Medical refrigerator intelligent monitoring system and method based on Internet of Things and cloud platform
Through an intelligent monitoring system combining the Internet of Things and cloud platform, real-time analysis of static and dynamic operation data of medical refrigerators is solved, and the problem that traditional manual inspections cannot detect abnormalities in a timely manner is realized, intelligent monitoring of medical refrigerators is achieved, and the safety and stability of medical items are ensured.
Patent Information
- Application Number
- CN202510682358.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-26
- Publication Date
- 2025-08-15
AI Technical Summary
Traditional medical refrigerator monitoring methods rely on manual regular inspections and cannot capture abnormal situations in real time, resulting in damage to medical items, resulting in waste of resources and potential risks.
Using intelligent monitoring methods based on the Internet of Things and cloud platforms, by receiving static and dynamic operating data of medical refrigerators, the first and second analysis models are used to judge static and dynamic abnormalities respectively, and output real-time or delayed prompt information.
It realizes timely alarms in abnormal operating state of medical refrigerators, ensures the quality stability of medical items, reduces the probability of false alarms in dynamic states, and reduces interference to medical work.
Smart Images
Figure CN120488615A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of refrigerators, and in particular to an intelligent monitoring system and method for medical refrigerators based on the Internet of Things and a cloud platform. Background Art
[0002] In modern healthcare, medical refrigerators are critical equipment for storing temperature-sensitive medical supplies, such as medicines, vaccines, and biologics. The stability of these refrigerators' operating conditions is directly linked to the quality and effectiveness of these medical supplies. Accurate temperature control plays a crucial role in maintaining the activity, efficacy, and safety of these medical supplies.
[0003] Traditional monitoring methods for medical refrigerators rely on regular manual inspections and the recording of temperature data. However, these inspections are often time-consuming and fail to capture operational anomalies in real time. For example, refrigeration system failures and unstable power supplies are difficult to detect in a timely manner, potentially damaging stored medical supplies and resulting in significant waste of medical resources and potential medical risks.
[0004] Therefore, how to achieve real-time, accurate and intelligent monitoring of medical refrigerators is a technical problem that needs to be solved urgently. Summary of the Invention
[0005] In response to the above technical problems, the present invention provides a medical refrigerator intelligent monitoring method, system, electronic equipment, computer storage medium and computer program product based on the Internet of Things and cloud platform to improve the efficiency and reliability of medical storage management, and to ensure medical safety and improve the quality of medical services.
[0006] The present invention discloses an intelligent monitoring method for medical refrigerators based on the Internet of Things and a cloud platform. The method comprises the following steps: receiving first operating data of the medical refrigerator in a static state and second operating data of the medical refrigerator in a dynamic state uploaded by an Internet of Things device; wherein, compared with the first operating data, the second operating data further includes mobile environment data; judging whether the medical refrigerator has a static abnormality based on the first operating data, and judging whether the medical refrigerator has a dynamic abnormality based on the second operating data; if a static abnormality exists, immediately outputting a first prompt message; if a dynamic abnormality exists, outputting a second prompt message after a first preset time period.
[0007] Optionally, receiving the second operating data of the medical refrigerator in a dynamic state uploaded by the Internet of Things device includes: determining whether the first operating data contains a signal that the mains power supply mode is cut off, and if so, generating and sending an additional signal to the medical refrigerator; wherein the additional signal is used to trigger the medical refrigerator to upload mobile environment data in addition to the first operating data; receiving the second operating data of the medical refrigerator in a dynamic state uploaded by the Internet of Things device.
[0008] Optionally, the generating and sending of an additional signal to the medical refrigerator includes: obtaining maintenance history data of the medical refrigerator, the maintenance history data including the duration between when the AC power supply mode is cut off and when it is restarted, and determining the shortest duration therein as the second preset duration; obtaining the type attributes of the medical items currently stored in the medical refrigerator, obtaining an additional adjustment duration based on the type attribute matching, and calculating the second preset duration minus the additional adjustment duration to obtain a third preset duration; upon detecting that the first operating data includes a signal that the AC power supply mode is cut off, and the accumulated duration reaches the third preset duration, generating and sending the additional signal to the medical refrigerator.
[0009] Optionally, the determining whether the medical refrigerator has a dynamic abnormality based on the second operating data includes: parsing the second operating data into basic data and the mobile environment data, the basic data and the first operating data containing the same data content; performing an abnormality analysis on the basic data using a first analysis model, and determining whether the medical refrigerator has a first dynamic abnormality based on the analysis results; performing a vibration analysis on the mobile environment data using a second analysis model, and if the analysis results indicate that the probability that the displacement of a specified component of the medical refrigerator exceeds an allowable range is higher than a probability threshold, determining that the medical refrigerator has a second dynamic abnormality.
[0010] Optionally, outputting a second prompt message after the first preset time includes: analyzing the running route of the medical refrigerator according to the mobile environment data, if the analysis shows that the running route is a departure route, setting the first preset time to the first time; if the analysis shows that the running route is a return route, setting the first preset time to the second time; the first time is shorter than the second time; if the dynamic abnormality still exists after the first preset time, outputting the second prompt message.
[0011] The present invention also discloses an intelligent monitoring system for medical refrigerators based on the Internet of Things and a cloud platform. The system includes a processing device and a storage device. The computer code stored in the storage device is called and executed by the processing device to implement the following steps: receiving first operating data of the medical refrigerator in a static state and second operating data of the medical refrigerator in a dynamic state uploaded by an Internet of Things device; wherein, compared with the first operating data, the second operating data also includes mobile environment data; judging whether the medical refrigerator has a static abnormality based on the first operating data, and judging whether the medical refrigerator has a dynamic abnormality based on the second operating data; if a static abnormality exists, outputting a first prompt message immediately; if a dynamic abnormality exists, outputting a second prompt message after a first preset time period.
[0012] The present invention also discloses an electronic device, comprising: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement any of the above methods.
[0013] The present invention further discloses a computer storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement any of the above methods.
[0014] The present invention further discloses a computer program product, which includes computer code. When the computer code is executed by a processor of an electronic device, any of the above methods is implemented.
[0015] The present invention has at least one beneficial effect: By comprehensively collecting and analyzing operational data in both static and dynamic states, the above-mentioned solution provides timely warnings of abnormal refrigerator operation, effectively ensuring the quality and stability of medical supplies stored within hospitals. Furthermore, the present invention adjusts the duration of dynamic alarms, thereby reducing the probability of false alarms and minimizing the disruption to medical work caused by unnecessary alarms. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.
[0017] Figure 1 It is a schematic diagram of the planar knife marks existing on the automotive accessories disclosed in the embodiment of the present invention.
[0018] Figure 2 This is a schematic diagram of determining whether a medical refrigerator has a dynamic abnormality based on second operating data according to an embodiment of the present invention.
[0019] Figure 3 This is a structural diagram of an intelligent monitoring system for medical refrigerators based on the Internet of Things and a cloud platform disclosed in an embodiment of the present invention. DETAILED DESCRIPTION
[0020] The following specific embodiments illustrate the implementation of this application. Those familiar with the art can easily understand the other advantages and functions of this application from the content disclosed in this specification. Obviously, the described embodiments are part of the embodiments of this application, but not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.
[0021] In addition, the technical features involved in the different embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.
[0022] like Figure 1 As shown, in response to the above technical problems, an embodiment of the present invention discloses an intelligent monitoring method for medical refrigerators based on the Internet of Things and a cloud platform, the method comprising the following steps: S100, receiving first operating data of the medical refrigerator in a static state and second operating data of the medical refrigerator in a dynamic state uploaded by an Internet of Things device; wherein, compared with the first operating data, the second operating data also includes mobile environment data.
[0023] In this step, the medical refrigerator is equipped with an IoT device, which connects it to a cloud platform. This allows the refrigerator to transmit its operating data to the cloud platform, which then determines whether the refrigerator is experiencing any anomalies and issues an alarm. The primary operating data primarily includes, but is not limited to, information reflecting the refrigerator's operating status under stable conditions, such as its internal temperature, refrigeration system pressure, and power consumption. The secondary operating data may also include positioning information and acceleration data reflecting its motion status.
[0024] The medical refrigerator of the present invention is a portable, mobile refrigerator that can be used routinely at a fixed location within a hospital (stationary use), typically powered by mains electricity. Alternatively, it can be temporarily transferred to an ambulance for emergency medical missions (mobile use), powered by a battery or unpowered (in which case the refrigerator becomes an insulated box) during transfer, and then connected to the ambulance's power supply after transfer.
[0025] When the refrigerator is in a static state, the IoT device collects and uploads primary operating data. When the refrigerator enters a dynamic state, such as when being transported to an ambulance, the IoT device collects and uploads secondary operating data. This data, in addition to the basic static data, also includes data related to the mobile environment, such as the impact of vehicle vibration on the stability of refrigerator components and the rapid changes in ambient temperature during movement. The cloud platform receives this operational data uploaded by the IoT devices in different states and prepares for subsequent analysis and judgment.
[0026] S200: Determine whether the medical refrigerator has a static abnormality based on the first operating data, and determine whether the medical refrigerator has a dynamic abnormality based on the second operating data.
[0027] In this step, after receiving the aforementioned operating status data, the cloud platform analyzes the data according to pre-defined judgment rules. The cloud platform compares the first operating data with a pre-defined range of normal operating data under static conditions to determine whether the medical refrigerator has static anomalies. For example, if the internal temperature of the refrigerator exceeds the temperature range set for storing medicines, vaccines, etc. during fixed-term hospital use, or if the refrigeration system pressure is outside the normal operating range, the cloud platform will determine that a static anomaly exists.
[0028] For the second operating data, due to the complex and changing environment of the refrigerator in a dynamic state, its normal operating data range differs from that in a static state. The cloud platform analyzes and judges data based on the normal data standards specifically set for dynamic conditions. For example, if vibrations during the ambulance's operation cause internal components to move beyond the allowable range, or if the ambient temperature fluctuates rapidly, the refrigerator may be unable to maintain an acceptable internal temperature range at a near future time, resulting in a dynamic anomaly.
[0029] S300: If a static abnormality exists, a first prompt message is output immediately; if a dynamic abnormality exists, a second prompt message is output after a first preset time period.
[0030] In this step, if the cloud platform determines a static anomaly exists, it immediately generates and outputs the first prompt message. This is because, in fixed hospital usage scenarios, medical refrigerators store medical supplies under stringent temperature and other conditions. Any anomaly could quickly affect the quality of the items, necessitating immediate notification to relevant medical staff or equipment management personnel so they can take timely action, such as inspecting the refrigerator for malfunctions or relocating stored items.
[0031] If a dynamic anomaly is determined, the cloud platform will output a second prompt message after the first preset time. This is because in dynamic scenarios such as ambulances, the vehicle is performing emergency rescue missions, and medical staff are focused on treatment. In addition, some data fluctuations in dynamic environments may be temporary and recoverable. The purpose of setting the first preset time is to give the refrigerator a certain amount of time to self-adjust and adapt to the dynamic environment (during this process, the temperature environment in the medical refrigerator remains in a reasonable range). If the anomaly has not been eliminated after this time, the prompt message will be output to avoid excessive interference information due to short-term fluctuations. At the same time, it can also ensure that medical staff are reminded in time when the anomaly persists to ensure the safety of medical supplies.
[0032] The above-mentioned solution of the present invention comprehensively collects and analyzes operating data in both static and dynamic states of the medical refrigerator, enabling timely alerts for abnormal operating conditions of the refrigerator, effectively ensuring the quality and stability of medical supplies stored within the hospital. Furthermore, the present invention adjusts the duration of dynamic alarms, thereby reducing the probability of false alarms and minimizing the disruption to medical work caused by unnecessary alarms.
[0033] It is understandable that the above-mentioned solution of the present invention can be applied to a cloud platform, or to a management terminal connected to a cloud platform. The management terminal can be a terminal of a medical institution or a management structure of a medical institution, and is not specifically limited.
[0034] Optionally, receiving the second operating data of the medical refrigerator in a dynamic state uploaded by the Internet of Things device includes: determining whether the first operating data contains a signal that the mains power supply mode is cut off, and if so, generating and sending an additional signal to the medical refrigerator; wherein the additional signal is used to trigger the medical refrigerator to upload mobile environment data in addition to the first operating data; receiving the second operating data of the medical refrigerator in a dynamic state uploaded by the Internet of Things device.
[0035] In this embodiment, when the cloud platform receives data about a medical refrigerator in its dynamic state, it determines whether the first operating data contains a signal indicating that the mains power supply mode has been disconnected. Because medical refrigerators rely on mains power in their static state and switch to battery power or no power (for use as incubators) in their dynamic state, the mains power disconnection signal becomes a key state transition indicator. When the cloud platform detects this signal, it indicates that the medical refrigerator has likely transitioned from its static state within the hospital to a mobile state, such as when a staff member disconnects the refrigerator from the mains power and begins carrying it to the ambulance. During this phase, the refrigerator's operating environment is about to undergo complex changes.
[0036] Next, if the cloud platform detects a loss of utility power, it generates and sends an additional signal to the refrigerator, specifically triggering it to upload its mobile environment data in addition to the primary operating data. Thus, in a dynamic state, the cloud platform can rely on both the primary operating data (basic static data such as internal temperature and refrigeration system pressure) and the mobile environment data to determine if the refrigerator is experiencing an anomaly and issue an alert.
[0037] It is understandable that dynamic anomalies actually also include the aforementioned static anomalies of medical refrigerators. In addition, they also include anomalies caused by environmental problems such as the displacement of internal components of the refrigerator exceeding the allowable range due to vibration, or the rapid change of ambient temperature and the prediction that the refrigerator will not be able to maintain the internal temperature within a reasonable range after reaching a near future time.
[0038] Optionally, the generating and sending of an additional signal to the medical refrigerator includes: obtaining maintenance history data of the medical refrigerator, the maintenance history data including the duration between when the AC power supply mode is cut off and when it is restarted, and determining the shortest duration therein as the second preset duration; obtaining the type attributes of the medical items currently stored in the medical refrigerator, obtaining an additional adjustment duration based on the type attribute matching, and calculating the second preset duration minus the additional adjustment duration to obtain a third preset duration; upon detecting that the first operating data includes a signal that the AC power supply mode is cut off, and the accumulated duration reaches the third preset duration, generating and sending the additional signal to the medical refrigerator.
[0039] In this embodiment, the medical refrigerator may need to be disconnected from the mains power supply during normal use, such as for maintenance, periodic debugging, and other operations. In order to prevent the cloud platform from mistaking the signal of the mains power supply mode being disconnected as the medical refrigerator switching from static use to mobile use, it is necessary to distinguish this. Specifically: the cloud platform first obtains the maintenance history data of the medical refrigerator. The maintenance history data records the duration of the medical refrigerator from the time the mains power supply mode was disconnected to the time it was restarted, that is, the duration of the mains power supply being disconnected. The shortest duration among these represents the fastest case in which the medical refrigerator has restored power after a mains power outage in the past, and is used as the second preset duration.
[0040] The cloud platform also obtains the type attributes of the medical supplies currently stored in the medical refrigerator. This information is also transmitted to the cloud platform by the medical refrigerator via an IoT device. Because different types of medical supplies have varying tolerances to environmental conditions such as temperature, an additional adjustment time is determined based on these type attributes. For example, human albumin, ceftriaxone sodium for injection, and rabies vaccine are highly sensitive to temperature fluctuations (i.e., have a narrower suitable storage temperature range), requiring a relatively longer additional adjustment time. Thus, when the cumulative duration of mains power disconnection reaches a relatively shorter third preset duration (third preset duration = second preset duration - additional adjustment time), the medical refrigerator is immediately determined to be transitioning from static to mobile use, initiating subsequent acquisition of the second operating data and analysis of anomalies.
[0041] If, however, the medicines stored are more tolerant to temperature fluctuations (i.e., suitable for a wider storage temperature range), the additional adjustment period will be slightly shorter. Thus, only when the cumulative duration of mains power disconnection reaches a relatively longer third preset duration is it determined that the medical refrigerator is about to transition from static to mobile use, and the subsequent acquisition of the second operating data and abnormality analysis is initiated.
[0042] Therefore, the present invention uses the maintenance history data of the medical refrigerator to determine the basic waiting time, that is, the second preset time. At the same time, it also comprehensively considers the type attributes of the medical items stored in the medical refrigerator, and then determines a more reasonable and accurate third preset time, thereby achieving a more accurate judgment on whether the medical refrigerator is about to switch from a static state to a dynamic state.
[0043] Optionally, the determining whether the medical refrigerator has a dynamic abnormality based on the second operating data includes: parsing the second operating data into basic data and the mobile environment data, the basic data and the first operating data containing the same data content; performing an abnormality analysis on the basic data using a first analysis model, and determining whether the medical refrigerator has a first dynamic abnormality based on the analysis results; performing a vibration analysis on the mobile environment data using a second analysis model, and if the analysis results indicate that the probability that the displacement of a specified component of the medical refrigerator exceeds an allowable range is higher than a probability threshold, determining that the medical refrigerator has a second dynamic abnormality.
[0044] In this embodiment, after receiving the second operational data uploaded by the medical refrigerator in a dynamic state, the cloud platform first parses it into basic data and mobile environment data. The basic data and the first operational data (i.e., the internal temperature, refrigeration system pressure, and power consumption of the medical refrigerator detected in the dynamic state) contain the same data content as the first operational data (i.e., the internal temperature, refrigeration system pressure, and power consumption of the medical refrigerator detected in the dynamic state. It should be understood that the dynamic state refers to the time the medical refrigerator is not in use at a fixed location, i.e., the time when it is transferred from a fixed location to an ambulance, the time when it is in use by an ambulance, and the time when it is transferred from an ambulance to a fixed location). These data reflect the refrigerator's basic operating status, such as the internal temperature, refrigeration system pressure, and power consumption, the only difference being that this data is acquired in the dynamic state. The mobile environment data, on the other hand, is data specific to the dynamic state, such as positioning information reflecting the refrigerator's motion and acceleration data (which can be used to analyze the intensity of vibration transmitted to the medical refrigerator from vehicle movement).
[0045] like Figure 2 As shown, the cloud platform uses a first analysis model to perform anomaly analysis on the basic data. The first analysis model is pre-set based on the normal range and variation patterns of the basic data of the medical refrigerator in its dynamic state. The first analysis model can also be used to perform anomaly analysis on the first operating data. By comparing the actual received basic data with the standards in the first analysis model, the analysis results determine whether the medical refrigerator has a first dynamic anomaly. For example, if, in the dynamic state, the internal temperature of the medical refrigerator exceeds the temperature range set for storing medicines, vaccines, etc. in that state, or if the refrigeration system pressure is outside the normal operating pressure range, the first analysis model will determine the presence of a first dynamic anomaly.
[0046] The cloud platform uses a second analysis model to perform vibration analysis on the mobile environment data. This model analyzes the vibration data generated by the movement of personnel or the operation of an ambulance in a medical refrigerator's dynamic state and predicts the probability of component displacement. If the second analysis model indicates that the probability of a specified component displacement exceeding the allowable range exceeds a threshold, such as shifting shelves storing medical supplies or opening the refrigerator door due to vibration (poor seal), the refrigerator is considered to have a second dynamic anomaly.
[0047] Among them, the first analysis model can adopt conventional classification models, such as SVM, random forest, etc., while the second analysis model is preferably constructed using Transformer, Resnet (such as Resnet-50, ResNet-101, ResNet-152, etc.), and is fully trained using training data constructed based on actual vibration data of the same medical refrigerator at different amplitudes and corresponding component displacement conditions. The specific training process will not be repeated here.
[0048] Taking the SVM as an example, the structure of the first analysis model is as follows: 1. Data input layer: Input data: Basic data from medical refrigerators, including real-time monitoring data such as temperature, humidity, voltage, current, operating hours, and number of door openings. It also includes the mains power supply status (e.g., whether it is disconnected) and the operating status of the backup power supply (e.g., battery).
[0049] 2. Feature Engineering Layer: Time Series Feature Extraction: For time series data (such as temperature changes over time), extract statistical features within the sliding window: mean, variance, maximum, minimum, range (reflecting the range of data fluctuations); slope (trend change), kurtosis / skewness (distribution shape); difference between adjacent moments (rate of change).
[0050] State feature encoding: One-hot encoding is performed on discrete state data (such as the power supply mode "mains power / backup power") and converted into a numerical vector that can be recognized by the model.
[0051] 3. SVM core classification layer: Model: For a binary classification problem (determining whether the first dynamic anomaly exists), a binary SVM is used. If it is necessary to distinguish the types of anomalies (such as temperature anomalies and power supply anomalies), it can be expanded to a multi-classification SVM (such as a One-vs-One or One-vs-Rest strategy).
[0052] Kernel function: For data with a linear correlation, such as voltage anomalies and temperature anomalies, a linear kernel function is used. For data with a nonlinear correlation, such as temperature changes and door opening times and operating time, a nonlinear kernel function (such as RBF kernel and polynomial kernel) is used.
[0053] Decision boundary construction: By maximizing the margin between samples, we find the optimal hyperplane in the feature space to separate normal data from abnormal data. For nonlinear problems, we use kernel functions to map low-dimensional features to high-dimensional space to achieve linear separability.
[0054] 4. Abnormality judgment logic layer: The binary classification model outputs a probability value or category label (for example, "0" indicates normal, "1" indicates the presence of the first dynamic abnormality). If the output probability is higher than the preset threshold (such as 0.8), it is judged as abnormal.
[0055] If it is a multi-classification model, it can be further mapped to specific anomaly types: Label "1": Temperature anomaly (for example, outside the range of 2-8°C). Label "2": Power supply anomaly (for example, the mains power is cut off and the backup power supply is not activated). Label "3": Operating parameter anomaly (for example, an abnormal increase in the number of door openings causing humidity fluctuations).
[0056] An example workflow for the first analysis model is as follows: Basic data such as the temperature (25°C), humidity (70%), mains power status (off), and backup power status (not activated) of a medical refrigerator are received. The 5-minute sliding window mean (24°C), variance (3°C²), and rate of change of humidity (+5% / minute) are calculated. The power status is one-hot encoded (mains power off = 1, backup power not activated = 1). SVM classification: The model uses an RBF kernel function to map features to a high-dimensional space and determine whether the sample is located in an abnormal region. If the output probability is 0.9 (above the threshold of 0.8), the temperature rise is determined to be caused by a power supply anomaly (the first dynamic anomaly). An abnormality alert is generated, prompting medical staff to check the backup power supply or temperature control system.
[0057] Taking the Transformer as an example, the structure of the second analysis model is as follows: 1. Data Input Layer: Input data: Mobile environment data from a medical refrigerator, primarily time series data collected by a vibration sensor. This includes: Three-axis acceleration data (x, y, and z axes): reflects the intensity and direction of vibration during transportation (unit: m / s²). Angular velocity data (rotation rate around the x, y, and z axes): monitors the magnitude of device flipping or shaking. Timestamp: Sampling time accurate to the millisecond level, used for time series alignment.
[0058] 2. Position encoding layer: Sine-cosine encoding can be used. A position embedding vector is generated through training, added to the feature vector of the vibration data, and then input into the Transformer encoder module.
[0059] 3. Transformer Encoder Module: Multi-Head Self-Attention Layer: Captures the dependencies between different time steps in the vibration sequence (for example, whether a strong vibration at one moment is related to the continuous shaking at the previous and next moments).
[0060] Feed-Forward Network: Performs nonlinear transformation on the features output by the self-attention layer to capture more complex vibration pattern features.
[0061] Layer Normalization and Residual Connections: Layer Normalization normalizes the feature dimensions of each sample to stabilize the training process. Residual Connections alleviate the vanishing gradient problem in deep networks and ensure cross-layer information flow.
[0062] 4. Feature fusion and classification layer: Global feature extraction: Pooling operations (such as mean pooling and maximum pooling) are performed on all time step features output by the encoder to generate a global vibration feature vector of fixed length.
[0063] Classification head: The global features are mapped to the probability value of component displacement exceeding the range through the fully connected layer. If the probability is higher than the probability threshold (for example, 0.7), it is determined that the second dynamic anomaly exists (the displacement of the specified component exceeds the range).
[0064] An example workflow for the second analysis model is as follows: A three-axis acceleration sequence (e.g., x-axis: [-0.5, 1.2, 3.8, ...] m / s²) of a medical refrigerator being transported by an ambulance is received. The sequence length is T = 1000 (1 second of data). A sine-cosine encoding vector containing temporal information is generated for each time step and added to the acceleration data. A multi-head self-attention layer detects a correlation between the strong vibration at time steps 500-600 (the moment of emergency braking) and the sustained fluctuations in the preceding and following moments. The feedforward network further extracts the characteristic pattern of "short-term strong vibration + sustained high-frequency shaking." Global pooling is performed to generate a feature vector, which, after passing through a fully connected layer, outputs a probability of 0.85 (higher than the threshold of 0.7), indicating a risk of compressor component displacement exceeding the range (a second dynamic anomaly).
[0065] Optionally, outputting a second prompt message after the first preset time includes: analyzing the running route of the medical refrigerator according to the mobile environment data, if the analysis shows that the running route is a departure route, setting the first preset time to the first time; if the analysis shows that the running route is a return route, setting the first preset time to the second time; the first time is shorter than the second time; if the dynamic abnormality still exists after the first preset time, outputting the second prompt message.
[0066] In this embodiment, after determining that a medical refrigerator has a dynamic anomaly, the cloud platform waits for a first preset time period. If the first preset time period has passed, the cloud platform reanalyzes the newly acquired second operating data for a dynamic anomaly. If the analysis indicates that the dynamic anomaly still exists, the cloud platform outputs a second prompt message. This reduces the distraction caused by excessive prompts to medical staff, ensuring that they can focus on treatment.
[0067] At the same time, the present invention also sets the first preset duration to be the first duration or the second duration. Specifically, the positioning information of the medical refrigerator can be analyzed from the mobile environment data, and the actual operating route of the medical refrigerator can be mapped based on the positioning information. Since the location of the hospital is known, it can be determined whether the ambulance where the medical refrigerator is located is currently on a departure route (for example, the actual operating route indicates that the ambulance is extending in a direction away from the hospital, and the actual operating route is determined to be the route of the ambulance from the hospital to the rescue site) or a return route (the actual operating route indicates that the ambulance is extending in a direction close to the hospital, and the actual operating route is determined to be the route of the ambulance from the rescue site to the hospital after completing the rescue mission).
[0068] If the analysis shows that the running route is the departure route, the first preset duration is set to a shorter first duration. This is because the medical staff did not provide treatment during the process of setting off for the rescue site. In this case, the waiting time for the dynamic abnormality reminder can be appropriately shortened. This does not interfere with the medical staff and can still remind them to deal with the dynamic abnormality of the medical refrigerator in a timely manner.
[0069] If the analysis indicates that the route is a return route, the first preset duration is set to a longer second duration. This indicates that an emergency mission is generally ongoing, i.e., medical personnel are treating the patient in the ambulance. In this case, setting the first preset duration to a longer second duration can appropriately reduce interference from medical personnel.
[0070] By setting the first preset time differently according to the operation route, the intelligent monitoring method of medical refrigerators based on the Internet of Things and cloud platform becomes more flexible and practical, and can better adapt to the complex situations of ambulances in different operation stages, thereby ensuring the safety of medical supplies in medical refrigerators and the smooth implementation of emergency work.
[0071] It is understood that neither the first nor the second time duration should be too long, otherwise medical staff may not be aware of any abnormal conditions in the medical refrigerator for an extended period, potentially causing damage to the medical supplies inside due to high temperatures, collisions, or squeezing. Furthermore, the medical refrigerator should be equipped with a reminder that emits a corresponding sound and / or light when the refrigerator receives the first or second reminder information sent by the cloud platform.
[0072] like Figure 3 As shown, an embodiment of the present invention further discloses an intelligent monitoring system for medical refrigerators based on the Internet of Things and a cloud platform. The system includes a processing device and a storage device. The computer code stored in the storage device is called and executed by the processing device to implement the following steps: receiving first operating data of the medical refrigerator in a static state and second operating data of the medical refrigerator in a dynamic state uploaded by an Internet of Things device; wherein, compared with the first operating data, the second operating data also includes mobile environment data.
[0073] Based on the first operating data, it is determined whether the medical refrigerator has a static abnormality, and based on the second operating data, it is determined whether the medical refrigerator has a dynamic abnormality; if a static abnormality exists, a first prompt message is output immediately; if a dynamic abnormality exists, a second prompt message is output after a first preset time period.
[0074] An embodiment of the present invention further discloses an electronic device, comprising: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement the method described in the aforementioned embodiment.
[0075] An embodiment of the present invention further discloses a computer storage medium, wherein the computer storage medium stores a computer program, and the computer program is executed by a processor to implement the method described in the above embodiment.
[0076] An embodiment of the present invention further discloses a computer program product, which includes computer code. When the computer code is executed by a processor of an electronic device, the method described in the above embodiment is implemented.
[0077] The computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or apparatuses, or any suitable combination thereof. Alternatively, the computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media may include electrical connections based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.
[0078] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in the present invention can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved. This is not limited herein.
[0079] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.
Claims
1. An intelligent monitoring method for medical refrigerators based on the Internet of Things and a cloud platform, characterized by: The method includes the following steps: receiving first operating data of a medical refrigerator in a static state and second operating data of a medical refrigerator in a dynamic state uploaded by an Internet of Things device; wherein, compared with the first operating data, the second operating data also includes mobile environment data; judging whether the medical refrigerator has a static abnormality based on the first operating data, and judging whether the medical refrigerator has a dynamic abnormality based on the second operating data; if a static abnormality exists, immediately outputting a first prompt message; if a dynamic abnormality exists, outputting a second prompt message after a first preset time period.
2. The intelligent monitoring method for medical refrigerators based on the Internet of Things and a cloud platform according to claim 1, characterized in that: Receiving the second operating data of the medical refrigerator in a dynamic state uploaded by the Internet of Things device, including: determining whether the first operating data contains a signal that the mains power supply mode is cut off, and if so, generating and sending an additional signal to the medical refrigerator; wherein the additional signal is used to trigger the medical refrigerator to upload mobile environment data in addition to the first operating data; receiving the second operating data of the medical refrigerator in a dynamic state uploaded by the Internet of Things device.
3. The intelligent monitoring method for medical refrigerators based on the Internet of Things and a cloud platform according to claim 2, characterized in that: The generating and sending of the additional signal to the medical refrigerator includes: obtaining maintenance history data of the medical refrigerator, the maintenance history data including the duration between when the mains power supply mode is cut off and when it is restarted, and determining the shortest duration therein as the second preset duration; obtaining the type attributes of the medical items currently stored in the medical refrigerator, obtaining an additional adjustment duration based on the type attribute matching, and calculating the second preset duration minus the additional adjustment duration to obtain a third preset duration; and generating and sending the additional signal to the medical refrigerator when it is detected that the first operating data includes a signal that the mains power supply mode is cut off and the accumulated duration reaches the third preset duration.
4. The intelligent monitoring method for medical refrigerators based on the Internet of Things and a cloud platform according to claim 1, characterized in that: The determining whether a medical refrigerator has a dynamic abnormality based on the second operating data includes: parsing the second operating data into basic data and the mobile environment data, the basic data and the first operating data containing the same data content; performing an abnormality analysis on the basic data using a first analysis model, and determining whether the medical refrigerator has a first dynamic abnormality based on the analysis result; and performing a vibration analysis on the mobile environment data using a second analysis model, and if the analysis result indicates that the probability that the displacement of a specified component of the medical refrigerator exceeds an allowable range is higher than a probability threshold, determining that the medical refrigerator has a second dynamic abnormality.
5. The intelligent monitoring method for medical refrigerators based on the Internet of Things and a cloud platform according to claim 4, characterized in that: Outputting a second prompt message after the first preset time includes: analyzing the running route of the medical refrigerator according to the mobile environment data, and if the analysis shows that the running route is a departure route, setting the first preset time to the first time; if the analysis shows that the running route is a return route, setting the first preset time to the second time; the first time is shorter than the second time; if the dynamic abnormality still exists after the first preset time, outputting the second prompt message.
6. An intelligent monitoring system for medical refrigerators based on the Internet of Things and a cloud platform, comprising a processing device and a storage device, characterized in that: The computer code stored in the storage device is called and executed by the processing device to implement the following steps: receiving the first operating data of the medical refrigerator in a static state and the second operating data of the medical refrigerator in a dynamic state uploaded by the Internet of Things device; wherein, compared with the first operating data, the second operating data also includes mobile environment data; judging whether the medical refrigerator has a static abnormality based on the first operating data, and judging whether the medical refrigerator has a dynamic abnormality based on the second operating data; if a static abnormality exists, outputting a first prompt message immediately; if a dynamic abnormality exists, outputting a second prompt message after a first preset time period.
7. The intelligent monitoring system for medical refrigerators based on the Internet of Things and a cloud platform according to claim 6, characterized in that: Receiving the second operating data of the medical refrigerator in a dynamic state uploaded by the Internet of Things device, including: determining whether the first operating data contains a signal that the mains power supply mode is cut off, and if so, generating and sending an additional signal to the medical refrigerator; wherein the additional signal is used to trigger the medical refrigerator to upload mobile environment data in addition to the first operating data; receiving the second operating data of the medical refrigerator in a dynamic state uploaded by the Internet of Things device.
8. An electronic device comprising: At least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement the method according to any one of claims 1 to 5.
9. A computer storage medium storing a computer program, wherein: The computer program is executed by a processor to implement the method according to any one of claims 1 to 5.
10. A computer program product, characterized in that: The computer program product includes computer code, and when the computer code is executed by a processor of an electronic device, the method according to any one of claims 1 to 5 is implemented.