies from 96 minutes to 41 minutes, significantly reducing the time window for goods to be in unsafe temperature control zones.

[0127] Regarding alarm quality, traditional methods often trigger numerous false alarms during normal operations such as access control opening, with the false alarm rate reaching 38%. This invention, however, uses a heat flow gating fusion unit and a time-series model to extrapolate the temperature field of access control disturbances, distinguishing between access control behavior and actual heat flow changes, thus reducing this rate to 9%. Simultaneously, the total number of alarms decreased from 1240 to 512, but the proportion of alarms requiring actual on-site handling increased from 46% to 81%, indicating that alarms are more focused on real risks and reducing the burden of ineffective responses for on-duty personnel.

[0128] In terms of inspection efficiency, the inspection list generated based on temperature control risk assessment results reduces the number of inspection tasks from an average of 64 per day to 39, while the processing time for a single inspection is shortened from 22 minutes to 14 minutes. Combined with the data showing that the proportion of duplicate and invalid inspections decreased from 31% to 6%, it can be seen that this invention, through a risk-driven inspection plan generation mechanism, effectively avoids the problems of repetitive visits and low-value inspections that exist in traditional manual experience-based inspections.

[0129] From the perspective of overall closed-loop operation, the average time to complete the closed loop for temperature control anomalies has been shortened from 5.2 hours to 2.1 hours, and the traceability and completeness rate of refrigerated container operation status has increased to 98%. This result stems from the unified association and feedback update mechanism of on-site records, digital twins, alarm handling records, and inspection records, enabling each anomaly to form a complete data link, providing continuous and effective data support for subsequent risk assessment and model correction.

[0130] Based on the above data, it can be confirmed that the present invention has achieved significant improvements in several key dimensions, such as the timeliness of temperature control risk identification, alarm accuracy, and on-site management efficiency, by introducing a digital twin three-dimensional temperature field, door area scene-specific modeling, and risk-driven alarm and inspection mechanisms. This fully demonstrates its engineering practical value in complex refrigerated container supervision scenarios.

[0131] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for online monitoring of refrigerated containers based on digital twins, characterized in that, Includes the following steps: Collect the container number, company, cargo type, preset temperature control range, on-site time and location of the refrigerated container, generate on-site files and form a basic information database of refrigerated containers; Based on the basic information database of refrigerated containers, continuously monitor the data, construct a digital twin of refrigerated containers, and establish the three-dimensional temperature field digital twin status and operating condition status. A heat flow gating fusion unit and a time-series model for door access disturbance temperature field are configured in the digital twin of a refrigerated container. Door area scenario-specific training and loss design are performed on the door access status and external environmental conditions to generate three-dimensional temperature field prediction results and temperature control risk assessment results. The temperature control risk assessment results are linked with the on-site records. Alarms are generated according to alarm rules and role-based alarm push, such as temperature over-limit alarms, low power alarms, and equipment failure alarms. Alarm processing records and a list of items to be inspected are formed. Based on the list of items to be inspected, an inspection plan is developed and pushed to the mobile application system. On-site data verification and abnormal situation reporting are completed, and inspection records and abnormal handling results are generated and updated.

2. The online monitoring method for refrigerated containers based on digital twins according to claim 1, characterized in that, The generation of the refrigerated container basic information database specifically includes: Collect refrigerated container entry instructions, record container number, company, cargo type, preset temperature control range, entry time, on-site location and contact information, and generate a set of original basic information data for refrigerated containers; The original basic information data set of refrigerated containers is subjected to field integrity verification, content legality verification, and container number uniqueness verification. It is then converted into a standardized record set of basic information. Records that fail the verification are marked as abnormal records to be processed, and an on-site status identifier field is generated. The standardized record set of basic information is aggregated using the container number as the primary key. All standardized records of the same container number within one on-site period are integrated into an on-site file, and the corresponding on-site period number and on-site duration fields are generated. A basic information database for refrigerated containers was built based on on-site records; By linking the refrigerated container basic information database with the terminal layout data, spatial mapping records composed of container number, on-site location, on-site duration, and on-site status identifier are written into the refrigerated container basic information database, forming an on-site archive spatial mapping set.

3. The online monitoring method for refrigerated containers based on digital twins according to claim 1, characterized in that, The generation of the three-dimensional temperature field digital twin state and operating condition state specifically includes: Search the on-site archives and create a corresponding digital twin of the refrigerated container for each on-site archive, forming a set of digital twins of the refrigerated containers corresponding to the on-site refrigerated containers; Configure monitoring data access channels for the digital twin set of refrigerated containers, perform timestamp completion, container number matching, field format standardization and outlier identification, and convert them into a standardized set of monitoring data records; The standardized monitoring data record set is associated with the refrigerated container digital twin set according to the container number to obtain the three-dimensional temperature field digital twin status and operating condition status that are dynamically updated over time. Establish a state time series for the three-dimensional temperature field digital twin state and operating condition state in chronological order; The latest three-dimensional temperature field digital twin status and operating condition status are linked and written back to the on-site archives to form a set of status index records that can be retrieved by container number, on-site location, or company.

4. The online monitoring method for refrigerated containers based on digital twins according to claim 1, characterized in that, The generation of the temperature control risk assessment results specifically includes: Retrieve on-site records from the refrigerated container basic information database, and read the three-dimensional temperature field digital twin state time series and operating condition state time series from the refrigerated container digital twin set to generate a door area scene training sample set; In a digital twin of a refrigerated container, a heat flow gating fusion unit is instantiated to generate a heat flow gating feature sequence, which is then associated with the corresponding three-dimensional temperature field digital twin state time slice to form a heat flow gating input sequence. In the digital twin of a refrigerated container, a time-series model for the temperature field of access control disturbance is configured. The training sample set of the door area scene is divided into multiple door area scene subsets according to cargo type, preset temperature control range and external environmental conditions. Training data sets for the time-series model of the temperature field of access control disturbance are constructed respectively. For each subset of door area scenarios, a door area scenario-specific training and loss design scheme is set in the access control disturbance temperature field time series extrapolation model to obtain the converged access control disturbance temperature field time series extrapolation model parameters, which are then deployed to the corresponding refrigerated container digital twins. The current access control status, external environmental conditions and the latest three-dimensional temperature field digital twin status are obtained from online monitoring. The heat flow gating fusion unit generates a heat flow gating feature sequence and inputs it into the access control disturbance temperature field time series extrapolation model to obtain the three-dimensional temperature field prediction result time series. Based on the time series of three-dimensional temperature field prediction results and the preset temperature control range in the refrigerated container basic information database, the predicted temperature distribution of each future time interval is analyzed on a time-by-time basis to generate temperature control risk assessment results, which are then written into the digital twin of the refrigerated container and linked to the corresponding on-site file.

5. The online monitoring method for refrigerated containers based on digital twins according to claim 1, characterized in that, The generation of the alarm processing record and the list of items to be inspected specifically includes: Read the temperature control risk assessment results and retrieve the corresponding on-site files to generate a risk association record set, which serves as input for alarm determination; Each risk-related record in the set is processed, the temperature risk score is calculated, the temperature exceedance alarm level and the temperature exceedance alarm triggering identifier are determined according to the interval, and a set of candidate records for temperature exceedance alarms is generated. By combining the operating status in the digital twin of the refrigerated container, a threshold judgment is made on the power status field to generate a power risk index, which is then matched with the power alarm conditions to generate a set of candidate records for low power alarm. Perform fault mode matching on the equipment operating status field to generate equipment fault risk indicators, and match them with equipment fault alarm conditions to generate a set of equipment fault alarm candidate records; The candidate records for temperature over-limit alarms, low power alarms, and equipment malfunction alarms are merged into a total set of alarm candidate records. This set of alarm records is then compiled and supplemented with a summary of temperature control risk assessment results and a summary of on-site records. Read the alarm record set to generate a role-based alarm push task set, and push it to the monitoring terminal and mobile application system. Generate an initial alarm processing record for each alarm record, recording the alarm generation time, target role, push channel and initial processing status. During the alarm handling process, the system receives alarm confirmation operations, processing progress information, and processing result information returned by the monitoring terminal and mobile application system, updates the status of the alarm handling records, and forms a set of alarm handling records with handling status identifiers. Based on the set of alarm handling records marked as pending inspection, a pending inspection list is generated according to box number, location, alarm type, and alarm level, and the on-site files are associated.

6. The online monitoring method for refrigerated containers based on digital twins according to claim 1, characterized in that, The generation of the inspection records and anomaly handling results specifically includes: Read the list to be inspected and retrieve the on-site files in the refrigerated container basic information database. Combine the container number, on-site location, alarm type and alarm level to generate a set of inspection tasks. Form an inspection plan according to the regional order and time order and write it into the mobile application system to obtain the inspection plan distribution results. Load the on-site archives and digital twin summary information of the refrigerated container corresponding to the inspection task into the mobile application system, collect temperature, power indication, equipment operation indication and access control status on site, compare with the status and operating condition summary information of the three-dimensional temperature field digital twin, and generate the initial data of the inspection record. When abnormal temperature, abnormal power, or equipment failure is detected during the inspection, the mobile application system records the abnormality type, on-site reading, abnormality description, and image information to form an abnormality reporting record. This record is then linked with the initial data of the inspection record and the on-site archives to obtain an initial set of abnormality handling results. The initial data of the inspection records and the initial set of anomaly handling results are fed back to the digital twin of the refrigerated container and the basic information database of the refrigerated container. The status of the three-dimensional temperature field digital twin and the operating condition status are corrected and updated. The inspection records and anomaly handling result summaries are added to the on-site archives, and the alarm handling record set is updated to generate a new list to be inspected.

7. A digital twin-based online monitoring system for refrigerated containers, comprising the digital twin-based online monitoring method for refrigerated containers as described in any one of claims 1 to 6, characterized in that, include: The basic information management module is used to collect information such as refrigerated container number, company, cargo type, preset temperature control range, on-site time and on-site location, generate on-site files and form a basic information database for refrigerated containers. The digital twin monitoring module is used to access temperature, operating status, power consumption, and access control status monitoring data based on the refrigerated container basic information database, construct a digital twin of the refrigerated container, and establish a three-dimensional temperature field digital twin status and operating condition status. The door area scenario analysis module is used to configure the heat flow gate control fusion unit and the door access disturbance temperature field time series extrapolation model in the digital twin of the refrigerated container. It performs door area scenario-specific training and loss design on the door access status and external environmental conditions, and generates three-dimensional temperature field prediction results and temperature control risk assessment results. The alarm management module is used to associate temperature control risk assessment results with on-site records, configure alarm rules and role-based alarm push to generate alarms for excessive temperature, low power and equipment failure, and form alarm handling records and a list of items to be inspected. The inspection management and mobile application module is used to formulate inspection plans based on the list of items to be inspected and push them to the mobile application system, perform on-site data verification and abnormal situation reporting, and generate inspection records and abnormal handling results to update the refrigerated container basic information database and refrigerated container digital twin.

Images