Refrigerator key device state early warning method, detection early warning system and equipment

Abstract

The invention discloses a refrigerator key device state early warning method, a detection early warning system and equipment, and the method comprises the steps: synchronously collecting the voltage, current, active power and reactive power of a medical refrigerator power supply loop at high frequency, and carrying out the data preprocessing through a power supply fluctuation compensation algorithm; establishing a load curve reference of a medical exclusive working condition, and extracting typical load characteristic frequencies of a compressor, a fan and a defrosting heater under each working condition; based on an improved independent component analysis algorithm, the whole machine load curve is disassembled into a compressor sub-curve, a fan sub-curve and a defrosting heater sub-curve; and the compressor sub-curve, the fan sub-curve and the defrosting heater sub-curve correspondingly give out state early warning according to a grading early warning mechanism. Through power supply side multi-parameter acquisition and load curve analysis, early abnormality identification of devices such as a compressor and a fan is realized, meanwhile, medical cold chain data tracing and supervision linkage requirements are met, and medical material storage safety is guaranteed.

Description

Technical Field

[0001] This invention relates to the field of data monitoring and management technology for medical cold chain equipment, specifically to a method, detection and early warning system, and equipment for early warning of the status of key components in refrigerators. It is applicable to medical refrigerators such as medical refrigerators, medical freezers, vaccine refrigerators, and low-temperature storage boxes for biological samples. It enables early identification and graded early warning of anomalies in the operating status of their core components (compressor, condenser fan, evaporator fan, LED lighting, defrost heater, etc.), while meeting the compliance requirements of medical cold chain supervision and extending to the entire life cycle operation and maintenance field of medical cold chain equipment. Background Technology

[0002] As a core terminal device in the medical cold chain system, medical refrigerators are responsible for storing special medical supplies such as vaccines, biological samples, diagnostic reagents, and blood products. The operational stability of their key components directly determines the safety and effectiveness of the stored medical supplies. With the establishment of standardized rules for the management and operation of drug quality, vaccine storage, and transportation, existing medical refrigerators not only need to maintain a precise temperature control environment, but also need to have the ability to monitor component status, provide early warning of faults, and ensure data traceability.

[0003] Current solutions for monitoring the condition of key components in medical refrigerators still have many industry-specific limitations: First, dedicated sensors are susceptible to failure due to the medical environment. Traditional solutions rely on temperature sensors to monitor the compressor housing temperature and speed sensors to monitor the fan operation. However, medical refrigerators require regular disinfection, sterilization, and cryogenic maintenance. High concentrations of disinfectants and ultra-low temperatures of -80°C can easily cause corrosion of sensor probes and freezing cracks of components, leading to distorted monitoring data. A large number of medical refrigerator monitoring failures are due to sensor damage. Second, the monitoring dimensions are disconnected from medical operating conditions. Medical refrigerators have special operating conditions such as vaccine storage, cold chain verification, and power outage backup. Traditional solutions only monitor a single physical quantity and cannot correlate the coupling relationship between operating conditions and device load. For example, during the cold chain verification stage, the compressor needs to run at a high load continuously. The load fluctuations caused by early bearing wear are easily misjudged as normal operating condition fluctuations, missing the early warning opportunity. Third, the early warning is delayed and does not meet the requirements of medical supervision. Existing technologies mostly trigger alarms after the temperature inside the box exceeds the threshold due to device failure. At this time, medical supplies are already at risk of deterioration. Furthermore, the linkage between device status data and the cold chain supervision platform has not been achieved, making it difficult to meet the GSP's compliance requirements for "full-process risk early warning". Fourth, it cannot distinguish multi-component coordinated failures. The refrigeration system of a medical refrigerator is a closed-loop system in which the compressor, fan, and defrost heater work together. Traditional isolated monitoring modes cannot identify coordinated failures such as "fan jamming causing compressor overload", which can easily lead to difficulties in tracing the source of the failure.

[0004] Furthermore, the unique scenarios in the medical cold chain field exacerbate the difficulty of monitoring: portable medical refrigerators at vaccination sites need to be moved by vehicle, and power voltage fluctuations can easily mask device malfunctions; ultra-low temperature freezers in biobanks have long operating cycles, and the gradual load changes caused by device aging are difficult to identify using traditional threshold monitoring. In practice, many vaccine storage accidents are related to the failure to detect early malfunctions of medical refrigerator compressors in a timely manner, leading to vaccine inactivation; in biobanks used in biological research, temperature control failure caused by fan jamming is also a significant cause of sample loss.

[0005] Meanwhile, existing solutions lack medical-grade data storage mechanisms, and the retention period for device status data is insufficient, failing to meet the compliance requirements for long-term retention of medical cold chain data.

[0006] Therefore, this application proposes a method for early warning of the status of key components in a refrigerator, which is adapted to the operating conditions of medical refrigerators, complies with regulatory standards, and can achieve early warning, in order to solve the above-mentioned technical problems. Summary of the Invention

[0007] The main objective of this invention is to overcome the shortcomings of traditional monitoring solutions, such as poor environmental adaptability, delayed early warning, and insufficient regulatory adaptation, in light of the specific characteristics of medical refrigerators. It provides a method, detection and early warning system, and equipment for early warning of the status of key refrigerator components. Through multi-parameter acquisition and load curve analysis on the power supply side, it enables early identification of anomalies in components such as compressors and fans. Simultaneously, it meets the requirements for data traceability and regulatory linkage in the medical cold chain, ensuring the safety of medical supplies storage, thereby solving the technical problems mentioned in the background art.

[0008] The present invention solves the above-mentioned technical problems by adopting the following technical solutions: A method for early warning of the status of key components in a refrigerator includes the following steps: Based on the medical electrical safety standard (IEC 60601-1), the voltage, current, active power and reactive power of the power circuit of the medical refrigerator are collected synchronously at high frequency, and the data is preprocessed by the power fluctuation compensation algorithm, which is adapted to the voltage unstable working conditions. Establish load curve benchmarks for medical-specific operating conditions such as cold chain verification, intensive vaccine storage, ultra-low temperature cryogenics, and power outage backup. Extract typical load characteristic frequency bands of compressors, fans, and defrost heaters under each operating condition, and finally construct a closed-loop system for the coordinated operation of multiple components of medical refrigerators. Based on the improved independent component analysis algorithm, the overall load curve of the closed-loop system of medical refrigerator with multiple components working together is decomposed into compressor sub-curve, fan sub-curve, and defrost heater sub-curve. A tiered early warning mechanism is constructed to develop specific intervention strategies for different medical scenarios, and status warnings are issued for the compressor sub-curve, fan sub-curve, and defrost heater sub-curve according to the tiered early warning mechanism. The data is stored in Flash memory to ensure that device status data and early warning records are stored in an immutable manner, and to export compliant reports in accordance with medical regulatory requirements.

[0009] Preferably, the power fluctuation compensation algorithm uses an adaptive recursive least squares (RLS) compensation algorithm to correct voltage, current, and power data in real time, in order to solve the problem of data distortion caused by grid voltage fluctuations (within ±10%). The specific operation process of the power fluctuation compensation algorithm for data preprocessing includes: Calculate the original power The calculation formula is as follows:

[0010] in, for The raw voltage collected at all times, for The raw current collected at all times, for Power factor at any time (directly output by the metering chip); The weighting factor of the raw power is adaptively and dynamically adjusted using RLS to offset the impact of voltage fluctuations on power calculation and output the true load power value. The calculation formula is as follows:

[0011] in, for The time-adaptive weight vector (4×1 dimension) includes corrected weights for voltage, current, original power, and power factor. Preferably, the adaptive weight vector The adjustment and update formula using RLS adaptive dynamic adjustment is:

[0012]

[0013]

[0014] in, for The time-time input vector (4×1 dimension) is the original combination of collected parameters. for Real-time reference power (obtained by matching the rated load of the medical storage refrigerator with real-time operating conditions, such as the reference power of the LED light when unloaded). The forgetting factor (values ​​0.95–0.99, default 0.98) is used to control the impact of historical data on current weights, balancing stability and dynamic response. for The time-varying covariance matrix (4×4) is used to reflect the reliability of the weight estimation.

[0015] Preferably, the extraction process for the typical load characteristic frequency bands of the compressor, fan, and defrost heater under each operating condition specifically includes: Experimental data were collected from three types of typical medical storage refrigerators (vaccine refrigerator, -80℃ ultra-low temperature sample box, and laboratory refrigerator) under multiple operating conditions. Five core operating conditions were tested for each model, and data was continuously collected for 24 hours (1kHz sampling rate) for each operating condition. During the experiment, the start-stop status of each device (such as compressor start / stop, fan speed regulation, and defrost heater on / off) was recorded simultaneously to establish a data-device status mapping label. The collected data is detrended and denoised (Kalman filtering) to remove extreme outliers (such as zero values ​​caused by sudden voltage drops). The dataset is split according to the working condition label, and each working condition generates an independent time-series data segment (e.g., the cold chain verification working condition data segment is 8 hours long, corresponding to the complete process of dropping from 25℃ to 2℃). For each operating condition data segment, perform Fast Fourier Transform (FFT) to convert the time domain signal into a frequency domain signal, extract the load characteristic frequency band of each device, and establish a four-dimensional mapping table of operating condition-device-characteristic frequency band-key parameters. Finally, statistical analysis was performed on the experimental data of the three models, and the 95% confidence interval was taken as the normal characteristic frequency band and parameter threshold of the device under each operating condition, so as to form a special load characteristic benchmark library for medical storage refrigerators.

[0016] Preferably, the experimental data acquisition conditions include: Operating condition 1: Idle / full load, as a normal storage operating condition; Operating condition 2: Continuous cooling, with the temperature dropping from 25°C to 2°C, serving as a cold chain verification condition; Operating condition 3: Frequent door opening, 10 minutes / time, as the operating condition for vaccine storage; Operating Condition 4: Full-load long-term operation, with the internal temperature of the chamber remaining stable at 2℃, serving as a condition for dense sample storage; Operating condition 5: Automatic defrosting start-run-end, serving as the defrosting cycle condition; Furthermore, the start-stop status of each device (such as compressor start / stop, fan speed regulation, and defrost heater on / off) is recorded synchronously during the experiment to establish a data-device status mapping label.

[0017] Preferably, extracting the load characteristic frequency bands of each device includes: Compressor: The characteristic frequency band during the start-up phase is 5-10Hz (corresponding to the starting current ripple), and the characteristic frequency band during steady-state operation is 0.1-1Hz (corresponding to periodic load fluctuations). Condensing / evaporating fans: Low-speed operation characteristic frequency band 1-3Hz, high-speed operation characteristic frequency band 3-8Hz, speed regulation process characteristic frequency band 8-15Hz; Defrosting heater: The characteristic frequency band during the start-up phase is 0.05 to 0.1 Hz (power rises slowly), and the characteristic frequency band during steady-state operation is 0 Hz (constant power). Extract key parameters corresponding to the characteristic frequency bands: peak power, power fluctuation rate, duration, and rise / fall slope, such as the peak power range of compressor startup (190W~210W) and the steady-state power fluctuation rate of the fan ≤3%.

[0018] Preferably, the extraction of the load characteristic frequency bands of each device is based on an improved independent component analysis algorithm. For the closed-loop system of a medical refrigerator with multiple devices working collaboratively, a set of decomposition formulas is constructed as follows: Used to transmit the overall load signal Disassembled into compressor sub-signals Fan sub-signal Defrosting heater sub-signal LED light signal The linear superposition of, where It is a mixed matrix. For time, The noise signal is used to accurately break down the overall load curve into the compressor sub-curve, fan sub-curve, and defrost heater sub-curve.

[0019] Preferably, the improved independent component analysis algorithm performs the following decomposition process for the overall load signal: The centralized processing operation for the overall load signal is performed using the following formula: ,in, For expectation operators; The overall load signal after centralized processing The formula for whitening treatment is as follows:

[0020] in, This is the model data after whitening treatment. , and These are the covariance matrices. The eigenvector matrix and eigenvalue matrix, after whitening, satisfy... , It is the identity matrix; Objective function with introduced operating condition weighting factors: Based on the conventional ICA objective function (maximizing non-Gaussianity):

[0021] The improved objective function is:

[0022] in, For non-linear activation functions, the logistic function is generally chosen. For operating condition weighting factors, This is the current operating condition number, such as k=2 for cold chain verification operating condition, with a value ranging from 0.3 to 1.5. It is determined by the operating condition-characteristic frequency band mapping table. In the mapping table, the higher the signal strength of the characteristic frequency band, the greater its weight. for Reference sub-signals of the device under operating conditions (retrieved from the reference feature library); Based on the adaptive learning rate iterative solution of the objective function during the weight matrix update process, we have:

[0023] For the adaptive learning rate during the weight matrix update process, =0.01 represents the initial learning rate. =1000 is the iteration decay constant), at which point the weight matrix update formula is: ; The final sub-signal reconstruction calculation is as follows:

[0024] in, for weight matrix The Column, corresponding to the first Disassembly weight of each component.

[0025] Preferably, the key parameters also need to be explained during the disassembly process as follows: The mixing matrix A is used to reflect the mixing ratio of the load signals of each device, and is determined by the circuit topology and device power ratio of the medical storage refrigerator. Operating condition weighting factor It is used to enhance the characteristic signals of the target device under the current operating conditions and suppress the interference signals of other operating conditions. For example, under the cold chain verification operating conditions, the compressor weight ω1=1.5 and the fan weight ω2=0.5. Adaptive learning rate This is used to indicate that a larger learning rate is used in the early stages of iteration for rapid convergence, and the learning rate is reduced in the later stages to improve accuracy and avoid oscillations.

[0026] In another aspect, the present invention also discloses a critical component status detection and early warning system for a refrigerator, installed inside the refrigerator, for performing the steps of any of the methods described above, including: The power acquisition module is equipped with multiple sets to collect the voltage, current, active power, and reactive power of the refrigerator's power circuit, and to perform preprocessing operations on the specified data. The system environment construction module is used to construct the load characteristic frequency band of the device under specified operating conditions based on the collected data and the data stored in the system. The component analysis module, located on the micro control unit, is used to perform the splitting operation of specified sub-devices based on the load characteristic frequency band constructed by the system environment construction module. The early warning module, located on the micro control unit, is used to issue specified early warning signals based on the sub-curves split by the component analysis module. The storage module, located on the micro control unit, is used to store specified data within the system and display and adjust the output on the display module of the micro control unit according to the specified report format. The IoT module, located on the micro control unit, is used to drive and adjust specified parameters of various electrical components of the refrigerator, and to transmit the specified parameters of the electrical components to the cloud server in real time.

[0027] In another aspect, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0028] As can be seen from the above technical solution, the present invention provides a method, detection and early warning system, and equipment for early warning of the status of key components in a refrigerator. Compared with the prior art, the present invention has the following advantages: 1. This invention eliminates the need for dedicated sensors and achieves monitoring by collecting status data of multiple devices only through the power input terminal (power side acquisition terminal). This avoids damage to monitoring components caused by medical disinfection and ultra-low temperature environments, reducing the monitoring failure rate to below 5%. It is suitable for special scenarios such as medical ultra-low temperature freezers and solves the problem of traditional solutions relying on dedicated sensors, which are prone to failure in medical disinfection and ultra-low temperature environments. Ultimately, it improves environmental adaptability and monitoring stability.

[0029] 2. This invention identifies latent device faults (such as load fluctuations in the early stages of compressor bearing wear and current ripple changes in the precursors of fan jamming) by detecting minute distortions in the load curve. This advances the warning point to the nascent stage of the fault, effectively preventing the deterioration and loss of vaccines and biological samples due to device failure, significantly reducing the risk of medical cold chain storage accidents, and thus solving the problem of medical supply loss caused by delayed warnings, achieving early identification and warning of latent device faults.

[0030] 3. This invention, through component-level load curve analysis, can clearly identify whether the compressor overload is caused by its own fault or fan jamming, making it easier to accurately distinguish collaborative faults, achieve precise fault tracing, significantly shorten the operation and maintenance troubleshooting time of medical refrigerators, reduce operation and maintenance costs, solve the problem of difficulty in tracing the source of multi-component collaborative faults, and realize independent disassembly and correlation analysis of component loads.

[0031] It should be understood that the descriptions in this section are not intended to identify key or essential features of embodiments of the invention, nor are they intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Of course, implementing any product of the invention does not necessarily require achieving all of the advantages described above simultaneously. Attached Figure Description

[0032] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a schematic diagram of the medical scenario workflow of the present invention; Figure 2 This is a block diagram of the overall system structure of the present invention. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] For details in the embodiments, please refer to Figure 1 and Figure 2 .

[0035] like Figure 1 As shown, the refrigerator key component status early warning method proposed in this embodiment of the invention includes the following steps: S1. Based on the medical electrical safety standard (IEC 60601-1), the voltage, current, active power and reactive power of the power circuit of the medical refrigerator are collected synchronously at high frequency, and the data is preprocessed by the power fluctuation compensation algorithm, which is adapted to the unstable voltage condition.

[0036] The power fluctuation compensation algorithm employs an adaptive recursive least squares (RLS) compensation algorithm to correct voltage, current, and power data in real time, thereby addressing the data distortion problem caused by grid voltage fluctuations (within ±10%). The specific data preprocessing steps of the power fluctuation compensation algorithm include: Calculate the original power The calculation formula is as follows:

[0037] in, for The raw voltage collected at all times, for The raw current collected at all times, for Power factor at any time (directly output by the metering chip); The weighting factor of the raw power is adaptively and dynamically adjusted using RLS to offset the impact of voltage fluctuations on power calculation and output the true load power value. The calculation formula is as follows:

[0038] in, for The time-adaptive weight vector (4×1 dimension) includes corrected weights for voltage, current, original power, and power factor. Furthermore, the adaptive weight vector The adjustment and update formula using RLS adaptive dynamic adjustment is:

[0039]

[0040]

[0041] in, for The time-time input vector (4×1 dimension) is the original combination of collected parameters. for Real-time reference power (obtained by matching the rated load of the medical storage refrigerator with real-time operating conditions, such as the reference power of the LED light when unloaded). The forgetting factor (values ​​0.95–0.99, default 0.98) is used to control the impact of historical data on current weights, balancing stability and dynamic response. for The time-varying covariance matrix (4×4) is used to reflect the reliability of the weight estimation.

[0042] S2. Establish load curve benchmarks for medical-specific operating conditions such as cold chain verification, intensive vaccine storage, ultra-low temperature cryogenics, and power outage backup. Extract typical load characteristic frequency bands of compressors, fans, and defrost heaters under each operating condition, and finally construct a closed-loop system for the coordinated operation of multiple components of the medical refrigerator.

[0043] The experimental data acquisition conditions included: Operating condition 1: Idle / full load, as a normal storage operating condition; Operating condition 2: Continuous cooling, with the temperature dropping from 25°C to 2°C, serving as a cold chain verification condition; Operating condition 3: Frequent door opening, 10 minutes / time, as the operating condition for vaccine storage; Operating Condition 4: Full-load long-term operation, with the internal temperature of the chamber remaining stable at 2℃, serving as a condition for dense sample storage; Operating condition 5: Automatic defrosting start-run-end, serving as the defrosting cycle condition; Furthermore, the start-stop status of each device (such as compressor start / stop, fan speed regulation, and defrost heater on / off) is recorded synchronously during the experiment to establish a data-device status mapping label.

[0044] Therefore, the extraction process at this point also includes: Experimental data were collected from three types of typical medical storage refrigerators (vaccine refrigerator, -80℃ ultra-low temperature sample box, and laboratory refrigerator) under multiple operating conditions. Five core operating conditions were tested for each model, and data was continuously collected for 24 hours (1kHz sampling rate) for each operating condition. During the experiment, the start-stop status of each device (such as compressor start / stop, fan speed regulation, and defrost heater on / off) was recorded simultaneously to establish a data-device status mapping label. The collected data is detrended and denoised (Kalman filtering) to remove extreme outliers (such as zero values ​​caused by sudden voltage drops). The dataset is split according to the working condition label, and each working condition generates an independent time-series data segment (e.g., the cold chain verification working condition data segment is 8 hours long, corresponding to the complete process of dropping from 25℃ to 2℃). For each operating condition data segment, perform Fast Fourier Transform (FFT) to convert the time domain signal into a frequency domain signal, extract the load characteristic frequency band of each device, and establish a four-dimensional mapping table of operating condition-device-characteristic frequency band-key parameters. Finally, statistical analysis was performed on the experimental data of the three models, and the 95% confidence interval was taken as the normal characteristic frequency band and parameter threshold of the device under each operating condition, so as to form a special load characteristic benchmark library for medical storage refrigerators.

[0045] Furthermore, the extraction of the load characteristic frequency bands of each device at this time includes: Compressor: The characteristic frequency band during the start-up phase is 5-10Hz (corresponding to the starting current ripple), and the characteristic frequency band during steady-state operation is 0.1-1Hz (corresponding to periodic load fluctuations). Condensing / evaporating fans: Low-speed operation characteristic frequency band 1-3Hz, high-speed operation characteristic frequency band 3-8Hz, speed regulation process characteristic frequency band 8-15Hz; Defrosting heater: The characteristic frequency band during the start-up phase is 0.05 to 0.1 Hz (power rises slowly), and the characteristic frequency band during steady-state operation is 0 Hz (constant power). Extract key parameters corresponding to the characteristic frequency bands: peak power, power fluctuation rate, duration, and rise / fall slope, such as the peak power range of compressor startup (190W~210W) and the steady-state power fluctuation rate of the fan ≤3%.

[0046] S3. Based on the improved independent component analysis algorithm, the overall load curve of the closed-loop system of the medical refrigerator with multiple components working together is decomposed into compressor sub-curve, fan sub-curve, and defrost heater sub-curve.

[0047] Specifically, at this point, based on an improved independent component analysis algorithm, a set of decomposition formulas is constructed for the closed-loop system of a medical refrigerator with multiple components working collaboratively, by extracting the load characteristic frequency bands of each device. Used to transmit the overall load signal Disassembled into compressor sub-signals Fan sub-signal Defrosting heater sub-signal LED light signal The linear superposition of, where It is a mixed matrix. For time, The noise signal is used to accurately break down the overall load curve into the compressor sub-curve, fan sub-curve, and defrost heater sub-curve.

[0048] Furthermore, the improved independent component analysis algorithm's decomposition process for the overall load signal is as follows: The centralized processing operation for the overall load signal is performed using the following formula: ,in, For expectation operators; The overall load signal after centralized processing The formula for whitening treatment is as follows:

[0049] in, This is the model data after whitening treatment. , and These are the covariance matrices. The eigenvector matrix and eigenvalue matrix, after whitening, satisfy... , It is the identity matrix; Objective function with introduced operating condition weighting factors: Based on the conventional ICA objective function (maximizing non-Gaussianity):

[0050] The improved objective function is:

[0051] in, For non-linear activation functions, the logistic function is generally chosen. For operating condition weighting factors, This is the current operating condition number, such as k=2 for cold chain verification operating condition, with a value ranging from 0.3 to 1.5. It is determined by the operating condition-characteristic frequency band mapping table. In the mapping table, the higher the signal strength of the characteristic frequency band, the greater its weight. for Reference sub-signals of the device under operating conditions (retrieved from the reference feature library); The results, compared with the conventional ICA objective function before the improvement, are shown in Table 1 below: Table 1: Comparison of ICA Algorithm Improvements

[0052] Based on the adaptive learning rate iterative solution of the objective function during the weight matrix update process, we have:

[0053] For the adaptive learning rate during the weight matrix update process, =0.01 represents the initial learning rate. =1000 is the iteration decay constant), at which point the weight matrix update formula is: ; The final sub-signal reconstruction calculation is as follows:

[0054] in, for weight matrix The Column, corresponding to the first Disassembly weight of each component.

[0055] In addition, it should be noted that the following key parameters need to be explained during the disassembly process: The mixing matrix A is used to reflect the mixing ratio of the load signals of each device, and is determined by the circuit topology and device power ratio of the medical storage refrigerator. Operating condition weighting factor It is used to enhance the characteristic signals of the target device under the current operating conditions and suppress the interference signals of other operating conditions. For example, under the cold chain verification operating conditions, the compressor weight ω1=1.5 and the fan weight ω2=0.5. Adaptive learning rate This is used to indicate that a larger learning rate is used in the early stages of iteration for rapid convergence, and the learning rate is reduced in the later stages to improve accuracy and avoid oscillations.

[0056] S4. Finally, a hierarchical early warning mechanism is constructed to develop exclusive intervention strategies for different medical scenarios, and to issue status warnings for the compressor sub-curve, fan sub-curve, and defrost heater sub-curve according to the hierarchical early warning mechanism.

[0057] S5. Use Flash storage for data to ensure the immutable storage of device status data and early warning records, and export compliant reports in accordance with medical regulatory requirements.

[0058] In summary, this method identifies latent device faults by detecting minute distortions in the load curve (such as load fluctuations in the early stages of compressor bearing wear and current ripple changes in the precursors of fan jamming). By advancing the warning point to the nascent stage of the fault, it can effectively prevent the deterioration and loss of vaccines and biological samples due to device failure, significantly reduce the risk of medical cold chain storage accidents, and thus solve the problem of medical supply losses caused by delayed warnings, achieving early identification and warning of latent device faults.

[0059] Furthermore, by disassembling the load curves at the component level, it is possible to clearly identify whether the compressor overload is caused by its own fault or by fan jamming, which facilitates the accurate differentiation of collaborative faults, enables precise fault tracing, significantly shortens the maintenance and troubleshooting time of medical refrigerators, reduces maintenance costs, solves the problem of difficulty in tracing the source of multi-component collaborative faults, and realizes independent disassembly and correlation analysis of component loads.

[0060] On the other hand, such as Figure 2 As shown, this invention also discloses a critical component status detection and early warning system for a refrigerator, installed inside the refrigerator, for executing the steps of any of the methods described above. In specific implementation, its module units specifically include: (1) Power acquisition module, which is set up in multiple groups, consisting of medical power parameter acquisition unit and medical working condition data preprocessing unit. The medical power parameter acquisition unit is used to acquire the voltage, current, active power and reactive power of the refrigerator power circuit, and the medical working condition data preprocessing unit performs the preprocessing operation of the specified data.

[0061] The medical power parameter acquisition unit is integrated into the power input terminal of the medical refrigerator and complies with the IEC60601-1 medical electrical safety standard. It includes a medical isolation voltage transformer, a current transformer, a power metering chip, and a microcontroller. Its core functions are: to synchronously acquire voltage, current, active power, and reactive power data at high frequency; and to have a built-in power fluctuation filtering module to adapt to voltage fluctuation conditions. The acquired data is encrypted and encapsulated before being transmitted to the preprocessing unit.

[0062] The medical operating condition data preprocessing unit is located on the main control board of the medical refrigerator, including medical-grade digital filtering and operating condition data alignment.

[0063] (2) System environment construction module, including medical-specific load curve fitting unit, used to construct load characteristic frequency bands of devices under specified operating conditions based on the collected data and the data stored in the system.

[0064] The medical-specific load curve fitting unit, based on a medical operating condition dataset, completes the fitting of load curves and the decomposition of device sub-curves. The steps are as follows: (2a) Medical working condition classification: The data is classified into medical-specific working condition datasets such as routine storage, cold chain verification, vaccine warehousing, and defrosting according to the working condition labels; (2b) Generation of whole machine load curve: With time as the horizontal axis and power as the vertical axis, fit the continuous load curve of the whole machine under different medical working conditions; (2c) Device sub-curve decomposition: Combining the device state model matching unit, for the closed-loop refrigeration system of the medical refrigerator "compressor-fan-defrost heater", the whole machine curve is decomposed into the sub-curves of each device, and the core characteristic parameters under the medical scenario are extracted (such as the power fluctuation rate of the compressor during continuous operation under the cold chain verification condition and the power rise slope of the defrost heater).

[0065] (3) Component analysis module, located on the micro control unit, includes a device state model matching unit, which is used to perform the splitting operation of specified sub-devices based on the load characteristic frequency band constructed by the system environment construction module.

[0066] The device state model matching unit has a built-in medical storage refrigerator device load benchmark library and medical operating condition deviation assessment component. At its core, the device state model matching unit includes medical refrigerator device load benchmarks and medical operating condition deviation assessments. (3a) The medical refrigerator component load benchmark library pre-stores the normal load characteristic range of different models of medical refrigerators (such as vaccine refrigerators and ultra-low temperature sample boxes), and imports qualified medical cold chain verification data of the same model of equipment.

[0067] (3b) Medical condition deviation assessment compares the real-time sub-curve characteristic parameters with the benchmark library, calculates the deviation value, and simultaneously associates the medical condition weights. The specific steps are as follows: L1. Medical Storage Refrigerator Component Load Benchmark Library: Pre-stores the normal load characteristic range and sub-curve templates of components for different models of medical storage refrigerators (such as vaccine refrigerators and ultra-low temperature sample boxes), and imports qualified medical cold chain verification data of the same model of equipment. For example, under the cold chain verification conditions of a certain model of vaccine refrigerator compressor: peak starting power 190W~210W, steady-state power 85W~95W, power fluctuation rate ≤2%; L2. Medical Operating Condition Deviation Assessment Module: Calculates the deviation values ​​between real-time feature parameters and the benchmark library according to the "device-operating condition" dimension. Specific calculation method: For numerical parameters (such as peak startup power):

[0068] in, For the first The relative deviation of each characteristic parameter These are the feature parameter values ​​extracted from the real-time sub-curve. The median value of the parameter thresholds under the corresponding working conditions in the benchmark library; For interval-type parameters (such as power volatility): (If the real-time value exceeds the upper limit) or (If the real-time value is below the lower limit); Overall device deviation: ,in The weights for the characteristic parameters are (e.g., the weight for peak startup power is 0.3, and the weight for power volatility is 0.2, which are determined from experimental data using the analytic hierarchy process). L3. Operating Condition Weighting: For key operating conditions such as cold chain verification and vaccine warehousing, the overall deviation will be considered. Multiply by a magnification factor of 1.2 to ensure that no anomalies under critical operating conditions are missed.

[0069] (4) Early warning module, located on the micro control unit, including medical graded early warning and monitoring linkage unit, used to issue specified early warning signals based on the sub-curves after being split by the component analysis module.

[0070] The medical-grade tiered early warning and regulatory linkage unit has preset four-level early warning thresholds specific to medical scenarios and supports integration with medical regulatory platforms, which specifically include: Attention level (deviation value 5%-10%): This is determined to be a minor abnormality of the device (such as LED light aging). The local control panel of the refrigerator will display a yellow warning, and the data will be recorded to the traceability unit. Warning level (deviation value 10%-20%): Determined as a moderate abnormality of the device (such as slight wear of the compressor bearing), and a warning message is pushed to the operation and maintenance terminal; Alarm level (deviation value 20%-30%): If the device is determined to be severely abnormal (such as fan jamming), a local buzzer alarm will be triggered on the refrigerator, and an information monitoring platform will be pushed to remind maintenance personnel to conduct an emergency investigation. Emergency Level (deviation > 30%): If the device is identified as malfunctioning (such as compressor startup failure), the medical emergency linkage interface is triggered, and an emergency alarm is sent to medical staff.

[0071] (5) Storage module, located on the micro control unit, including medical cold chain data storage and traceability unit, used to store specified data in the system and display and adjust the output on the display module of the micro control unit according to the specified report format.

[0072] The medical cold chain data storage and traceability unit uses Flash memory to meet medical data storage compliance requirements. (5a) The stored content includes power acquisition data, load curves, early warning records, and operating condition indicators. The storage period is ≥5 years, and the data cannot be tampered with. (5b) Supports exporting compliant reports according to medical regulatory requirements, and provides a MODBUS interface to connect with the medical device management system to achieve real-time data upload and traceability; (5c) Built-in model iteration module updates the benchmark library every 30 days based on the historical compliant operation data of the medical refrigerator, and the aging characteristics of the adapter components gradually change under long-term medical working conditions.

[0073] (6) The IOT module is located on the micro control unit and is used to drive and adjust the specified parameters of each electrical component of the refrigerator, and to transmit the specified parameters of the electrical components to the cloud server in real time.

[0074] The connection relationships between the units are as follows: a. The signal output terminal of the medical power supply parameter acquisition unit is electrically connected to the signal input terminal of the medical working condition data preprocessing unit; b. The output of the medical operating condition data preprocessing unit is communicatively connected to the input of the medical-specific load curve fitting unit; c. The output of the medical-specific load curve fitting unit is connected to the input of the device state model matching unit and the input of the medical cold chain data storage and traceability unit, respectively. d. The output of the device state model matching unit is connected to the input of the medical-grade hierarchical early warning and regulatory linkage unit; e. The feedback end of the medical cold chain data storage and traceability unit is connected to the model update end of the device status model matching unit, and its data upload end is connected to the external medical cold chain supervision platform.

[0075] In addition, the system is also equipped with a medical emergency linkage interface, which is used to connect to medical backup power supplies and refrigerator emergency temperature control modules to realize emergency intervention after early warning.

[0076] In another embodiment, the power acquisition module is not limited to using a wired acquisition unit, but can also adopt a medical wireless acquisition module to adapt to medical scenarios (such as a medical Bluetooth BLE module, which complies with the safety standards of medical wireless devices). In this case, it can be used for the renovation of old medical refrigerators without disassembling the power circuit, and can realize non-contact data acquisition during disinfection.

[0077] In another embodiment, a medical supplies transfer linkage module can be added during the emergency warning process. When an emergency warning is triggered, a supplies transfer prompt is automatically sent to the management terminal of the surrounding backup medical refrigerators, and the current temperature control data of the refrigerators is pushed to provide a basis for decision-making for the transfer of medical supplies.

[0078] In another embodiment, a medical refrigerator cluster monitoring cloud platform can be added to enable comparative analysis of load curves of multiple medical refrigerators within a hospital / disease control center, identify common component anomalies in the same batch of medical refrigerators (such as early batch failures of compressors in a batch of vaccine refrigerators), and generate a cluster medical cold chain risk assessment report.

[0079] In another embodiment, an automatic cold chain verification trigger module can be added. When an abnormal compressor load is detected, the cold chain verification process is automatically paused and an anomaly report is pushed to ensure the compliance of the verification data.

[0080] In another embodiment, auxiliary medical data collection interfaces such as temperature and humidity inside the medical refrigerator, door opening and closing frequency, and vaccine storage volume can be added. The power load curve can be integrated with the above data for analysis. For example, when the compressor load is abnormal and the cooling rate inside the refrigerator is lower than the medical standard threshold, the compressor cooling efficiency can be accurately determined, and the risk level can be assessed in conjunction with the vaccine storage volume.

[0081] The following operating procedures are adopted in sequence during the actual use of this system: Power data acquisition: When the medical refrigerator is powered on, the acquisition unit starts. Data preprocessing: denoising and filtering + medical condition label alignment; Operating condition identification: Receive operating condition instructions such as cold chain verification / vaccine warehousing, and divide the dataset; Curve fitting: Generate the overall curve and break it down into sub-curves for medical-specific components; Component matching: Compare with the medical refrigerator benchmark library and calculate the deviation according to the operating condition weight; Warning determination: Determine the warning level as Level IV; Output traceability: Triggers local / remote early warnings, and in emergency situations, activates backup power and material transfer prompts; saves data and uploads it to the monitoring platform, and regularly iterates the medical benchmark database to achieve process loop.

[0082] In summary, this method and system eliminates the need for dedicated sensors and achieves monitoring by collecting status data of multiple devices only through the power input terminal (power side acquisition terminal). This avoids damage to monitoring components caused by medical disinfection and ultra-low temperature environments, reducing the monitoring failure rate to below 5%. It is suitable for special scenarios such as medical ultra-low temperature freezers and solves the problem of traditional solutions relying on dedicated sensors that are prone to failure in medical disinfection and ultra-low temperature environments, ultimately improving environmental adaptability and monitoring stability.

[0083] In addition, the system also establishes a load benchmark library for specific operating conditions such as cold chain verification and vaccine storage to avoid false alarms under special operating conditions. It can adapt to the full operating condition monitoring needs of medical refrigerators, thereby solving the problem that existing technologies cannot adapt to the special operating conditions of medical refrigerators (cold chain verification, vaccine storage, vehicle movement) and are prone to misjudging the status of devices.

[0084] Furthermore, with its built-in medical-grade data storage module, device status data and early warning records can be retained for ≥5 years. It also supports connection to the National Vaccine Cold Chain Traceability Platform and hospital medical equipment management systems, enabling real-time data upload and traceability, and complying with GSP, GMP and other regulatory requirements. This solves the problem of monitoring data not being able to connect to the medical cold chain supervision platform and not meeting the compliance requirements of GSP and other regulations, thus meeting medical regulatory compliance in actual use.

[0085] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.

[0086] In another aspect, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0087] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute the steps in any of the refrigerator key component status early warning methods in the above embodiments.

[0088] It is understood that the system provided in the embodiments of the present invention corresponds to the method provided in the embodiments of the present invention, and the explanation, examples and beneficial effects of the relevant content can be referred to the corresponding parts of the above methods.

[0089] This application also provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, communication interface, and memory communicate with each other via the communication bus. Memory, used to store computer programs; When the processor executes the program stored in the memory, it implements the steps in the above-mentioned method for early warning of the status of key refrigerator components.

[0090] The communication bus mentioned in the above-mentioned electronic devices can be a standard bus for interconnecting peripheral components or an extended industrial standard structure bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc.

[0091] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0092] The memory may include random access memory or non-volatile memory, such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0093] The processors mentioned above can be general-purpose processors, including central processing units, network processors, etc.; they can also be digital signal processors, application-specific integrated circuits, field-programmable gate arrays or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0094] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented entirely or partially as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium, an optical medium, or a semiconductor medium, etc.

[0095] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0096] Furthermore, it should be noted that if any directional indication (such as up, down, left, right, front, back, etc.) is involved in the embodiments of the present invention, the directional indication is only used to explain the relative positional relationship and movement of each component in a specific posture. If the specific posture changes, the directional indication will also change accordingly.

[0097] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the meaning of "and / or" throughout the text includes three parallel solutions; for example, "A and / or B" includes solution A, solution B, or a solution where both A and B are satisfied simultaneously. Furthermore, in the embodiments of this invention, "multiple" refers to two or more. Moreover, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

Claims

1. A method for early warning of the status of key components in a refrigerator, characterized in that, include: The voltage, current, active power, and reactive power of the power circuit of the medical refrigerator are collected at high frequency and preprocessed using a power fluctuation compensation algorithm. Establish a load curve benchmark for medical-specific operating conditions, extract typical load characteristic frequency bands of compressor, fan, and defrost heater under each operating condition, and finally construct a closed-loop system for the coordinated operation of multiple components in a medical refrigerator. Based on the improved independent component analysis algorithm, the overall load curve of the closed-loop system of medical refrigerator with multiple components working together is decomposed into compressor sub-curve, fan sub-curve, and defrost heater sub-curve. A tiered early warning mechanism is established, which issues status warnings for the compressor sub-curve, fan sub-curve, and defrost heater sub-curve according to the tiered early warning mechanism. The data is stored in Flash and exported in accordance with medical regulatory requirements to create compliant reports.

2. The refrigerator key component status early warning method as described in claim 1, characterized in that, The specific operation process of data preprocessing in the power fluctuation compensation algorithm includes: Calculate the original power The calculation formula is as follows: in, for The raw voltage collected at all times, for The raw current collected at all times, for Power factor at any given moment; The weighting factor of the raw power is adaptively and dynamically adjusted using RLS to offset the impact of voltage fluctuations on power calculation and output the true load power value. The calculation formula is as follows: in, for The time-adaptive weight vector includes corrected weights for voltage, current, original power, and power factor.

3. The refrigerator key component status early warning method as described in claim 2, characterized in that, The adaptive weight vector The adjustment and update formula using RLS adaptive dynamic adjustment is: in, for The input vector at each time step is the original combination of acquired parameters. for Constant reference power, The forgetting factor is used to control the influence of historical data on current weights. for The time-varying covariance matrix is ​​used to reflect the reliability of the weight estimation.

4. The refrigerator key component status early warning method as described in claim 1, characterized in that, The extraction process for the typical load characteristic frequency bands of the compressor, fan, and defrost heater under each operating condition specifically includes: Collect experimental data of a typical medical storage refrigerator under multiple operating conditions and simultaneously record the start-stop status of each device to establish a data-device status mapping label. The collected data is detrended and denoised to remove extreme outliers; The dataset is split according to the working condition label, and each working condition generates an independent time series data segment; For each operating condition, a fast Fourier transform is performed on the data segment to convert the time-domain signal into a frequency-domain signal, extract the load characteristic frequency band of each device, and establish a four-dimensional mapping table of operating condition-device-characteristic frequency band-key parameters. Finally, statistical analysis of the experimental data was conducted to form a baseline library of load characteristics.

5. The refrigerator key component status early warning method as described in claim 4, characterized in that, The experimental data acquisition conditions include: Operating condition 1: Idle / full load, as a normal storage operating condition; Operating condition 2: Continuous cooling, with the temperature dropping from 25°C to 2°C, serving as a cold chain verification condition; Operating condition 3: Frequent door opening, 10 minutes / time, as the operating condition for vaccine storage; Operating Condition 4: Full-load long-term operation, with the internal temperature of the chamber remaining stable at 2℃, serving as a condition for dense sample storage; Operating condition 5: Automatic defrosting start-run-end, serving as the defrosting cycle condition; Furthermore, the start-stop status of each device was recorded synchronously during the experiment to establish a data-device status mapping label.

6. The refrigerator key component status early warning method as described in claim 4, characterized in that, Extracting the load characteristic frequency bands of each device includes: Compressor: The characteristic frequency band during startup is 5-10Hz, and the characteristic frequency band during steady-state operation is 0.1-1Hz; Condensing / evaporating fans: Low-speed operation characteristic frequency band 1-3Hz, high-speed operation characteristic frequency band 3-8Hz, speed regulation process characteristic frequency band 8-15Hz; Defrosting heater: characteristic frequency band during startup 0.05~0.1Hz, characteristic frequency band during steady-state operation 0Hz; Extract key parameters corresponding to the characteristic frequency bands: peak power, power fluctuation rate, duration, and rise / fall slope.

7. The refrigerator key component status early warning method as described in claim 6, characterized in that, During the extraction of the load characteristic frequency bands of each device, an improved independent component analysis algorithm is used to construct a set of decomposition formulas. Used to transmit the overall load signal Disassembled into compressor sub-signals Fan sub-signal Defrosting heater sub-signal LED light signal The linear superposition of, where It is a mixed matrix. For time, The noise signal is used to break down the overall load curve into sub-curves for the compressor, fan, and defrost heater.

8. The refrigerator key component status early warning method as described in claim 7, characterized in that, The improved independent component analysis algorithm performs the following decomposition process for the overall load signal: The centralized processing operation for the overall load signal is performed using the following formula: ,in, For expectation operators; The overall load signal after centralized processing The formula for whitening treatment is as follows: in, This is the model data after whitening treatment. , and These are the covariance matrices. The eigenvector matrix and eigenvalue matrix, after whitening, satisfy... , It is the identity matrix; The objective function that introduces the working condition weighting factor is: in, It is a non-linear activation function. For operating condition weighting factors, This is the current operating condition number. for Reference sub-signal of the device under operating conditions; Based on the adaptive learning rate iterative solution of the objective function during the weight matrix update process, we have: For the adaptive learning rate during the weight matrix update process, This is represented as the initial learning rate. (where is the iterative decay constant), at which point the weight matrix update formula is: ； The final sub-signal reconstruction calculation is as follows: in, for weight matrix The Column, corresponding to the first Disassembly weight of each component.

9. A status detection and early warning system for key components of a refrigerator, characterized in that, The device, located inside a refrigerator, is used to perform the steps of the method as described in any one of claims 1 to 8, including: The power acquisition module is used to collect the voltage, current, active power, and reactive power of the refrigerator's power circuit and perform preprocessing operations on the specified data. The system environment construction module is used to construct the load characteristic frequency band of the device under specified operating conditions based on the collected data and the data stored in the system. The component analysis module, located on the micro control unit, is used to perform the splitting operation of specified sub-devices based on the load characteristic frequency band constructed by the system environment construction module. The early warning module, located on the micro control unit, is used to issue specified early warning signals based on the sub-curves split by the component analysis module. The storage module, located on the micro control unit, is used to store specified data within the system and display and adjust the output on the display module of the micro control unit according to the specified report format. The IoT module, located on the micro control unit, is used to drive and adjust specified parameters of various electrical components of the refrigerator, and to transmit the specified parameters of the electrical components to the cloud server in real time.

10. A computer device, characterized in that, It includes a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 8.

Images