Sealing method, system and device for two-phase immersion liquid cooling data center

Abstract

The invention discloses a sealing method, system and device for a two-phase immersion liquid cooling data center, and relates to the technical field of data center sealing processing, and the method comprises the steps: 1, collecting sealing data in real time, carrying out the data preprocessing, and carrying out the real-time transmission and standardized storage of the sealing data; 2, the sealing state is judged in real time according to the received sealing data, and self-adaptive sealing compensation adjustment measures are taken; 3, the adjusted sealing data are collected in real time, adjustment deviation judgment of self-adaptive sealing compensation adjustment is carried out, and deviation correction measures are taken based on the adjustment deviation judgment result; and 4, performing visual intelligent response based on the sealing data and the sealing state before and after the sealing compensation adjustment. The problem that long-term good sealing is difficult to achieve due to the fact that a sealing piece dynamically changes from loose to tight and then to loose due to working condition changes of pressure, temperature and vibration fluctuation in the operation process of the data center is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of data center sealing treatment, and particularly to a sealing method, system and device for a two-phase immersion liquid-cooled data center. Background Technique

[0002] With the development of big data, cloud computing and high-performance computing, the power consumption of data centers has been continuously rising. The traditional air-cooled heat dissipation method has been difficult to meet the requirements of high-density servers for cooling efficiency and energy consumption control. Due to its superior heat transfer performance, noise reduction and energy-saving advantages, the two-phase immersion liquid-cooling technology has gradually become an important cooling solution for the new generation of green data centers. In the application of two-phase immersion liquid-cooled data centers, the long-term reliability of the sealing system is directly related to the safe operation of the equipment. Commonly used coolants (such as fluorinated liquids) are volatile and have certain chemical corrosiveness. Once the seal fails and the coolant leaks, it will not only cause equipment short circuits and data loss, but may also pose serious hazards to human health and environmental safety.

[0003] For example, the invention patent with the publication number CN103838337B discloses a data center server cooling cabinet, which includes a server cabinet, a phase change heat transfer system, a server housing and a spring piece for fixing the server housing; the phase change heat transfer system includes a plurality of heat exchangers installed in the server cabinet, an outdoor cooling device and a connecting pipeline; the plurality of heat exchangers are respectively installed on each layer of brackets for placing servers in the server cabinet; the server housing is placed on top of the plurality of heat exchangers; the server is sealed inside the server housing; the entire data center server cooling cabinet has no air cooling, the server heat generation unit is in direct contact with the server housing, and the server housing is in direct contact with the plurality of heat exchangers installed in the server cabinet, achieving contact heat transfer and improving the heat transfer efficiency; the data center server cooling cabinet of the present invention can increase the placement density of servers, has a low construction cost, a short construction period, and greatly reduces the operating cost of the entire data center.

[0004] For example, the invention patent with the publication number CN113849054A discloses an immersion liquid cooling heat dissipation system applied to a data center, which includes a cooling box body with a coolant built therein. A rotating shaft is sealed and rotatably arranged in the cooling box body. A cooling air chamber for the coolant vapor to enter is fixed at the upper end of the rotating shaft. The rotation of the rotating shaft is controlled by a first driving mechanism. A cooling disturbance mechanism is arranged on the cooling air chamber. The cooling disturbance mechanism includes a heat dissipation plate hinged on the side wall of the cooling air chamber and capable of flipping up and down. An elastic contraction member with both ends hermetically connected to both of them is arranged between the heat dissipation plate and the cooling air chamber. A cooling flow channel communicated with the cooling air chamber through the elastic contraction member is arranged in the heat dissipation plate. The up and down flipping of the heat dissipation plate is controlled by a second driving mechanism. A fan blade for disturbing the air on the side wall of the cooling box body as the cooling air chamber rotates is arranged on the side of the heat dissipation plate close to the cooling box body. This immersion liquid cooling heat dissipation system can significantly reduce energy consumption and equipment costs, and is safer and more environmentally friendly.

[0005] However, in the process of implementing the technical solution of the invention in the embodiments of the present application, it is found that the above technology has at least the following technical problems: In the actual operating environment of a two-phase immersion liquid cooling data center, the sealing device often needs to withstand the repeated fluctuations of the coolant pressure, the periodic temperature changes caused by the heat generated by server equipment, as well as the vibrations of cabinets, pipelines and external mechanical disturbances and other multiple complex working conditions for a long time. These factors will cause the seal to continuously experience dynamic force changes from loose to tight and then to loose during the working process, resulting in the inability to ensure long-term good sealing.

[0006] Therefore, in view of the above problems, there is an urgent need for a sealing method, system and device for a two-phase immersion liquid cooling data center. Summary of the Invention

[0007] Technical problems to be solved Aiming at the deficiencies of the prior art, the present invention provides a sealing method, system and device for a two-phase immersion liquid cooling data center, which solves the problem that during the operation of the data center, the seal undergoes dynamic changes from loose to tight and then to loose due to the changes in working conditions such as pressure, temperature and vibration fluctuations, resulting in difficulty in achieving long-term good sealing.

[0008] Technical solution: To achieve the above objectives, the present invention is realized through the following technical solutions: A sealing method for a two-phase immersion liquid-cooled data center, including: Step 1, real-time collect sealing data through multi-type integrated sensors, perform data preprocessing on the real-time sealing data, establish a sealing database, and perform real-time transmission and standardized storage of the sealing data according to time and location information; Step 2, receive the preprocessed real-time sealing data and perform structured arrangement, judge the sealing state in real time according to the received sealing data, and take adaptive sealing compensation adjustment measures according to the sealing state; Step 3, receive the sealing compensation adjustment instruction in real time, and real-time collect the adjusted sealing data, judge the adjustment deviation of the adaptive sealing compensation adjustment according to the adjusted sealing data, and implement deviation correction measures based on the adjustment deviation judgment result; Step 4, perform visual intelligent response based on the sealing data and sealing state before and after the sealing compensation adjustment.

[0009] Further, the specific process of real-time collecting sealing data through multi-type integrated sensors, performing data preprocessing on the real-time sealing data, establishing a sealing database, and performing real-time transmission and standardized storage of the sealing data according to time and location information is as follows: Use multi-type sensors to real-time collect the sealing data of the device, and the sealing data includes: real-time temperature, pressure, vibration data, and displacement data of the adjustment mechanism; Check whether the sealing data is missing, and for the missing sealing data, use the linear interpolation method to complete the sealing data. For the completed sealing data, perform denoising and filtering operations by combining median filtering and mean filtering, identify and remove dynamic outliers of the sealing data through the sliding window detection algorithm, and further perform preliminary normalization processing on the sealing data; Package the preprocessed sealing data according to the time stamp and sensor location, store it in the sealing database, and send it to the data analysis and decision-making module in real time through the bus.

[0010] Further, receive the preprocessed real-time seal data and perform structured collation. The specific process of real-time judging the seal state based on the received seal data is as follows: Receive the real-time seal data from the dynamic sensing module, perform preliminary decoding and verification on the seal data, and perform structured collation according to the source, acquisition time, and data type of the seal data; Obtain the pressure data, and calculate the average value of the pressure data under normal leak-free conditions within one month to obtain the target pressure value; Obtain the temperature data, and calculate the average value of the temperature data under normal leak-free conditions within one month to obtain the target temperature value; Obtain the vibration data, sum the squares of all vibration data within the time window length and then calculate the average value to obtain the root mean square value of the vibration acceleration of the seal measurement point as the average vibration intensity; At the same time, based on the sliding time window, calculate the historical vibration data distribution, and take the 95th percentile value as the vibration standard value; Subtract the current moment pressure from the target pressure value and divide by the target pressure value to obtain the pressure deviation ratio; Subtract the current moment temperature from the target temperature value and divide by the target temperature value to obtain the temperature deviation ratio; Square and sum the pressure deviation ratio and the temperature deviation ratio respectively, and then take the square root to obtain the multi-factor deviation distance; Add the ratio of the average vibration intensity to the vibration standard value to a constant one to obtain the vibration adjustment factor; Multiply the multi-factor deviation distance by the vibration adjustment factor to obtain the seal abnormality judgment value; Store the seal abnormality judgment value in the seal database, and compare the seal abnormality judgment value with the abnormality threshold in real time, and take seal health grading response measures according to the comparison result of the abnormality threshold.

[0011] Further, the specific process of taking seal health grading response measures according to the comparison result of the abnormality threshold is as follows: When the seal abnormality judgment value is less than or equal to the abnormality threshold, it is determined that the seal state is normal, and only routine monitoring and seal data recording are performed, and a health report is generated regularly to support predictive maintenance; When the seal abnormality judgment value is greater than the abnormality threshold, it is determined that the seal state is abnormal, and enter the adaptive compensation adjustment step, automatically encrypt the acquisition and monitoring period of the seal data, and refine the time window; Automatically identify continuous small-amplitude and short-term seal abnormality changes, capture early seal abnormality signals, and automatically fine-tune the seal pressure, fine-tune the displacement, and lower the actuator action threshold; At the same time, intelligently mark and classify the seal data during the abnormal occurrence period and automatically synchronize it to the abnormal event archive; According to the characteristics of the abnormal event, automatically generate a structured abnormal report and push it to the operation and maintenance personnel.

[0012] Furthermore, the specific process of taking adaptive seal compensation adjustment measures according to the seal state is as follows: Calculate the seal anomaly judgment value in real time, add the seal anomaly judgment value of the previous moment and the seal anomaly judgment value of the current moment and take the average to obtain the average seal state value; Subtract the current moment pressure from the target pressure value to obtain the pressure deviation value; Multiply the average seal state value by the pressure deviation value to obtain the seal compensation adjustment value; When the seal compensation adjustment value is greater than the adjustment threshold, it is determined that there is a risk of seal leakage, and a compensation adjustment instruction to increase the seal pressure and tighten the seal structure is automatically issued; When the seal compensation adjustment value is less than or equal to the adjustment threshold, it is determined that there is a risk of seal damage, and an instruction to reduce pressure and loosen the adjustment is automatically issued; After the seal adjustment instruction is executed, new seal data is automatically collected, the seal anomaly judgment value is calculated in real time, and the adjustment effect is checked. If the seal state returns to normal, it is automatically recorded as a successful adjustment; If the abnormal seal state is not lifted, the compensation adjustment is automatically cycled and the compensation amount is gradually increased until the seal state returns to normal; If the seal state is still abnormal after three consecutive adjustments, automatically perform multi-dimensional feature statistics and comparison on the historical data of pressure, temperature, vibration, and displacement during the abnormal period, and use correlation analysis based on the pressure deviation ratio, temperature deviation ratio, and vibration adjustment factor to automatically identify the parameter with the largest change amplitude and the most significant deviation from the normal trend during the abnormal period, and automatically attribute it to the types mainly caused by pressure fluctuation, temperature impact, vibration disturbance, and multi-factor coupling. At the same time, the alarm level is automatically upgraded, an abnormal attribution report is generated, the main influencing factors leading to the invalidity of this adjustment are clearly marked and pushed to the operation and maintenance personnel; Build an abnormal event archive, record all seal data, seal anomaly judgment values, and seal compensation adjustment values during this period, and synchronously record the attribution report into the abnormal event archive; After each adjustment, write the seal anomaly judgment value, seal compensation adjustment value, and adjustment effect into the seal database; Automatically count and analyze successful adjustment and failure cases, optimize pressure and anomaly judgment parameters, and generate a seal health trend report, abnormal frequency statistics, and maintenance window suggestions.

[0013] Furthermore, the specific process of receiving the seal compensation adjustment instruction in real time, collecting the adjusted seal data in real time, and judging the adjustment deviation of the adaptive seal compensation adjustment based on the adjusted seal data is as follows: receiving the adjustment command issued by the data analysis and decision-making module in real time, the SMA actuator adjusts the current and temperature according to the adjustment command to achieve fine seal adjustment; the hydraulic unit accurately adjusts the pressure according to the adjustment command to compensate for the actual pressure change of the seal interface; obtaining the displacement data of the adjustment mechanism in the seal data, based on the sliding time window length, obtaining the displacement amount after the adjustment of the historical adjustment component and calculating the average value to obtain the expected displacement amount of the adjustment component; subtracting the current pressure from the target pressure value, taking the absolute value, and then dividing by the target pressure value to obtain the absolute pressure deviation ratio; subtracting the current displacement data from the expected displacement amount of the adjustment component, taking the absolute value, and then dividing by the expected displacement amount to obtain the absolute displacement deviation ratio; multiplying the absolute pressure deviation ratio by the pressure weight factor to obtain the pressure deviation value, multiplying the absolute displacement deviation ratio by the displacement weight factor to obtain the displacement deviation value, and adding the pressure deviation value and the displacement deviation value to obtain the comprehensive deviation value; adding the ratio of the average vibration intensity to the vibration standard value to a constant one to obtain the vibration adjustment factor; multiplying the comprehensive deviation value by the vibration adjustment factor to obtain the seal adjustment deviation correction value.

[0014] Furthermore, the specific process of implementing the deviation correction measure based on the adjustment deviation judgment result is as follows: comparing the seal adjustment deviation correction value with the deviation threshold in real time. When the seal adjustment deviation correction value is less than the deviation threshold, it is considered that the seal compensation adjustment is qualified, the seal is normal, and the further adjustment operation is automatically stopped, and the normal monitoring mode is entered, and the normal seal data collection and periodic seal state self-check are restored; and the record of this adjustment is archived; when the seal adjustment deviation correction value is greater than or equal to the deviation threshold, it is considered that the seal compensation adjustment is unqualified and the seal is abnormal, and the adjustment amplitude is automatically increased and the adjustment interval is shortened; the seal data is collected again in real time, and the sampling frequency of the seal data is increased, and the seal adjustment deviation correction value is calculated cyclically until the seal adjustment deviation correction value is less than the deviation threshold; if the seal adjustment deviation correction value is greater than or equal to the deviation threshold for three consecutive times, through the sliding window analysis method, comparing the dynamic change curves of the seal data before and after the adjustment, automatically searching for the similarity with the historical seal abnormal events in the abnormal event archive, if there is a match, the main cause of the abnormality is automatically marked; starting the actuator health detection, and automatically comparing the volatility and continuity of the seal data collected by the sensor, if drift, loss and mutation are found, a maintenance alarm is automatically issued, and after the cause is clearly attributed, it is automatically pushed to the operation and maintenance end to remind specific attention to the main cause parameters and related components; pushing the alarm information and the main cause of the abnormality to the operation and maintenance personnel through the information platform, and synchronizing all abnormal seal data, adjustment attempts, execution status and response feedback to the abnormal event archive.

[0015] Furthermore, the specific process of visual intelligent response based on the seal data and seal status before and after seal compensation adjustment is as follows: according to the seal data and seal status before and after adjustment, the dynamic trend curve of the seal data is displayed in real time; automatically detect and highlight abnormal points, and push sound and pop-up warnings according to the abnormal classification; at the same time, support the query and report generation of historical seal data and abnormal events, intelligently analyze the seal health trend and give maintenance suggestions; support operation and maintenance personnel to initiate temporary adjustment and maintenance requests through the interface, and all operations and abnormal processes are automatically recorded and archived.

[0016] The second aspect of the present invention provides a sealing system for a two-phase immersion liquid-cooled data center, including: a dynamic perception module, a data analysis and decision-making module, an adaptive execution module, and a status visualization module: among them, the dynamic perception module is used to collect seal data in real time through a multi-type integrated sensor, perform data preprocessing on the real-time seal data, establish a seal database, and transmit and standardize the storage of the seal data in real time according to time and position information; the data analysis and decision-making module is used to receive the preprocessed real-time seal data and perform structured sorting, judge the seal status in real time according to the received seal data, and take adaptive seal compensation adjustment measures according to the seal status; the adaptive execution module is used to receive the seal compensation adjustment instruction in real time, collect the adjusted seal data in real time, judge the adjustment deviation of the adaptive seal compensation adjustment according to the adjusted seal data, and implement deviation correction measures based on the adjustment deviation judgment result; the status visualization module is used to perform visual intelligent response based on the seal data and seal status before and after seal compensation adjustment.

[0017] The third aspect of the present invention provides a sealing device for a two-phase immersion liquid-cooled data center, including: an integrated sensor array, a micro intelligent actuator unit, an adaptive seal ring assembly, and an embedded control and communication terminal: among them, the integrated sensor array is used to integrate multiple sensors, collect seal data in real time, and provide all-round perception data of the seal health status; the micro intelligent actuator unit is used to receive the preprocessed real-time seal data and perform structured sorting, drive the seal ring to tighten and relax, and use intelligent driving components to perform adaptive seal compensation adjustment measures on the seal ring; the adaptive seal ring assembly is used to receive the seal compensation adjustment instruction in real time, collect the adjusted seal data in real time, and realize real-time tightening and loosening adjustment of the seal surface according to the adjusted seal data to ensure a long-term healthy seal state; the embedded control and communication terminal is used to integrate multiple modules to perform local criterion calculation, adjustment instruction generation, automatic control execution, and two-way communication with the cloud based on the seal data and seal status before and after seal compensation adjustment, and perform intelligent closed-loop control and remote operation and maintenance interaction.

[0018] Beneficial effects The present invention has the following beneficial effects: (1) In the present invention, by real-time sensing of multiple environmental fluctuations such as pressure, temperature, and vibration, and automatically identifying the dynamic seal failure trend from loose to tight and then to loose through multi-parameter criteria, and autonomously performing compensation adjustment and deviation correction, it fundamentally overcomes the problem that traditional static seals are difficult to adapt to dynamic working conditions, and effectively eliminates the leakage risk caused by seal failure due to fluctuations in the liquid cooling environment.

[0019] (2) In the present invention, by collecting, archiving, and analyzing the health trends of seal status and adjustment effects throughout the process, supporting automatic marking of anomalies, generation of health reports, and suggestions for maintenance windows, predictive maintenance and intelligent operation and maintenance of the seal system are realized, greatly improving the operation and maintenance efficiency and seal reliability.

[0020] (3) In the present invention, by constructing a self-learning optimization mechanism for weight factors and anomaly criterion parameters based on historical seal data, automatically analyzing the seal status, adjustment effects, and seal data during long-term operation, dynamically adjusting the weights of each parameter and the criterion threshold, high self-adaptability is achieved, continuously improving the long-term steady-state performance and self-evolution ability of the seal system in diverse environments, and breaking through the technical bottlenecks of rigid parameters and lagging response in traditional seal systems.

[0021] (4) In the present invention, by integrating a status visualization module, supporting real-time trend display, abnormal high-light warning, and one-key query of historical data and events, operation and maintenance personnel can remotely issue adjustment and maintenance requests at any time, and all anomalies and operation processes are automatically archived, greatly improving the visualization management ability and full-life cycle traceability of the seal system in the data center.

[0022] Of course, it is not necessary for any product implementing the present invention to achieve all the above-mentioned advantages simultaneously. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a flowchart of a sealing method for a two-phase immersion liquid cooling data center; Figure 2 It is a structural diagram of a sealing system for a two-phase immersion liquid cooling data center; Figure 3 It is a dynamic trend diagram of seal abnormal deviation correction and pressure displacement. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0024] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. As understood by those skilled in the art, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0025] Please refer to Figures 1 - 3 , an embodiment of the present invention provides a technical solution: a sealing method, system and device for a two-phase immersion liquid-cooled data center, including: Step 1, collect sealing data in real time through multi-type integrated sensors, perform data preprocessing on the real-time sealing data, establish a sealing database, and perform real-time transmission and standardized storage of the sealing data according to time and location information; Step 2, receive the preprocessed real-time sealing data and perform structured arrangement, judge the sealing state in real time according to the received sealing data, and take adaptive sealing compensation adjustment measures according to the sealing state; Step 3, receive the sealing compensation adjustment instruction in real time, and collect the adjusted sealing data in real time, judge the adjustment deviation of the adaptive sealing compensation adjustment according to the adjusted sealing data, and implement deviation correction measures based on the adjustment deviation judgment result; Step 4, perform visual intelligent response based on the sealing data and sealing state before and after the sealing compensation adjustment.

[0026] Specifically, the specific process of collecting sealing data in real time through multi-type integrated sensors, performing data preprocessing on the real-time sealing data, establishing a sealing database, and performing real-time transmission and standardized storage of the sealing data according to time and location information is as follows: Use multi-type sensors, including pressure sensors, temperature sensors, acceleration sensors and displacement sensors, to collect the sealing data of the device in real time. The sealing data includes: real-time temperature, pressure, vibration data and the displacement data of the adjustment mechanism; Check whether the sealing data is missing. For the missing sealing data, use linear interpolation method to complete the sealing data. For the completed sealing data, perform denoising and filtering operations by combining median filtering and mean filtering, and perform multi-level filtering optimization for high-frequency noise and slow-varying interference respectively to improve the stability and representativeness of the signal; Identify and eliminate dynamic outliers in the sealing data through the sliding window detection algorithm. By setting the parameters of multiple window lengths, the abnormal detection sensitivity under various working conditions can be flexibly adapted, and short-term extreme outliers and abnormal interferences caused by sensor failures can be automatically eliminated; Further perform preliminary normalization processing on the sealing data. According to the measurement ranges and historical distributions of various sensors, perform standardized transformation on the data of different parameters to eliminate the numerical influence between different measurement dimensions and provide a unified scale for subsequent algorithm criteria and multi-parameter analysis; Package the preprocessed sealing data according to the time stamp and sensor position, store it in the sealing database, and send it to the data analysis and decision-making module in real time through the bus.

[0027] In this implementation scheme, multi-type integrated sensors are used to collect multi-dimensional data of temperature, pressure, vibration, and displacement of the sealing system in real time. By combining linear interpolation for completion, filtering for noise reduction, sliding window anomaly rejection, and normalization for multiple data preprocessing methods, the integrity, accuracy, and robustness of the sealing data are effectively improved. The data is uniformly packaged according to the timestamp and sensor location and stored in the sealing database in real time, which not only realizes efficient data standardization management but also provides a solid data foundation for subsequent intelligent analysis, criterion calculation, and closed-loop control. It significantly enhances the perception ability and dynamic response level of the sealing system to abnormal states under variable working conditions, providing strong support for sealing health monitoring and intelligent decision-making.

[0028] Specifically, the specific process of receiving the preprocessed real-time sealing data and performing structured arrangement and judging the sealing state according to the received sealing data in real time is as follows: Receive the real-time sealing data from the dynamic perception module, perform preliminary decoding and verification on the sealing data, and perform structured arrangement according to the source, acquisition time, and data type of the sealing data. Establish unique identifiers for the sealing data of different channels and different measuring points to ensure the accuracy of data traceability and the efficiency of retrieval; Obtain the pressure data, and calculate the average value of the pressure data under normal leakage-free working conditions within one month to obtain the target pressure value; Obtain the temperature data, and calculate the average value of the temperature data under normal leakage-free working conditions within one month to obtain the target temperature value; Obtain the vibration data, sum the squares of all vibration data within the time window length and then calculate the average value to obtain the root mean square value of the vibration acceleration of the sealing measuring point as the average vibration intensity; At the same time, based on the sliding time window, calculate the distribution of historical vibration data, and take the value at the 95th percentile as the vibration standard value, which can effectively distinguish short-term accidental vibration from systematic abnormal vibration and improve the sensitivity of early abnormal identification; Subtract the current pressure from the target pressure value and divide by the target pressure value to obtain the pressure deviation ratio; Subtract the current temperature from the target temperature value and divide by the target temperature value to obtain the temperature deviation ratio; Square and sum the pressure deviation ratio and the temperature deviation ratio respectively, and then take the square root to obtain the multi-factor deviation distance; Add the ratio of the average vibration intensity to the vibration standard value to the constant one to obtain the vibration adjustment factor. Adding to the constant one is to ensure that even if the vibration risk is zero, the vibration adjustment factor itself always has a value of 1, and the overall sealing anomaly judgment value will not disappear due to the vibration term being zero, thus ensuring the safety, stability, and engineering applicability of the criterion; Multiply the multi-factor deviation distance by the vibration adjustment factor to obtain the sealing anomaly judgment value, which comprehensively reflects the overall deviation degree of the sealing state from the target working condition, and through the dynamic adjustment of the vibration factor, more sensitively reflects the risk amplification effect brought by minor leakage, relaxation, or mechanical shock; Store the sealing anomaly judgment value in the sealing database, and compare the sealing anomaly judgment value with the anomaly threshold in real time, and take sealing health grading response measures according to the comparison result of the anomaly threshold.

[0029] The specific formula for the sealing abnormality judgment value is: ; Where, Indicates the seal abnormality judgment value, which is used to judge the health of the seal status at the current moment. It comprehensively reflects the synergistic and amplifying effect of the three major influencing factors of seal pressure deviation, temperature deviation and vibration disturbance on the seal performance. The larger the abnormality judgment value, the greater the deviation of the seal status from the ideal working condition, the stronger the disturbance, and the higher the risk of seal failure. Indicates the pressure at the current moment, reflecting the actual force on the seal; Indicates the target pressure value, reflecting the ideal pressure that needs to be maintained for sealing, ensuring the reliability baseline of the seal; Indicates the current temperature, dynamically reflects the current temperature of the sealing area, and monitors environmental and process fluctuations; Indicates the target temperature value, representing the ideal temperature expected in the sealing design, ensuring the sealing performance under thermal expansion and contraction; Indicates at time and before and after The average vibration intensity within an event is a standard indicator that measures the overall vibration disturbance intensity of the sealing part over a period of time. The greater the average vibration intensity, the more severe the mechanical disturbance to the sealing part, the greater the impact on the sealing reliability, and the higher the risk of seal failure. Indicates the vibration standard value, which serves as the vibration reference level for healthy sealing conditions; Represents the vibration adjustment factor, which is used to automatically amplify the overall abnormal impact according to the vibration risk. The greater the vibration, the higher the risk of sealing abnormality, and automatic early warning and adjustment are carried out.

[0030] In this implementation, by structuring the pre-processed real-time sealing data and dynamically setting target pressure, temperature, and vibration reference values in combination with historical healthy operating condition data, accurate quantification and dynamic identification of the sealing status are achieved. The composite criterion of multi-factor deviation distance and vibration adjustment factor can comprehensively reflect the synergistic effects of multiple influencing factors such as pressure, temperature, and vibration on the sealing status, improving the accuracy and sensitivity of abnormality determination. Sealing abnormality determination values are generated and archived in real time, supporting health graded response and intelligent closed-loop regulation, effectively enhancing the sealing system's ability to intelligently identify abnormal conditions under complex operating conditions, implement graded control, and proactively prevent and control risks.

[0031] Specifically, the specific process of taking sealed health grading response measures according to the abnormal threshold comparison result is as follows: when the sealed abnormal judgment value is less than or equal to the abnormal threshold, it is determined that the sealed state is normal, and only routine monitoring and sealed data recording are carried out, and a health report is generated regularly to support predictive maintenance, ensuring that the system can operate efficiently under healthy working conditions, avoiding ineffective regulation, and providing a data basis for the whole-life cycle health management and trend analysis of the equipment through the regular report output; when the sealed abnormal judgment value is greater than the abnormal threshold, it is determined that the sealed state is abnormal, and it enters the adaptive compensation adjustment step, automatically encrypts the acquisition and monitoring period of the sealed data, refines the time window, and captures the dynamic change process of the sealed performance more precisely, providing timely and accurate data support for abnormal response; automatically identifies continuous small-amplitude and short-term sealed abnormal changes, captures early sealed abnormal signals, and automatically micro-adjusts the sealed pressure, micro-adjusts the displacement, and reduces the actuator action threshold, actively discovers and intervenes in latent faults and initial deterioration phenomena, and effectively delays the sealed failure process; at the same time, intelligently marks and classifies and archives the sealed data during the abnormal occurrence period, and automatically synchronizes it to the abnormal event archive; according to the characteristics of the abnormal event, automatically generates a structured abnormal report and pushes it to the operation and maintenance personnel.

[0032] In this implementation plan, through the hierarchical comparison of the sealed abnormal judgment value and the threshold, an intelligent hierarchical response to the sealed health state is realized. In the normal state, the system automatically enters routine monitoring and data archiving, and regularly generates a health report to support predictive maintenance; in the abnormal state, the system can not only adaptively adjust the acquisition and monitoring frequency and dynamically refine the abnormal response window, but also has the ability of early abnormal automatic identification and fine-tuning compensation, can actively intervene at the initial stage of potential hazards, and effectively block the risk expansion. Greatly improves the early warning, hierarchical control and operation and maintenance automation level of the sealed system, and significantly enhances the reliability and independent safety prevention and control ability of the system.

[0033] Specifically, the specific process of taking adaptive sealing compensation adjustment measures according to the sealing status is: calculate the sealing abnormality judgment value in real time, add the sealing abnormality judgment value of the previous moment and the sealing abnormality judgment value of the current moment and calculate the average value to obtain the average sealing status value, effectively smooth out occasional fluctuations, and improve the stability and anti-interference ability of the adjustment criterion; subtract the current moment pressure from the target pressure value to obtain the pressure deviation value, the pressure deviation value reflects the direct difference between the current working condition and the ideal state, and is the core basis for evaluating the sealing performance deviation and potential risks; multiply the average sealing status value by the pressure deviation value to obtain the sealing compensation adjustment value, dynamically combine the overall deviation with the instantaneous pressure abnormality, and realize the adaptive coupling of the adjustment amplitude and the degree of abnormality; when the sealing compensation adjustment value is greater than the adjustment threshold, it is determined that there is a risk of sealing leakage, and automatically issues a compensation adjustment instruction to increase the sealing pressure and tighten the sealing structure; when the sealing compensation adjustment value is less than or equal to the adjustment threshold, it is determined that there is a risk of seal damage, and automatically issues a decompression and loosening adjustment instruction; real-time prevention and control of two common risks of sealing leakage and structural damage to avoid human misjudgment and response delay. After the sealing adjustment instruction is executed, new sealing data is automatically collected, the sealing abnormality judgment value is calculated in real time, and the adjustment effect is checked. If the sealing state returns to normal, it is automatically recorded as a successful adjustment; if the sealing abnormality is not relieved, compensation adjustment is automatically performed in a cycle and the compensation amount is gradually increased until the sealing state returns to normal. This type of closed-loop adaptive adjustment process ensures the system's dynamic response capability to continuous abnormalities until the sealing performance is fully restored; if the sealing state is still abnormal after three consecutive adjustments, the pressure, temperature, vibration, and displacement historical data of the abnormal period are automatically statistically analyzed and compared in multiple dimensions, and based on the pressure deviation ratio, temperature, and displacement, the system automatically calculates the characteristics of the sealing data. The deviation ratio and vibration adjustment factor use correlation analysis to automatically identify the parameters with the largest change amplitude and the most significant deviation from the normal trend during the abnormal period, and automatically attribute them to pressure fluctuation, temperature shock, vibration disturbance and multi-factor coupling types. At the same time, the alarm level is automatically upgraded and an abnormal attribution report is generated. The attribution report generation mechanism adopts multi-parameter feature deviation scoring and time series correlation analysis algorithm to automatically compare the fluctuation amplitude, mean shift and change trend of each parameter in the abnormal section, and combines the abnormal pattern clustering of historical cases to intelligently classify and sort the main causes of the abnormality, so as to achieve accurate attribution of abnormalities in complex working conditions. The main influencing factors that caused the invalid adjustment are clearly marked and pushed to the operation and maintenance personnel, providing them with targeted intervention suggestions, reducing troubleshooting time and improving processing efficiency; building an abnormal event archive to record all sealing data, sealing abnormality judgment values and sealing compensation adjustment values for the period, and synchronously recording the attribution report into the abnormal event archive; after each adjustment, the sealing abnormality judgment value, sealing compensation adjustment value and adjustment effect are written into the sealing database; automatically count and analyze successful adjustment and failure cases, optimize pressure and abnormality judgment parameters, and generate sealing health trend reports, abnormality frequency statistics and maintenance window recommendations.

[0034] Among them, the specific formula for the seal compensation adjustment value is: ; In the formula, represents the seal compensation adjustment value, which is used to control the pressure adjustment of the sealing device; represents the seal abnormality judgment value at the current moment, which quantitatively reflects the degree of seal abnormality; represents the abnormality judgment value at the previous moment, which reflects the short-term trend and inertia; represents the pressure at the current moment, which reflects the actual force on the seal at present; represents the target pressure value, which reflects the ideal pressure that the seal needs to maintain.

[0035] In this implementation scheme, by calculating and dynamically updating the seal abnormality judgment value in real time, and intelligently issuing pressurization or depressurization adjustment commands, the adaptive compensation and closed-loop control of the seal state are realized. When multiple adjustments are ineffective, the system automatically attributes the main cause of the abnormality and upgrades the alarm, generates a detailed report and synchronizes the operation and maintenance. All data and adjustment effects are automatically archived, supporting parameter self-optimization and health trend analysis. This method significantly improves the intelligent adjustment, abnormality diagnosis and self-optimization capabilities of the seal system, and enhances the long-term security guarantee of the data center.

[0036] Specifically, the specific process of receiving the seal compensation adjustment command in real time, collecting the seal data after adjustment in real time, and judging the adjustment deviation of the adaptive seal compensation adjustment according to the adjusted seal data is as follows: receiving the adjustment command issued by the data analysis and decision-making module in real time, the SMA actuator adjusts the current and temperature according to the adjustment command to realize fine adjustment of the seal; the hydraulic unit accurately adjusts the pressure according to the adjustment command to compensate for the actual pressure change of the seal interface; obtaining the displacement data of the adjustment mechanism in the seal data, based on the sliding time window length, obtaining the displacement amount after the adjustment of the historical adjustment component and calculating the average value to obtain the expected displacement amount of the adjustment component. This windowing and mean extraction method can adapt to different working conditions fluctuations and effectively avoid adjustment errors caused by short-term abnormalities or actuator misoperations; subtracting the pressure at the current moment from the target pressure value, taking the absolute value, and then dividing by the target pressure value to obtain the absolute pressure deviation ratio; subtracting the displacement data at the current moment from the expected displacement amount of the adjustment component, taking the absolute value, and then dividing by the expected displacement amount to obtain the absolute displacement deviation ratio; multiplying the absolute pressure deviation ratio by the pressure weight factor to obtain the pressure deviation value, multiplying the absolute displacement deviation ratio by the displacement weight factor to obtain the displacement deviation value, and adding the pressure deviation value and the displacement deviation value to obtain the comprehensive deviation value; adding the ratio of the average vibration intensity to the vibration standard value to a constant one to obtain the vibration adjustment factor, and the vibration adjustment factor is , which can sensitively reflect the amplifying effect of dynamic disturbance on the reliability of sealing adjustment, and effectively prevent the accumulation of risks caused by external impact, fatigue and equipment failure; the sealing adjustment deviation correction value is obtained by multiplying the comprehensive deviation value with the vibration adjustment factor.

[0037] Among them, the specific formula for the seal adjustment deviation correction value is: ; Where, The seal adjustment deviation correction value indicates the comprehensive adjustment deviation of the adaptive actuator at the current moment, which is used to determine whether the adjustment meets the standard and whether further compensation correction is required. Indicates the pressure at the current moment, reflecting the actual force on the seal; Indicates the target pressure value, reflecting the ideal pressure that needs to be maintained for sealing, ensuring the reliability baseline of the seal; Indicates the displacement of the regulating component after adjustment, reflecting the actual displacement state of the actuator after adjustment; Indicates the expected displacement of the adjustment component, reflecting the ideal displacement target to be achieved in this round of adjustment; Indicates at time and before and after The average vibration intensity within an event is a standard indicator that measures the overall vibration disturbance intensity of the sealing part over a period of time. The greater the average vibration intensity, the more severe the mechanical disturbance to the sealing part, the greater the impact on the sealing reliability, and the higher the risk of seal failure. Indicates the vibration standard value, which serves as the vibration reference level for healthy sealing conditions; represents the pressure weight factor. Based on the historical pressure recovery speed after adjustment and the success rate of pressure anomaly relief, a single-variable sensitivity analysis method is used to fix the displacement weights. Under different pressure weight factor values, the average time and number of times the pressure recovers to the target pressure value after each adjustment are calculated. The pressure weight factor that achieves the fastest and best pressure anomaly recovery is selected as the optimal pressure weight factor, and its value range is between 0.5 and 1.0; The displacement weight factor is the displacement weight factor. Based on the actual displacement arrival rate after adjustment and the response efficiency of the sealing mechanism in the historical data, a linear regression analysis method is adopted. The historical sealing data and the sealing compensation adjustment value are used to establish a regression relationship between the displacement deviation and the adjustment effect. By minimizing the displacement adjustment error, the displacement weight factor that can best improve the sealing mechanism's action accuracy and efficiency is deduced. As the optimal displacement weight factor, the value range is between 0.3 and 1.0.

[0038] Set the pressure weight factor to 0.6, the displacement weight factor to 0.4, the target pressure value to 1, the expected displacement to 5, and the vibration standard value to 0.2. Take five sampling time points. As time changes, with the same weight factors, calculate the seal adjustment deviation correction value based on the pressure data at different times, the displacement amount after adjustment of the adjustment component, and the vibration data. As shown in Table 1, the data table of the seal adjustment deviation correction value.

[0039] Table 1 Data table of the seal adjustment deviation correction value

[0040] As Figure 3 shown, it is the seal abnormal deviation correction and pressure-displacement dynamic trend chart provided by the embodiment of the present application. The abscissa in the figure is the sampling time point, the left ordinate is the seal adjustment deviation correction value, the right ordinate is the pressure and displacement data values, and the deviation threshold is set to 0.15. According to Table 1 and Figure 3 it can be seen that the seal adjustment deviation correction values before and after adjustment, as well as the trends of the pressure and displacement data changing with time. And at the moment of seal abnormality, both the pressure and displacement reach relative extreme values, and at the same time, the vibration data is also relatively high. After the subsequent pressure and displacement recover, the seal adjustment deviation correction value also drops immediately.

[0041] In this implementation scheme, by collecting the adjusted seal data in real time, combining the target pressure and the expected displacement, comprehensively calculating the deviation of the pressure and displacement, and superimposing the vibration influence, the seal adjustment deviation is dynamically determined. The system realizes the precise adaptive correction of the seal adjustment based on the multi-parameter weighted and vibration adaptive adjustment factor, significantly improves the response speed and compensation accuracy of the adjustment, and effectively enhances the health stability and risk prevention and control ability of the seal system under complex dynamic working conditions.

[0042] Specifically, the specific process of implementing deviation correction measures based on the adjustment deviation judgment result is as follows: The sealing adjustment deviation correction value is compared with the deviation threshold in real time. When the sealing adjustment deviation correction value is less than the deviation threshold, it is considered that the sealing compensation adjustment is qualified, the sealing is normal, and the further adjustment operation is automatically stopped, and the system switches to the normal monitoring mode, resumes normal sealing data collection and periodic sealing status self-check, avoids ineffective and excessive adjustment, and ensures the efficient operation of the system; and archives the current adjustment record; when the sealing adjustment deviation correction value is greater than or equal to the deviation threshold, it is considered that the sealing compensation adjustment is unqualified, the sealing is abnormal, and the adjustment amplitude is automatically increased and the adjustment interval is shortened; the sealing data is collected again in real time, and the sampling frequency of the sealing data is increased, and the sealing adjustment deviation correction value is calculated cyclically until the sealing adjustment deviation correction value is less than the deviation threshold, ensuring that the system can perform intelligent intervention with higher frequency and greater response amplitude when an abnormality occurs, improving the timeliness of adjustment and the abnormal self-healing ability; if the sealing adjustment deviation correction value is greater than or equal to the deviation threshold for three consecutive times, through the sliding window analysis method, the dynamic change curves of the sealing data before and after adjustment are compared, and the similarity with the historical sealing abnormal events in the abnormal event library is automatically searched. If there is a match, the main cause of the abnormality is automatically marked; the actuator health detection is started, and the volatility and continuity of the sealing data collected by the sensor are automatically compared. If drift, loss, and mutation are found, a maintenance alarm is automatically issued, and after the cause is determined, it is automatically pushed to the operation and maintenance terminal to remind specific attention to the main cause parameters and related components; the judgment criteria for actuator health detection include: whether the deviation between the response amplitude of the actuator feedback displacement and pressure and the theoretical value exceeds the set threshold after the adjustment instruction is issued, whether the response time is abnormally delayed, whether the actuator has three or more consecutive ineffective actions or the feedback signal is continuously abnormal, whether the actuator status signal is lost, the fluctuation is too large, and mutation occurs. As long as any condition is met, it can be automatically determined that the actuator may fail and the status is abnormal, and the corresponding maintenance and alarm process is started. The alarm information and the main cause of the abnormality are pushed to the operation and maintenance personnel through the information platform, and all abnormal sealing data, adjustment attempts, execution status, and response feedback are synchronized to the abnormal event library.

[0043] In this implementation plan, through the real-time comparison of the sealing adjustment deviation correction value and the threshold, the precise closed-loop control and hierarchical response of the sealing adjustment are realized. When the deviation is qualified, the system automatically switches to normal monitoring; when the deviation is abnormal, the system automatically increases the adjustment amplitude and the sampling frequency, and corrects cyclically until it returns to normal. For the situation of continuous adjustment failure, the system can automatically compare historical abnormal events, intelligently mark the main cause of the abnormality, and start the self-check of the actuator and sensor. When an abnormality is found, a maintenance alarm and a main cause prompt are automatically pushed. All abnormal data and response processes are archived in real time, providing a basis for subsequent precise cause tracing and operation and maintenance optimization. This method greatly improves the intelligent diagnosis, rapid self-healing, and automated closed-loop operation and maintenance capabilities of the sealing system.

[0044] Specifically, the specific process of visual intelligent response based on the seal data and seal status before and after seal compensation adjustment is as follows: According to the seal data and seal status before and after adjustment, the dynamic trend curve of the seal data is displayed in real time, intuitively reflecting the whole process of the change of pressure, displacement, and vibration parameters of the seal system over time, helping operation and maintenance personnel to insight into the working condition fluctuations and adjustment effects, and timely identify operation anomalies and potential problems; automatically detect and highlight the abnormal points, and push sound and pop-up warnings according to the abnormal classification; at the same time, support the query and report generation of historical seal data and abnormal events, intelligently analyze the seal health trend and give maintenance suggestions; support operation and maintenance personnel to initiate temporary adjustment and maintenance requests through the interface, and all operations and abnormal processes are automatically recorded and archived, realizing the leap from passive monitoring of seal health to active management, intelligent warning, data-driven operation and maintenance, and full life cycle traceability, significantly enhancing the seal safety and operation and maintenance efficiency of the data center.

[0045] In this implementation plan, through the real-time visual display of the data and seal status before and after seal compensation adjustment, the intuitive monitoring of the operation status of the seal system and the rapid identification of anomalies are realized. The system can not only dynamically present multi-dimensional trend curves, highlight abnormal points and classify intelligent warnings, but also support the efficient retrieval of historical data and abnormal events, automatic report generation and maintenance suggestion push. Operation and maintenance personnel can conveniently initiate adjustment or maintenance operations, and all operations and abnormal processes are automatically archived, greatly improving the digital and standardized levels of visual management, intelligent warning and full-process operation and maintenance of the seal system.

[0046] Refer to Figure 2 As shown, the second aspect of the present invention provides a seal system for a two-phase immersion liquid-cooled data center, which is applied to the above-mentioned seal method for a two-phase immersion liquid-cooled data center, including: a dynamic perception module, a data analysis and decision-making module, an adaptive execution module, and a status visualization module: Among them, the dynamic perception module is used to collect seal data in real time through multi-type integrated sensors, perform data preprocessing on the real-time seal data, establish a seal database, and transmit and standardize the storage of the seal data in real time according to time and location information; the data analysis and decision-making module is used to receive the preprocessed real-time seal data and perform structured sorting, judge the seal status in real time according to the received seal data, and take adaptive seal compensation adjustment measures according to the seal status; the adaptive execution module is used to receive the seal compensation adjustment instruction in real time, collect the adjusted seal data in real time, judge the adjustment deviation of the adaptive seal compensation adjustment according to the adjusted seal data, and implement deviation correction measures based on the adjustment deviation judgment result; the status visualization module is used to perform visual intelligent response based on the seal data and seal status before and after seal compensation adjustment.

[0047] In this implementation scheme, through the integration of modules such as dynamic perception, intelligent analysis, adaptive execution, and visualization linkage, the efficient acquisition, fine preprocessing, and standardized management of sealed data are realized. The system can dynamically determine the seal state based on multi-dimensional real-time data and intelligently generate compensation adjustment strategies to achieve closed-loop adaptive adjustment and abnormal self-correction. Through the whole-process data archiving and trend visualization, it supports the intuitive monitoring, hierarchical warning, and operation and maintenance decision-making of the seal state. Overall, it improves the intelligent diagnosis, autonomous adjustment, and digital management capabilities of the seal system in complex environments, and significantly enhances the seal safety and operation reliability of the data center.

[0048] The third aspect of the present invention provides a sealing device for a two-phase immersion liquid-cooled data center, which is applied to the above-mentioned sealing method for a two-phase immersion liquid-cooled data center, and includes: an integrated sensor array, a micro intelligent actuator unit, an adaptive sealing ring assembly, and an embedded control and communication terminal. Among them, the integrated sensor array is used to integrate multiple sensors to collect seal data in real time and provide all-round perception data of the seal health state; the micro intelligent actuator unit is used to receive the preprocessed real-time seal data, perform structured sorting, drive the sealing ring to tighten and relax, and use intelligent driving components to perform adaptive seal compensation adjustment measures on the sealing ring; the adaptive sealing ring assembly is used to receive seal compensation adjustment instructions in real time, collect the adjusted seal data in real time, and realize real-time tightening and loosening adjustment of the seal surface according to the adjusted seal data to ensure a long-term healthy seal state; the embedded control and communication terminal is used to integrate multiple modules to perform local criterion calculation, adjustment instruction generation, automatic control execution, and two-way communication with the cloud based on the seal data and seal state before and after seal compensation adjustment, and perform intelligent closed-loop control and remote operation and maintenance interaction.

[0049] In this implementation scheme, the real-time perception of multi-dimensional parameters of the seal system is realized through the integrated sensor array, providing data support for the comprehensive monitoring of the seal health state. The micro intelligent actuator unit can accurately drive the sealing ring to implement automatic adjustment, cooperate with the adaptive sealing ring assembly, and realize the dynamic tightening and loosening adjustment of the seal surface, significantly improving the long-term reliability of the seal. The embedded control and communication terminal is responsible for local intelligent criterion calculation, adjustment control, and remote two-way interaction with the cloud, ensuring the closed-loop intelligent control and efficient operation and maintenance response of the seal adjustment. Overall, this solution greatly improves the self-perception, self-adjustment, and remote intelligent management capabilities of the seal system, providing a solid guarantee for the safe and stable operation of the data center.

[0050] It should be noted that, in this document, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprising", "including" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device.

[0051] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific embodiments described. As understood by those skilled in the art, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the present invention, so that those skilled in the art can well understand and utilize the present invention. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. A sealing method for a two-phase immersion liquid-cooled data center, characterized in that, Including: Step 1: Real-time collect seal data through multi-type integrated sensors, perform data preprocessing on the real-time seal data, establish a seal database, and perform real-time transmission and standardized storage of the seal data according to time and location information; Step 2: Receive the preprocessed real-time seal data and perform structured arrangement, judge the seal state in real time according to the received seal data, and take adaptive seal compensation adjustment measures according to the seal state; Step 3: Receive the seal compensation adjustment instruction in real time, and collect the adjusted seal data in real time. Judge the adjustment deviation of the adaptive seal compensation adjustment according to the adjusted seal data, and implement deviation correction measures based on the adjustment deviation judgment result; Step 4: Perform visual intelligent response based on the seal data and seal state before and after seal compensation adjustment.

2. A sealing method for a two-phase immersion liquid-cooled data center according to claim 1, characterized in that, The specific process of real-time collecting seal data through multi-type integrated sensors, performing data preprocessing on the real-time seal data, establishing a seal database, and performing real-time transmission and standardized storage of the seal data according to time and location information is as follows: Use multi-type sensors to collect the seal data of the device in real time. The seal data includes: real-time temperature, pressure, vibration data, and displacement data of the adjustment mechanism; Check whether the seal data is missing. For the missing seal data, use the linear interpolation method to complete the seal data. For the completed seal data, perform denoising and filtering operations by combining median filtering and mean filtering. Identify and remove dynamic outliers from the seal data through the sliding window detection algorithm, and further perform preliminary normalization processing on the seal data; Package the preprocessed seal data according to the time stamp and sensor location, store it in the seal database, and send it to the data analysis and decision-making module in real time through the bus.

3. A sealing method for a two-phase immersion liquid-cooled data center according to claim 1, characterized in that, The specific process of receiving the preprocessed real-time seal data and performing structured arrangement, and judging the seal state in real time according to the received seal data is as follows: Receive the real-time seal data from the dynamic perception module, perform preliminary decoding and verification on the seal data, and perform structured arrangement according to the source, collection time, and data type of the seal data; Obtain the pressure data, and calculate the average value of the pressure data under normal leak-free conditions within one month to obtain the target pressure value; obtain the temperature data, and calculate the average value of the temperature data under normal leak-free conditions within one month to obtain the target temperature value; obtain the vibration data, sum the squares of all vibration data within the time window length and then calculate the average value to obtain the root mean square value of the vibration acceleration of the seal measurement point as the average vibration intensity; at the same time, based on the sliding time window, calculate the distribution of historical vibration data, and take the value at the 95th percentile as the vibration standard value; Subtract the current pressure from the target pressure value and divide by the target pressure value to obtain the pressure deviation ratio; Subtract the current temperature from the target temperature value and divide by the target temperature value to obtain the temperature deviation ratio; Square the pressure deviation ratio and the temperature deviation ratio respectively, sum them up, and then take the square root to obtain the multi-factor deviation distance; add the ratio of the average vibration intensity to the vibration standard value to a constant to get the vibration adjustment factor; multiply the multi-factor deviation distance by the vibration adjustment factor to obtain the seal abnormality judgment value. Store the seal abnormality judgment value in the seal database, and compare the seal abnormality judgment value with the abnormality threshold in real time. Take seal health grading response measures according to the comparison result of the abnormality threshold.

4. A sealing method for a two-phase immersion liquid-cooled data center according to claim 3, characterized in that, The specific process of taking seal health grading response measures according to the comparison result of the abnormality threshold is as follows: When the seal abnormality judgment value is less than or equal to the abnormality threshold, it is determined that the seal state is normal. Only conduct routine monitoring and seal data recording, generate health reports regularly, and support predictive maintenance. When the seal abnormality judgment value is greater than the abnormality threshold, it is determined that the seal state is abnormal. Enter the adaptive compensation adjustment step. Automatically encrypt the acquisition and monitoring cycle of seal data and refine the time window. Automatically identify continuous small-amplitude and short-term seal abnormality changes, capture early seal abnormality signals, and automatically fine-tune the seal pressure, fine-tune the displacement, and lower the actuator action threshold. At the same time, intelligently mark and classify the seal data during the abnormal occurrence period and automatically synchronize it to the abnormal event archive. Automatically generate a structured abnormal report and push it to the operation and maintenance personnel according to the characteristics of the abnormal event.

5. A sealing method for a two-phase immersion liquid-cooled data center according to claim 1, characterized in that, The specific process of taking adaptive seal compensation adjustment measures according to the seal state is as follows: Calculate the seal abnormality judgment value in real time. Add the seal abnormality judgment value at the previous moment and the seal abnormality judgment value at the current moment and take the average to obtain the average seal state value. Subtract the pressure at the current moment from the target pressure value to obtain the pressure deviation value. Multiply the average seal state value by the pressure deviation value to obtain the seal compensation adjustment value. When the seal compensation adjustment value is greater than the adjustment threshold, it is determined that there is a risk of seal leakage, and automatically issue a compensation adjustment instruction to increase the seal pressure and tighten the seal structure. When the seal compensation adjustment value is less than or equal to the adjustment threshold, it is determined that there is a risk of seal component damage, and automatically issue an instruction to reduce pressure and loosen the adjustment. After the seal adjustment instruction is executed, automatically collect new seal data, calculate the seal abnormality judgment value in real time, and check the adjustment effect. If the seal state returns to normal, automatically record it as a successful adjustment. If the abnormal seal state is not lifted, automatically perform compensation adjustment in a loop and gradually increase the compensation amount until the seal state returns to normal. If the sealing state is still abnormal after three consecutive adjustments, automatically conduct multi-dimensional feature statistics and comparison on the historical data of pressure, temperature, vibration, and displacement during the abnormal period. Using correlation analysis based on the pressure deviation ratio, temperature deviation ratio, and vibration adjustment factor, automatically identify the parameter with the largest change amplitude and the most significant deviation from the normal trend during the abnormal period, and automatically attribute it to the types mainly caused by pressure fluctuation, temperature shock, vibration disturbance, and multi-factor coupling. At the same time, automatically upgrade the alarm level, generate an abnormal attribution report, clearly mark the main influencing factors that cause the invalidity of this adjustment, and push them to the operation and maintenance personnel; construct an abnormal event archive, record all the sealing data, sealing abnormal judgment values, and sealing compensation adjustment values during this period, and synchronously record the attribution report into the abnormal event archive; After each adjustment, write the sealing abnormal judgment value, sealing compensation adjustment value, and adjustment effect into the sealing database; automatically count and analyze successful and failed cases, optimize the pressure and abnormal judgment parameters, and generate a sealing health trend report, abnormal frequency statistics, and maintenance window suggestions.

6. A sealing method for a two-phase immersion liquid-cooled data center according to claim 1, characterized in that, The specific process of receiving the sealing compensation adjustment instruction in real time, collecting the sealing data after adjustment in real time, and judging the adjustment deviation of the adaptive sealing compensation adjustment according to the adjusted sealing data is as follows: Receive the adjustment command sent by the data analysis and decision-making module in real time. The SMA actuator adjusts the current and temperature according to the adjustment command to achieve fine adjustment of the seal; the hydraulic unit accurately adjusts the pressure according to the adjustment command to compensate for the actual pressure change of the sealing interface; Obtain the displacement data of the adjustment mechanism in the sealing data. Based on the sliding time window length, obtain the displacement amount of the historical adjustment component after adjustment and calculate the average value to obtain the expected displacement amount of the adjustment component; Subtract the current pressure from the target pressure value, take the absolute value, and then divide by the target pressure value to obtain the absolute pressure deviation ratio; subtract the current displacement data from the expected displacement amount of the adjustment component, take the absolute value, and then divide by the expected displacement amount to obtain the absolute displacement deviation ratio; Multiply the absolute pressure deviation ratio by the pressure weight factor to obtain the pressure deviation value, multiply the absolute displacement deviation ratio by the displacement weight factor to obtain the displacement deviation value, and add the pressure deviation value and the displacement deviation value to obtain the comprehensive deviation value; add the ratio of the average vibration intensity to the vibration standard value to a constant one to obtain the vibration adjustment factor; Multiply the comprehensive deviation value by the vibration adjustment factor to obtain the sealing adjustment deviation correction value.

7. A sealing method for a two-phase immersion liquid-cooled data center according to claim 1, characterized in that, The specific process of implementing the deviation correction measure based on the adjustment deviation judgment result is as follows: Compare the sealing adjustment deviation correction value with the deviation threshold in real time. When the sealing adjustment deviation correction value is less than the deviation threshold, it is considered that the sealing compensation adjustment is qualified, the seal is normal, automatically stop further adjustment operations, switch to the normal monitoring mode, resume normal sealing data collection and periodic sealing state self-check; and archive the record of this adjustment; When the seal adjustment deviation correction value is greater than or equal to the deviation threshold, it is considered that the seal compensation adjustment fails and the seal is abnormal. The adjustment amplitude is automatically increased and the adjustment interval is shortened; the seal data is collected in real time again, and the sampling frequency of the seal data is increased. The seal adjustment deviation correction value is calculated cyclically until the seal adjustment deviation correction value is less than the deviation threshold; If the seal adjustment deviation correction value is greater than or equal to the deviation threshold for three consecutive times, through the sliding window analysis method, the dynamic change curves of the seal data before and after the adjustment are compared, and the similarity with the historical seal abnormal events in the abnormal event library is automatically searched. If there is a match, the main cause of the abnormality is automatically marked; The actuator health detection is started, and the volatility and continuity of the seal data collected by the sensor are automatically compared. If drift, loss, and mutation are found, a maintenance alarm is automatically issued, and after the attribution is clarified, it is automatically pushed to the operation and maintenance terminal to remind specific attention to the main cause parameters and related components; the alarm information and the main cause of the abnormality are pushed to the operation and maintenance personnel through the information platform, and all abnormal seal data, adjustment attempts, execution status, and response feedback are synchronized to the abnormal event library.

8. A sealing method for a two-phase immersion liquid-cooled data center according to claim 1, characterized in that, The specific process of the visual intelligent response based on the seal data and the seal state before and after the seal compensation adjustment is as follows: According to the seal data and the seal state before and after the adjustment, the dynamic trend curve of the seal data is displayed in real time; the abnormal points are automatically detected and highlighted, and sound and pop-up warnings are pushed according to the abnormal classification; at the same time, it supports the query and report generation of historical seal data and abnormal events, intelligently analyzes the seal health trend and gives maintenance suggestions; It supports the operation and maintenance personnel to initiate temporary adjustment and maintenance requests through the interface, and all operations and abnormal processes are automatically recorded and archived.

9. A sealing system for a two-phase immersion liquid-cooled data center, applying the sealing method for a two-phase immersion liquid-cooled data center according to any one of claims 1-8, comprising: The dynamic perception module, the data analysis and decision-making module, the adaptive execution module and the status visualization module, characterized in that: Among them, the dynamic perception module is used to collect seal data in real time through a multi-type integrated sensor, perform data preprocessing on the real-time seal data, establish a seal database, and transmit and standardize the storage of the seal data according to time and location information; The data analysis and decision-making module is used to receive the preprocessed real-time seal data and perform structured sorting, judge the seal state in real time according to the received seal data, and take adaptive seal compensation adjustment measures according to the seal state; The adaptive execution module is used to receive the seal compensation adjustment instruction in real time, collect the seal data after the adjustment in real time, judge the adjustment deviation of the adaptive seal compensation adjustment according to the seal data after the adjustment, and implement the deviation correction measure based on the adjustment deviation judgment result; The status visualization module is used to perform visual intelligent response based on the seal data and the seal state before and after the seal compensation adjustment.

10. A sealing device for a two-phase immersion liquid-cooled data center, applying the sealing method for a two-phase immersion liquid-cooled data center according to any one of claims 1-8, comprising: The integrated sensor array, the micro intelligent actuator unit, the adaptive seal ring assembly and the embedded control and communication terminal, characterized in that: Among them, the integrated sensor array is used to integrate multiple sensors, collect seal data in real time, and provide all-round perception data of the seal health state; The described micro intelligent actuator unit is used to receive preprocessed real-time sealing data, structure it, drive the sealing ring to tighten and loosen, and use intelligent driving components to perform adaptive sealing compensation adjustment measures on the sealing ring; The described adaptive sealing ring assembly is used to receive sealing compensation adjustment instructions in real time, collect the adjusted sealing data in real time, and realize real-time tightening and loosening adjustment of the sealing surface according to the adjusted sealing data to ensure a long-term healthy sealing state; The described embedded control and communication terminal is used to integrate multiple modules to perform local criterion calculation, adjustment instruction generation, automatic control execution, and two-way communication with the cloud based on the sealing data and sealing state before and after sealing compensation adjustment, and perform intelligent closed-loop control and remote operation and maintenance interaction.

Images