A method for detecting refrigeration faults in ultra-low temperature freezers

Abstract

This invention relates to the field of fault detection technology, specifically a method for detecting refrigeration faults in ultra-low temperature freezers. This invention collects and outputs external sensor data and refrigeration system data through an information sensing layer, overcoming the limitations of relying on a single external sensor. Furthermore, the external sensing unit and the refrigeration sensing unit collect data specifically; when an external sensor malfunctions, fault detection can be achieved through refrigeration system data, improving the stability and reliability of detection and avoiding detection failures due to sensor malfunctions. This invention identifies data stability indicators from external sensor data and performs health assessments on these indicators to obtain data health signals. It comprehensively analyzes the total abnormal values ​​of external sensors and the total abnormal values ​​of the refrigeration system to classify abnormalities into levels and generates corresponding fault alarm strategies. This approach matches the severity of the fault, reduces false alarms and missed alarms, and ensures the safe operation of the freezer.

Description

Technical Field

[0001] This invention relates to the field of fault detection technology, specifically a method for detecting refrigeration faults in ultra-low temperature freezers. Background Technology

[0002] Ultra-low temperature freezers are used to meet the rigid demand for extreme low-temperature preservation in special fields: In the biomedical field, they can maintain the activity and structural stability of biological samples such as stem cells, vaccines, and organs, and prevent protein denaturation or microbial growth; in the scientific research field, they can provide a stable low-temperature environment for the preservation of superconducting materials, low-temperature physics experiments, and deep-sea bacterial strains; in the pharmaceutical industry, they are key equipment for the long-distance transportation and storage of special reagents in cold chain logistics. Therefore, the detection of refrigeration failure is one of the important links. Once the refrigeration fails, the temperature will rise in a short time, causing the stored items to deteriorate in a short period of time, resulting in huge economic losses.

[0003] Existing methods for detecting refrigeration faults in ultra-low temperature freezers collect data using integrated multimodal data sensors and detect refrigeration faults based on various modal data. However, this over-reliance on added modal sensors ignores the fact that the freezer's refrigeration system itself is an information source that can provide data support for fault detection. The traditional method of relying solely on sensors to provide information cannot provide stable fault detection capabilities when the sensors malfunction. Summary of the Invention

[0004] This invention provides a method for detecting refrigeration faults in ultra-low temperature freezers, which solves the aforementioned technical problems.

[0005] The first aspect of this invention provides a method for detecting refrigeration faults in ultra-low temperature freezers, comprising the following steps:

[0006] Step 1: Collect information based on the information sensing layer of the freezer and output external sensor data and refrigeration system data; execute step 2 on the external sensor data and send the refrigeration system data to step 3;

[0007] As a further improvement of the present invention, information is collected and output based on the information sensing layer of the freezer, specifically as follows:

[0008] Step 1.1: Collect freezer temperature data, freezer pressure data, and freezer vibration data through the external sensing unit of the information sensing layer; the external sensing unit includes temperature sensors, pressure sensors, and vibration sensors; the temperature sensors collect freezer temperature data through temperature sensors deployed at preset locations; the pressure sensors collect freezer pressure data through pressure sensors deployed at preset locations; the vibration sensors collect freezer vibration data through vibration acceleration sensors deployed at preset locations.

[0009] Step 1.2: The refrigerant data, capillary thermal data, and mixing characteristic data of the freezer refrigeration system are collected by the refrigeration sensing unit of the information sensing layer; the refrigeration sensing unit collects refrigerant data through a microwave resonant sensor placed in the evaporator inlet pipe; the capillary thermal data is collected through a fiber optic grating sensor and input into a preset pipe structure analytical model to output the thermodynamic coupling data of the capillary; the mixing characteristic data of the lubricating mixture, including dielectric constant and conductivity, are collected through a built-in parallel plate capacitance sensor integrated into the compressor lubricating oil return port.

[0010] Step 2: Acquire external sensor data and identify the external sensor data using preset data processing technology to obtain data stability indicators, perform health assessment on the data stability indicators to obtain data health signals, and send the data health signals to Step 3;

[0011] As a further improvement to the present invention, a health assessment of data stability indicators is performed, the specific implementation of which is as follows:

[0012] Data stability indicators are identified to obtain data fluctuation amplitude, frequency of abnormal jumps, and physical indicators. Each indicator is normalized and mapped to the [0,1] interval to obtain sub-indicators. Each sub-indicator is input into a preset dynamic adjustment factor calculation formula to obtain the corresponding dynamic adjustment factor. Each sub-indicator and the corresponding dynamic adjustment factor are substituted into a preset two-factor dynamic correlation weighted formula to obtain the data health anomaly value. When the data health anomaly value is greater than the preset health anomaly upper limit, a refrigeration system data retrieval signal is generated; otherwise, a retrieval stop signal is generated.

[0013] Step 3: Perform fault detection analysis on external sensor data to obtain the total value of external sensor anomalies; output a refrigeration system data retrieval signal based on the data health signal, and detect the refrigeration system data retrieval signal through a preset signal detection unit. When the refrigeration system data retrieval signal is detected, execute step 3.1 to obtain fault detection evaluation information; when the retrieval stop signal is detected, execute step 3.2 to obtain fault detection evaluation information.

[0014] As a further improvement to the present invention, fault detection and analysis are performed on external sensor data, and the specific analysis content is as follows:

[0015] The freezer temperature data, freezer pressure data, and freezer vibration data are obtained based on external sensor data.

[0016] The current freezer temperature is obtained from the freezer temperature data. The corresponding set freezer temperature is retrieved from the database. The difference between the current and set freezer temperatures is calculated mathematically and recorded as the temperature difference. When the temperature difference exceeds the preset maximum allowable temperature difference, the corresponding moment is marked as an over-temperature moment, and the duration of the over-temperature moment is recorded as the over-temperature duration. The absolute value of the temperature difference is then compared with the maximum allowable temperature difference, and this ratio is recorded as the temperature fluctuation ratio. Both the current over-temperature duration and the temperature fluctuation ratio are input into a preset temperature risk coupling model. Calculate and output the temperature fault risk value TR; where, These are the temperature difference fluctuation ratio and the over-temperature duration, respectively. This is the preset time decay factor.

[0017] The refrigeration system pressure is obtained from the freezer pressure data. The refrigeration system pressure is monitored, and when the pressure exceeds the preset standard pressure range, the corresponding time is marked as a pressure jump moment. The number of pressure jump moments within a unit time period is counted to obtain the pressure jump count, and the historical average pressure jump rate is obtained. The pressure jump count and the average pressure jump count are then input into a preset Poisson probability verification model. Calculate the output pressure failure risk value PR; where... These represent the number of pressure jumps and the average pressure jump rate, respectively; Nb is the number of pressure jumps per unit time period; P represents the probability value of a specific abnormal event occurring, that is, the probability that the freezer pressure will experience a specified number of abnormal jumps within a unit time period.

[0018] Based on the real-time vibration values ​​obtained from the freezer vibration data, a preset number of sliding windows is determined. Vibration values ​​at corresponding continuous measurement times are obtained based on this number of sliding windows. The mean and standard deviation of the sliding window vibration are then mathematically calculated from each vibration value. These values ​​are then input into a preset sliding window abrupt change ratio coupling model. The vibration failure risk value ZR is calculated; where, These represent the values ​​of real-time vibration, mean vibration of the sliding window, and standard deviation of vibration of the sliding window, respectively. This is the preset mutation coefficient.

[0019] The risk values ​​for temperature failure, pressure failure, and vibration failure are input into a preset risk contribution adaptive weighted coupling model. Calculate and output the total value of external sensor anomalies, EA; where... For self-adaptive weights.

[0020] Furthermore, step 3.1: Based on the refrigeration system data and the total value of external sensor anomalies, a comprehensive analysis is performed to obtain fault detection and assessment information;

[0021] As a further improvement to the present invention, the specific execution steps of step 3.1 are as follows:

[0022] Step 3.1.1: Perform fault detection analysis on the refrigeration system data to obtain the total abnormal value of the refrigeration system. Specifically, this involves obtaining refrigerant data, capillary thermal data, and mixing characteristic data based on the refrigeration system data; analyzing the refrigerant data, capillary thermal data, and mixing characteristic data to obtain the gas-liquid ratio risk value, capillary coupling risk value, and mixing characteristic risk value; obtaining the gas-liquid ratio risk value, capillary coupling risk value, and mixing characteristic risk value, and normalizing each risk value to the [0, 1] interval; identifying each risk value; and inputting the normalized value corresponding to the non-zero risk value into the preset geometric mean and arithmetic mean deviation ratio model to calculate and output the total abnormal value of the refrigeration system.

[0023] Furthermore, the analysis of the refrigerant data, capillary thermal data, and mixing characteristic data specifically includes:

[0024] The current gas-liquid ratio is obtained based on refrigerant data. This current ratio is then matched against a preset ideal gas-liquid ratio range. If the current gas-liquid ratio exceeds the ideal range, the portion exceeding this range is marked as an abnormal gas-liquid index. This abnormal gas-liquid index is then divided into multiple abnormal gas-liquid index intervals based on preset intervals. Each abnormal gas-liquid index interval is assigned an abnormal gas-liquid amplification factor, which is greater than zero and positively correlated with the median index of the abnormal gas-liquid index interval. The current abnormal gas-liquid index is matched against each abnormal gas-liquid index interval to obtain the corresponding abnormal gas-liquid amplification factor. Finally, the abnormal gas-liquid index and the abnormal gas-liquid amplification factor are multiplied to calculate and output the gas-liquid ratio risk value.

[0025] The inlet and outlet temperature difference and inlet and outlet pressure difference of the capillary are obtained based on capillary thermal data; the dimensionless values ​​of the inlet and outlet temperature difference and inlet and outlet pressure difference are input into the preset energy imbalance coupling model for calculation and output of capillary coupling risk value.

[0026] Based on the mixing characteristic data, the dielectric constant and conductivity of the mixture are obtained from the actual measured data. Reference dielectric constant and reference conductivity, as well as the corresponding critical dielectric constant and critical conductivity, are extracted from the preset health status of the refrigeration system in the database. The dielectric constant, reference dielectric constant, and critical dielectric constant are input into a preset standardized deviation calculation formula to calculate and output the dielectric constant deviation value. Similarly, the conductivity deviation values ​​of the conductivity, reference conductivity, and critical conductivity are obtained. Both the dielectric constant deviation value and the conductivity deviation value are input into a preset cooperative deviation model to calculate and output the mixing characteristic risk value.

[0027] Step 3.1.2: A comprehensive analysis of the total external sensor anomaly value and the total refrigeration system anomaly value is performed to obtain fault detection and assessment information. Specifically, the analysis involves assigning a comprehensive weighting factor to both the total external sensor anomaly value and the total refrigeration system anomaly value, with the comprehensive weighting factor corresponding to the total refrigeration system anomaly value being greater than that corresponding to the total external sensor anomaly value. The total external sensor anomaly value and the total refrigeration system anomaly value are then substituted into a preset comprehensive calculation formula to calculate and output a comprehensive anomaly value. Finally, a preset fault detection and assessment mechanism is used to evaluate the comprehensive anomaly value to obtain fault detection and assessment information.

[0028] Furthermore, step 3.2: Outputting fault detection and evaluation information based on the total value of external sensor anomalies, specifically: directly outputting the total value of external sensor anomalies as the comprehensive detection anomaly value, and evaluating the comprehensive detection anomaly value through the fault detection and evaluation mechanism to obtain fault detection and evaluation information.

[0029] Furthermore, the fault detection and evaluation mechanism is as follows: An input step to obtain the comprehensive outlier value is taken; based on the input step, corresponding interval values ​​are set; based on the interval values ​​corresponding to the comprehensive outlier value, three comprehensive outlier value intervals are obtained, denoted as ZA1, ZA2, and ZA3 respectively; the current comprehensive outlier value ZA is matched with each comprehensive outlier value interval; when ZA∈ZA1, the fault detection and evaluation information is generated as low-level outlier; when ZA∈ZA2, the fault detection and evaluation information is generated as medium-level outlier; and when ZA∈ZA3, the fault detection and evaluation information is generated as high-level outlier.

[0030] Step 4: Generate corresponding fault alarm strategies based on fault detection and assessment information, and execute them accordingly. Specifically, the fault detection and assessment information is identified. If the corresponding level is low, the fault alarm strategy is to display it on the freezer's visual screen and store it in the local log. If the duration of the low-level fault exceeds the preset duration, it is upgraded to a medium-level fault. If the corresponding level is medium, the fault alarm strategy is to provide low-frequency audible and visual warnings and remote notifications. If the duration of the medium-level fault exceeds the preset duration, it is upgraded to a high-level fault. If the corresponding level is high, the fault alarm strategy is to provide high-frequency audible and visual warnings and multi-level remote notifications.

[0031] The beneficial effects of the technical solution provided by this invention compared with the prior art are as follows:

[0032] 1. This invention collects and outputs external sensor data and refrigeration system data through an information sensing layer, which solves the shortcomings of a single external sensor. Furthermore, the external sensing unit collects temperature, pressure, and vibration data, while the refrigeration sensing unit specifically collects data on refrigerant gas-liquid ratio, capillary thermodynamics, and lubricating oil mixture characteristics. When the external sensor malfunctions, fault detection can be achieved through refrigeration system data, improving the stability and reliability of detection and avoiding detection failures caused by sensor malfunctions.

[0033] 2. This invention identifies data stability indicators by identifying external sensor data, and performs health assessments on the data stability indicators to obtain data health signals. It comprehensively analyzes the total abnormal value of external sensors and the total abnormal value of the refrigeration system to classify the abnormality level and generate corresponding fault alarm strategies. This can match the severity of the fault, reduce false alarms and missed alarms, and ensure the safe operation of the freezer. Attached Figure Description

[0034] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. The following drawings are not deliberately drawn to scale according to the actual size, but are intended to show the main idea of ​​this application.

[0035] Figure 1 This is a flowchart of a method for detecting refrigeration faults in an ultra-low temperature freezer according to the present invention. Detailed Implementation

[0036] 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 some embodiments of the present invention, and not all embodiments. 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.

[0037] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 In one embodiment of the present invention, a method for detecting refrigeration faults in an ultra-low temperature freezer includes:

[0038] Step 1: Basic Construction of Fault Detection and Analysis Information Source: Based on the information sensing layer of the freezer, information is collected and output to external sensor data and refrigeration system data; Step 2 is executed on the external sensor data, and the refrigeration system data is sent to Step 3; the specific construction content is as follows:

[0039] Step 1.1, External sensor data acquisition: The external sensor unit of the information sensing layer acquires the freezer temperature data, freezer pressure data, and freezer vibration data.

[0040] External sensing units include temperature sensors, pressure sensors, and vibration sensors;

[0041] Temperature sensors obtain the temperature data of the freezer by deploying temperature sensors at preset locations (for example, by placing resistive or patch temperature sensors at key points of the freezer such as the freezer shelves, corners of the cabinet, and condenser outlets to obtain the corresponding temperature data).

[0042] Pressure sensors obtain freezer pressure data by collecting pressure data through pressure sensors deployed at preset locations (for example, installing pressure sensors at the high and low pressure sides of the refrigeration system and at both ends of the capillary tube in the freezer pressure distribution area and collecting the corresponding pressure data).

[0043] Vibration sensors obtain vibration data of the freezer by collecting vibration data through vibration acceleration sensors deployed at preset locations (for example, installing vibration acceleration sensors on the compressor housing and freezer body and collecting corresponding vibration data).

[0044] Step 1.2, Refrigeration System Data Acquisition: The refrigerant data, capillary thermal data, and mixing characteristic data of the freezer refrigeration system are acquired through the refrigeration sensing unit of the information sensing layer;

[0045] The refrigeration sensing unit acquires refrigerant data through a microwave resonant sensor placed in the evaporator inlet pipe;

[0046] Capillary thermal data is obtained by acquiring data from the capillary circuit using a fiber optic grating sensor. The capillary thermal data is then input into a preset pipeline structure analytical model (the pressure-temperature coupling data, i.e., thermodynamic coupling data, is calculated by the ratio of pressure and temperature in the capillary thermal data based on a preset coupling coefficient). The capillary thermal data is then output as thermodynamic coupling data.

[0047] Data on the mixing characteristics of the lubricating mixture, including dielectric constant and conductivity, are collected by a built-in parallel plate capacitive sensor integrated into the compressor's lubricating oil return port.

[0048] Step 2, Data Source Evaluation and Analysis: Acquire external sensor data and identify the external sensor data through preset data processing technology to obtain data stability indicators. Then, conduct a health assessment on the data stability indicators to obtain data health signals and send the data health signals to Step 3.

[0049] The health assessment of data stability indicators involves the following steps: identifying data stability indicators to obtain data fluctuation amplitude, frequency of abnormal jumps, and physical indicators; normalizing each indicator and mapping it to the [0, 1] interval to obtain sub-indicators; and labeling the normalized sub-indicators corresponding to data fluctuation amplitude, frequency of abnormal jumps, and physical indicators as Bo, Tp, and Wx, respectively; and inputting each sub-indicator into a preset dynamic adjustment factor calculation formula. The corresponding dynamic adjustment factor is obtained through calculation. Where zi∈{Bo, Tp, Wx}; k is a preset adjustment coefficient with a value of 0.5, used to keep the value of the dynamic adjustment factor within the range of [1, 1.5] to avoid excessive weight fluctuations; substitute each sub-indicator and its corresponding dynamic adjustment factor into the preset two-factor dynamic correlation weighting formula. The data health anomaly value JD is calculated; where, Each sub-indicator has a pre-set base weight value, which is determined based on domain experience (for example, the sum of all base weight values ​​is 1). (The values ​​are 0.3, 0.4, and 0.3 respectively). When the abnormal data value is greater than the preset abnormal data value limit, a data retrieval signal for the refrigeration system is generated; otherwise, a retrieval stop signal is generated.

[0050] The data processing technology involves mathematically calculating the standard deviation and mean of external sensor data corresponding to a preset time period; calculating a data ratio by comparing the absolute values ​​of the standard deviation and mean of the current preset unit time period, and marking this data ratio as the data fluctuation amplitude; calculating the difference between adjacent sampling data points within a unit time period, and comparing this difference with the corresponding preset multiple of the mean to output the frequency of abnormal jumps (for example, if the temperature difference between two adjacent sampling points is greater than twice the mean temperature data, the corresponding sampling point is marked as an abnormal jump, and the frequency of abnormal jumps is statistically calculated); comparing each external sensor data with the pre-stored physical reasonableness range values ​​in the database, and marking the portion exceeding the corresponding preset physical reasonableness range value as a physical indicator; and integrating the data fluctuation amplitude, abnormal jump frequency, and physical indicators of each external sensor data to obtain a data stability index.

[0051] Step 3, Fault Detection and Evaluation Analysis: Perform fault detection analysis on the external sensor data to obtain the total value of external sensor anomalies; output the refrigeration system data retrieval signal based on the data health signal, and detect the refrigeration system data retrieval signal through a preset signal detection unit. When the refrigeration system data retrieval signal is detected, execute step 3.1 to obtain fault detection and evaluation information; when the retrieval stop signal is detected, execute step 3.2 to obtain fault detection and evaluation information.

[0052] The specific analysis content for fault detection analysis of external sensor data is as follows:

[0053] The freezer temperature data, freezer pressure data, and freezer vibration data are obtained based on external sensor data.

[0054] The current freezer temperature is obtained from the freezer temperature data. The corresponding set freezer temperature is retrieved from the database. The difference between the current and set freezer temperatures is calculated mathematically and recorded as the temperature difference. When the temperature difference exceeds the preset maximum allowable temperature difference, the corresponding moment is marked as an over-temperature moment, and the duration of the over-temperature moment is recorded as the over-temperature duration. The absolute value of the temperature difference is then compared to the maximum allowable temperature difference, and this ratio is recorded as the temperature fluctuation ratio. Both the current over-temperature duration and the temperature fluctuation ratio are input into a preset temperature risk coupling model (calculated by coupling the deviation ratio of temperature from the threshold with a time decay factor). Calculate and output the temperature fault risk value TR; where, These are the temperature difference fluctuation ratio and the over-temperature duration, respectively. The preset time decay factor is set by domain experience and is 0.02. It is used to quantify the characteristic that the longer the over-temperature duration, the faster the risk accumulates. For example, when the temperature difference is 4℃ (the maximum allowable temperature difference is 4℃) and the over-temperature duration is 10 minutes, then TR≈1×1.22=1.22.

[0055] The refrigeration system pressure is obtained from the freezer pressure data. The refrigeration system pressure is monitored, and when the pressure exceeds the preset standard pressure range, the corresponding time is marked as a pressure jump moment. The number of pressure jump moments within a unit time period is counted to obtain the pressure jump count, and the historical average pressure jump rate is obtained. The pressure jump count and the average pressure jump count are then input into a preset Poisson probability verification model. Calculate the output pressure failure risk value PR; where... These represent the number of pressure jumps and the average pressure jump rate, respectively; Nb is the number of pressure jumps per unit time period; P represents the probability value of a specific abnormal event occurring, that is, the probability that the freezer pressure will experience a specified number of abnormal jumps within a unit time period.

[0056] Based on the real-time vibration values ​​obtained from the freezer vibration data, a preset number of sliding windows is determined. Vibration values ​​at corresponding continuous measurement times are obtained based on this number of sliding windows. The mean and standard deviation of the sliding window vibration are then mathematically calculated from each vibration value. These values ​​are then input into a preset sliding window abrupt change ratio coupling model. The vibration failure risk value ZR is calculated; where, These represent the values ​​of real-time vibration, mean vibration of the sliding window, and standard deviation of vibration of the sliding window, respectively. The preset mutation coefficient, which is set to 2, is based on domain experience and is used to control the sensitivity of local mutation detection.

[0057] The risk values ​​for temperature failure, pressure failure, and vibration failure are input into a preset risk contribution adaptive weighted coupling model. Calculate and output the total value of external sensor anomalies, EA; where... These are adaptive weights, used to output adaptive weights based on the proportion of a single risk value. These are the numerical values ​​corresponding to the risk values ​​for temperature failure, pressure failure, and vibration failure, respectively; it should be noted that... When all values ​​are equal to zero, the total value of external sensor anomalies is directly output as 0.

[0058] Step 3.1: Based on the refrigeration system data and the total abnormal values ​​of external sensors, a comprehensive analysis is performed to obtain fault detection and assessment information. The specific steps are as follows:

[0059] Step 3.1.1: Perform fault detection and analysis on the refrigeration system data to obtain the total abnormal value of the refrigeration system, specifically as follows:

[0060] Refrigerant data, capillary thermal data, and mixing characteristic data are obtained from refrigeration system data;

[0061] Based on refrigerant data, the current gas-liquid ratio is obtained. The current gas-liquid ratio is matched with a preset ideal gas-liquid ratio range. If the current gas-liquid ratio exceeds the ideal gas-liquid ratio range, the portion of the gas-liquid ratio exceeding the ideal range is marked as an abnormal gas-liquid index. The abnormal gas-liquid index is divided equally based on a preset index interval to obtain multiple abnormal gas-liquid index intervals. Each abnormal gas-liquid index interval is assigned an abnormal gas-liquid amplification coefficient (used to amplify the influence of high abnormal gas-liquid indices). The abnormal gas-liquid amplification coefficient is greater than zero and positively correlated with the median index of the abnormal gas-liquid index interval. The current abnormal gas-liquid index is matched with each abnormal gas-liquid index interval to obtain the corresponding abnormal gas-liquid amplification coefficient. The abnormal gas-liquid index and the abnormal gas-liquid amplification coefficient are multiplied to calculate and output the gas-liquid ratio risk value.

[0062] The inlet and outlet temperature difference and inlet and outlet pressure difference of the capillary are obtained based on capillary thermal data; the dimensionless values ​​of the inlet and outlet temperature difference and inlet and outlet pressure difference are then input into a preset energy imbalance coupling model. Calculate and output the capillary coupling risk value mk; where, These are the values ​​for the temperature difference between the inlet and outlet and the pressure difference between the inlet and outlet, respectively. These are the pre-designed inlet and outlet temperature difference thresholds and inlet and outlet pressure difference thresholds, respectively; The preset thermodynamic coupling coefficient is set based on the proportional constant of temperature and pressure changes under normal operating conditions.

[0063] Based on the mixing characteristic data, the dielectric constant and conductivity of the mixture are obtained from the actual measured data. Reference dielectric constant and reference conductivity, as well as the corresponding critical dielectric constant and critical conductivity, are extracted from the preset health status of the refrigeration system in the database. The dielectric constant, reference dielectric constant, and critical dielectric constant are input into a preset standardized deviation calculation formula to calculate and output the dielectric constant deviation value. Similarly, the conductivity deviation values ​​of the conductivity, reference conductivity, and critical conductivity are obtained. Both the dielectric constant deviation value and the conductivity deviation value are input into a preset cooperative deviation model to calculate and output the mixing characteristic risk value.

[0064] The above analysis uses the following formula:

[0065] Standardized deviation calculation formula: ;

[0066] Collaborative Deviation Model: ;

[0067] in, Let these be the dielectric constant and conductivity, respectively. These are the reference dielectric constant and critical dielectric constant, and the reference conductivity and critical conductivity, respectively; These are the dielectric constant deviation value and conductivity deviation value, respectively; mr is the risk value of mixed characteristics.

[0068] The gas-liquid ratio risk value, capillary coupling risk value, and mixing characteristic risk value are obtained, and each risk value is normalized to the interval [0, 1]. Each risk value is identified, and the normalized value corresponding to the non-zero risk value is input into a preset geometric mean vs. arithmetic mean deviation ratio model. The total abnormal value SA of the refrigeration system is calculated and output; where gm and am represent the geometric mean and arithmetic mean, respectively; and qy is the gas-liquid ratio risk value.

[0069] Step 3.1.2: Perform a comprehensive analysis of the total abnormal values ​​of the external sensors and the total abnormal values ​​of the refrigeration system to obtain fault detection and evaluation information. Specifically, assign a comprehensive weighting factor to both the total abnormal values ​​of the external sensors and the total abnormal values ​​of the refrigeration system, ensuring that the comprehensive weighting factor corresponding to the total abnormal value of the refrigeration system is greater than that corresponding to the total abnormal value of the external sensors. Substitute the total abnormal values ​​of the external sensors and the total abnormal values ​​of the refrigeration system into a preset comprehensive calculation formula. The comprehensive outlier ZA is calculated and output; where, These are the comprehensive weighting factors corresponding to the total abnormal values ​​of external sensors and the total abnormal values ​​of the refrigeration system, respectively; the fault detection and evaluation information is obtained by evaluating the comprehensive abnormal values ​​through a preset fault detection and evaluation mechanism.

[0070] Step 3.2: Output fault detection and evaluation information based on the total value of external sensor anomalies. Specifically, the total value of external sensor anomalies is directly output as the comprehensive detection anomaly value, and the comprehensive detection anomaly value is evaluated through the fault detection and evaluation mechanism to obtain the fault detection and evaluation information.

[0071] The fault detection and evaluation mechanism is as follows: An input step to obtain the comprehensive anomaly value is taken. Based on this input step, corresponding interval values ​​are set (i.e., different interval values ​​are assigned to the comprehensive anomaly values ​​output in steps 3.1.2 and 3.2 respectively). Based on the interval values ​​corresponding to the comprehensive anomaly values, three intervals of comprehensive anomaly values ​​are obtained, denoted as ZA1, ZA2, and ZA3 respectively (the minimum values ​​of the intervals corresponding to ZA1, ZA2, and ZA3 show an increasing trend). The current comprehensive anomaly value is matched with each interval. When ZA∈ZA1, a low-anomaly level fault detection and evaluation information is generated; when ZA∈ZA2, a medium-anomaly level fault detection and evaluation information is generated; and when ZA∈ZA3, a high-anomaly level fault detection and evaluation information is generated.

[0072] Step 4, Fault Alarm and Output: Based on the fault detection and evaluation information, a corresponding fault alarm strategy is generated and executed. Specifically, the fault detection and evaluation information is identified. If it corresponds to a low anomaly level, the fault alarm strategy is to display it on the freezer's visual screen and store it in the local log. If the duration of the low anomaly level exceeds the preset anomaly duration, it is upgraded to a medium anomaly level. If it corresponds to a medium anomaly level, the fault alarm strategy is to provide low-frequency audible and visual warnings and remote notifications. If the duration of the medium anomaly level exceeds the preset anomaly duration, it is upgraded to a high anomaly level. If it corresponds to a high anomaly level, the fault alarm strategy is to provide high-frequency audible and visual warnings and multi-level remote notifications (e.g., SMS, dial-up, and pop-up notifications to maintenance personnel and responsible persons).

[0073] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting refrigeration faults in ultra-low temperature freezers, characterized in that, Comprising the following steps: Step 1: based on the information perception layer of the refrigerator, information collection and output of external sensing data and refrigeration system data are performed; external sensing data is executed in step 2, and refrigeration system data is sent to step 3; Step 2: obtaining external sensing data, identifying the external sensing data to obtain data stability index, and then health evaluating the data stability index to obtain data health signal, and sending it to step 3; Step 3: fault detection analysis is performed on the external sensing data to obtain external sensing abnormal total value; Based on the data health signal, refrigeration system data retrieval signal and retrieval stop signal are outputted, and they are detected; when the refrigeration system data retrieval signal is detected, step 3.1 is executed to obtain fault detection evaluation information; when the retrieval stop signal is detected, step 3.2 is executed to obtain fault detection evaluation information; Step 3.1: based on the comprehensive analysis of refrigeration system data and external sensing abnormal total value, fault detection evaluation information is obtained; Step 3.2: based on the external sensing abnormal total value, fault detection evaluation information is outputted; Step 4: based on the fault detection evaluation information, corresponding fault alarm strategy is generated and executed; Wherein, the health evaluation of the data stability index to obtain the data health signal, its specific execution content is: identifying the data stability index to obtain data fluctuation amplitude, abnormal jump frequency and physical index, normalizing each index to map [0, 1] interval to obtain sub-index; each sub-index is inputted into the preset dynamic adjustment factor calculation formula to obtain the corresponding dynamic adjustment factor; Each sub-index and the corresponding dynamic adjustment factor are substituted into the preset double-factor dynamic correlation weighting formula to obtain the data health abnormal value; when the data health abnormal value is greater than the preset health abnormal upper limit, the refrigeration system data retrieval signal is generated; otherwise, the retrieval stop signal is generated.

2. The method of claim 1, wherein the method further comprises: The information collection and output of external sensing data and refrigeration system data based on the information perception layer of the refrigerator are as follows: Step 1.1: through the external sensing unit of the information perception layer, the refrigerator temperature data, the refrigerator pressure data and the refrigerator vibration data of the refrigerator are collected; the external sensing unit includes temperature sensor, pressure sensor and vibration sensor; the temperature sensor collects the temperature data of the refrigerator through the temperature sensor deployed at the preset position; the pressure sensor collects the pressure data of the refrigerator through the pressure sensor deployed at the preset position; the vibration sensor collects the vibration data of the refrigerator through the vibration acceleration sensor deployed at the preset position; Step 1.2: Collecting refrigerant data, capillary heat data, and mixed property data of the refrigerator refrigeration system through the refrigeration sensing unit of the information sensing layer; the refrigeration sensing unit collects refrigerant data through a microwave resonance sensor placed in the evaporator inlet pipeline; capillary heat data is collected through a fiber bragg grating sensor on the capillary pipeline, and the capillary heat data is input into a preset pipeline structure analysis model to output thermodynamic coupling data of the capillary; mixed property data of the lubricating mixture including dielectric constant and conductivity are collected through an internal parallel plate capacitor sensor integrated in the compressor lubricating oil return port.

3. The method of claim 2, wherein the method further comprises: The external sensor abnormal total value is obtained by performing fault detection analysis on the external sensor data, and the specific analysis content is as follows: According to the external sensor data, obtain the refrigerator temperature data, refrigerator pressure data and refrigerator vibration data; According to the refrigerator temperature data, obtain the current refrigerator temperature, obtain the corresponding current refrigerator set temperature based on the database, and perform mathematical calculation on the numerical values of the current refrigerator temperature and the refrigerator set temperature to obtain a difference value, which is recorded as a temperature difference value; when the temperature difference value exceeds the preset maximum allowed temperature difference, the corresponding time is marked as an over-temperature time, and the over-temperature time duration is obtained by counting the over-temperature time duration; Take the absolute value of the temperature difference value and the maximum allowed temperature difference to calculate a ratio, and record it as a temperature difference fluctuation ratio; the numerical values of the current over-temperature time duration and the temperature difference fluctuation ratio are input into a preset temperature risk coupling model to calculate and output a temperature fault risk value; According to the refrigerator pressure data, obtain the refrigeration system pressure value, monitor the refrigeration system pressure, and when the refrigeration system pressure value exceeds the preset standard pressure interval, mark the corresponding time as a pressure jump time, count the number of pressure jump times in a unit time period to obtain a pressure jump number, and obtain the average pressure jump rate in history; the pressure jump number and the average pressure jump number are input into a preset Poisson probability verification model to calculate and output a pressure fault risk value; According to the real-time vibration value obtained from the refrigerator vibration data, obtain a preset number of sliding window quantities, obtain the vibration values of the corresponding number of continuous measurement times based on the number of sliding window quantities, and perform mathematical calculation on each vibration value to obtain the sliding window vibration mean value and the sliding window vibration standard deviation; the numerical values of the real-time vibration value, the sliding window vibration mean value and the sliding window vibration standard deviation are input into a preset sliding window mutation ratio coupling model to calculate and obtain a vibration fault risk value; The temperature fault risk value, the pressure fault risk value and the vibration fault risk value are input into a preset risk contribution degree self-adaptive weight coupling model to calculate and output an external sensor abnormal total value.

4. The method of claim 3, wherein the method further comprises: The specific execution steps of step 3.1 are as follows: Step 3.1.1: performing fault detection analysis on the refrigeration system data to obtain a refrigeration system abnormal total value; Step 3.1.2: performing comprehensive analysis on the external sensor abnormal total value and the refrigeration system abnormal total value to obtain fault detection evaluation information.

5. The method of claim 4, wherein the method further comprises: The specific analysis content of the refrigeration system data is as follows: According to the refrigeration system data, refrigerant data, capillary heat data and mixture characteristic data are obtained; Based on the refrigerant data, a current gas-liquid ratio is obtained, and the current gas-liquid ratio is matched with a preset ideal gas-liquid ratio interval. If the current gas-liquid ratio exceeds the ideal gas-liquid ratio interval, the gas-liquid ratio value of the part exceeding the ideal gas-liquid ratio interval is marked as an abnormal gas-liquid index. The abnormal gas-liquid index is equally divided into multiple abnormal gas-liquid index intervals based on a preset index interval. Each abnormal gas-liquid index interval is assigned an abnormal gas-liquid amplification coefficient. The value of the abnormal gas-liquid amplification coefficient is greater than zero and positively correlated with the middle index of the abnormal gas-liquid index interval. The current abnormal gas-liquid index is matched with each abnormal gas-liquid index interval to obtain the corresponding abnormal gas-liquid amplification coefficient, and the product of the abnormal gas-liquid index and the abnormal gas-liquid amplification coefficient is calculated to output a gas-liquid ratio risk value. Based on the capillary heat data, the inlet and outlet temperature difference and the inlet and outlet pressure difference of the capillary are obtained. The dimensionless values of the inlet and outlet temperature difference and the inlet and outlet pressure difference are input into a preset energy imbalance coupling model to calculate and output a capillary coupling risk value. Based on the mixture characteristic data, the current actual measured dielectric constant and conductivity of the mixture are obtained. Based on the preset health state of the refrigeration system in the database, the reference dielectric constant and the reference conductivity, as well as the corresponding critical dielectric constant and the critical conductivity, are extracted. The dielectric constant, the reference dielectric constant and the critical dielectric constant are input into a preset standard deviation calculation formula to calculate and output a dielectric constant deviation value. Similarly, the conductivity deviation value of the conductivity, the reference conductivity and the critical conductivity is obtained. The dielectric constant deviation value and the conductivity deviation value are input into a preset cooperative deviation model to calculate and output a mixture characteristic risk value. The gas-liquid ratio risk value, the capillary coupling risk value and the mixture characteristic risk value are obtained, and each risk value is normalized to the [0, 1] interval. The non-zero risk value corresponding to the normalized value is input into a preset deviation ratio model of the geometric mean and the arithmetic mean to calculate and output a refrigeration system abnormal total value.

6. The method of claim 5, wherein the method further comprises: The external sensor abnormal total value and the refrigeration system abnormal total value are comprehensively analyzed to obtain fault detection evaluation information. The specific analysis content is that the external sensor abnormal total value and the refrigeration system abnormal total value are respectively assigned a comprehensive weight factor, and the value of the comprehensive weight factor corresponding to the refrigeration system abnormal total value is greater than the value of the comprehensive weight factor corresponding to the external sensor abnormal total value. The external sensor abnormal total value and the refrigeration system abnormal total value are substituted into a preset comprehensive calculation formula to calculate and output a detection comprehensive abnormal value. The detection comprehensive abnormal value is evaluated by a preset fault detection evaluation mechanism to obtain the fault detection evaluation information.

7. The method of claim 6, wherein the method further comprises: The external sensor abnormal total value is directly output as the detection comprehensive abnormal value, and the detection comprehensive abnormal value is evaluated by the fault detection evaluation mechanism to obtain the fault detection evaluation information.

8. The method of claim 7, wherein the method further comprises: The fault detection evaluation mechanism is: an input step of detecting comprehensive abnormal values, setting corresponding division interval values based on the input step, dividing three detection comprehensive abnormal value intervals based on the corresponding division interval values of the detection comprehensive abnormal values, and recording them as ZA1, ZA2, and ZA3 respectively, matching the current corresponding detection comprehensive abnormal value ZA with each detection comprehensive abnormal value interval, when ZA∈ZA1, generating fault detection evaluation information as low abnormal level; when ZA∈ZA2, generating fault detection evaluation information as medium abnormal level; when ZA∈ZA3, generating fault detection evaluation information as high abnormal level.

9. The method of claim 8, wherein the method further comprises: The fault detection evaluation information is generated based on the corresponding fault alarm strategy, and the fault alarm strategy is executed based on the fault alarm strategy. Specifically, the fault detection evaluation information is identified. If it corresponds to a low abnormal level, the fault alarm strategy is output to the refrigerator visual screen for display, and is stored in the local log. If the low abnormal level lasts for more than the preset abnormal duration, it is upgraded to a medium abnormal level. If it corresponds to a medium abnormal level, the fault alarm strategy is output to perform low-frequency sound and light warning and remote notification. If the medium abnormal level lasts for more than the preset abnormal duration, it is upgraded to a high abnormal level. If it corresponds to a high abnormal level, the fault alarm strategy is output to perform high-frequency sound and light warning and multi-level remote notification.

Images