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

Abstract

The present invention discloses a sealing method, system and device for a two-phase immersion liquid-cooled data center, which relates to the field of data center sealing processing technology. The method comprises: step one, real-time collection of sealing data, data preprocessing, real-time transmission and standardized storage of the sealing data; step two, real-time judgment of the sealing state based on the received sealing data, and taking adaptive sealing compensation adjustment measures; step three, real-time collection of adjusted sealing data, judgment of the adjustment deviation of the adaptive sealing compensation adjustment, and implementation of deviation correction measures based on the adjustment deviation judgment result; step four, visual intelligent response based on the sealing data and sealing state before and after the sealing compensation adjustment. The method solves the problem that during the operation of the data center, the seals undergo dynamic changes from loose to tight and then loose again due to changes in the working conditions of pressure, temperature and vibration fluctuations, making it difficult to achieve long-term good sealing.

Description

Technical Field

[0001] The present invention relates to the technical field of sealing processing for data centers, and in particular to a sealing method, system and device for a two-phase immersion liquid-cooled data center. Background Art

[0002] With the development of big data, cloud computing, and high-performance computing, data center power consumption continues to rise. Traditional air cooling methods are no longer able to meet the cooling efficiency and energy consumption control requirements of high-density servers. Two-phase immersion liquid cooling technology, with its superior heat transfer performance, noise reduction, and energy savings, is becoming a key cooling solution for the next generation of green data centers. In two-phase immersion liquid cooling data center applications, 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 chemically corrosive. If the seal fails and the coolant leaks, it will not only cause equipment short circuits and data loss, but also pose serious risks to human health and environmental safety.

[0003] For example, the invention patent with publication number CN103838337B discloses a data center server cooling cabinet, including a server cabinet, a phase change heat transfer system, a server casing and a spring sheet for fixing the server casing; the phase change heat transfer system includes multiple heat exchangers installed in the server cabinet, an outdoor cooling device and connecting pipes; the multiple heat exchangers are respectively installed on each layer of brackets where servers are placed in the server cabinet; the server casing is placed on the upper part of the multiple heat exchangers; the server is sealed inside the server casing; the entire data center server cooling cabinet has no air cooling, the server heating unit is in direct contact with the server casing, and the server casing is in direct contact with multiple heat exchangers installed in the server cabinet, so as to achieve contact heat transfer and improve heat transfer efficiency; the data center server cooling cabinet of the present invention can increase the placement density of servers, has low construction cost and short construction period, and greatly reduces the operating cost of the entire data center.

[0004] For example, the invention patent with publication number CN113849054A discloses an immersion liquid cooling heat dissipation system for use in a data center, comprising a cooling box with built-in cooling liquid, a rotating shaft provided in the cooling box for sealing rotation, a cooling air chamber for cooling liquid vapor to enter fixed on 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 provided on the cooling air chamber, the cooling disturbance mechanism comprises a heat sink hingedly provided on the side wall of the cooling air chamber and capable of being flipped up and down, an elastic shrinkage member is provided between the heat sink and the cooling air chamber, and a cooling flow channel is provided in the heat sink that is connected to the cooling air chamber through the elastic shrinkage member, the up and down flipping of the heat sink is controlled by a second driving mechanism, and fan blades are provided on the side of the heat sink close to the cooling box for disturbing the air on the side wall of the cooling box as the cooling air chamber rotates. The 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 solutions of the invention in the embodiments of the present application, the present application found that the above technology has at least the following technical problems:

[0006] In the actual operating environment of a two-phase immersion liquid-cooled data center, sealing devices often need to withstand long-term influences from multiple complex working conditions, including repeated fluctuations in cooling fluid pressure, temperature cycle changes caused by heat generation from server equipment, cabinet and pipeline vibration, and external mechanical disturbances. These factors will cause the seals to continuously experience dynamic force changes from loose to tight and then loose again during operation, making it impossible to ensure long-term good sealing.

[0007] Therefore, in response to the above problems, there is an urgent need for a sealing method, system and device for a two-phase immersion liquid-cooled data center. Summary of the Invention

[0008] Technical problems solved

[0009] In response to the shortcomings of the existing technology, the present invention provides a sealing method, system and device for a two-phase immersion liquid-cooled data center, which solves the problem that during the operation of the data center, the seals undergo dynamic changes from loose to tight and then loose again due to changes in working conditions such as pressure, temperature and vibration fluctuations, making it difficult to achieve long-term good sealing.

[0010] Technical solution. To achieve the above objectives, the present invention is implemented through the following technical solution: a sealing method for a two-phase immersion liquid-cooled data center, comprising: step one, real-time collection of sealing data through multi-type integrated sensors, data preprocessing of the real-time sealing data, establishment of a sealing database, and real-time transmission and standardized storage of the sealing data according to time and location information; step two, receiving the preprocessed real-time sealing data and performing structured organization, judging the sealing status in real time based on the received sealing data, and taking adaptive sealing compensation adjustment measures based on the sealing status; step three, real-time reception of sealing compensation adjustment instructions, and real-time collection of adjusted sealing data, performing adjustment deviation judgment of adaptive sealing compensation adjustment based on the adjusted sealing data, and implementing deviation correction measures based on the adjustment deviation judgment result; step four, performing visual intelligent response based on the sealing data and sealing status before and after the sealing compensation adjustment.

[0011] Furthermore, the sealing data is collected in real time by multi-type integrated sensors, the real-time sealing data is preprocessed, a sealing database is established, and the sealing data is transmitted and stored in a standardized manner in real time according to time and position information. The specific process is as follows: the sealing data of the device is collected in real time by multi-type sensors, and the sealing data includes: real-time temperature, pressure, vibration data and displacement data of the adjustment mechanism; the sealing data is checked to see if it is missing. For missing sealing data, the sealing data is supplemented by linear interpolation. The supplemented sealing data is denoised and filtered by combining median filtering and mean filtering. The dynamic outliers of the sealing data are identified and eliminated by the sliding window detection algorithm, and the sealing data is further preliminarily normalized; the preprocessed sealing data is uniformly packaged according to timestamps and sensor positions, stored in the sealing database, and sent to the data analysis and decision-making module in real time through the bus.

[0012] Furthermore, the pre-processed real-time sealing data is received and structured, and the specific process of judging the sealing status in real time based on the received sealing data is as follows: receiving the real-time sealing data from the dynamic sensing module, preliminarily decoding and verifying the sealing data, and structuring the sealing data according to the source, acquisition time and data type of the sealing data; obtaining pressure data, and calculating the average value of the pressure data under normal no-leakage working conditions within one month to obtain the target pressure value; obtaining temperature data, and calculating the average value of the temperature data under normal no-leakage working conditions within one month to obtain the target temperature value; obtaining vibration data, square-summing all vibration data within the time window length, and then calculating 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, the distribution of historical vibration data is calculated, and the 95th percentile value is taken as the vibration standard value; the target pressure value is subtracted from the current pressure and then divided by the target pressure value to obtain the pressure deviation ratio; the target temperature value is subtracted from the current temperature and then divided by the target temperature value to obtain the temperature deviation ratio; the pressure deviation ratio and the temperature deviation ratio are squared and summed, and then the square root is taken to obtain the multi-factor deviation distance; the ratio of the average vibration intensity to the vibration standard value is added to the constant 1 to obtain the vibration adjustment factor; the multi-factor deviation distance is multiplied by the vibration adjustment factor to obtain the sealing abnormality judgment value; the sealing abnormality judgment value is stored in the sealing database, and the sealing abnormality judgment value and the abnormality threshold are compared in real time, and sealing health classification response measures are taken according to the abnormality threshold comparison results.

[0013] Furthermore, the specific process of taking sealing health graded response measures based on the abnormality threshold comparison results is as follows: when the sealing abnormality judgment value is less than or equal to the abnormality threshold, the sealing state is judged to be normal, and only routine monitoring and sealing data recording are performed, and health reports are generated regularly to support predictive maintenance; when the sealing abnormality judgment value is greater than the abnormality threshold, it is judged that the sealing state is abnormal, and the adaptive compensation adjustment step is entered, the sealing data collection and monitoring cycle is automatically encrypted, and the time window is refined; continuous small-amplitude, short-term sealing abnormal changes are automatically identified, early sealing abnormality signals are captured, and the sealing pressure and displacement are automatically fine-tuned, and the actuator action threshold is lowered; at the same time, the sealing data of the abnormal period are intelligently marked and graded and archived, and automatically synchronized to the abnormal event archive; according to the characteristics of the abnormal event, a structured abnormality report is automatically generated and pushed to the operation and maintenance personnel.

[0014] Furthermore, the specific process of taking adaptive sealing compensation adjustment measures according to the sealing status is as follows: 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; subtract the current moment pressure from the target pressure value to obtain the pressure deviation value; multiply the average sealing status value and the pressure deviation value to obtain the sealing compensation adjustment value; 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; 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 status returns to normal, it is automatically recorded as a successful adjustment; if the sealing abnormality status is not relieved, the compensation adjustment is automatically circulated and the compensation amount is gradually increased until the sealing status returns to normal; if three consecutive If the sealing status is still abnormal after the first adjustment, the multi-dimensional feature statistics and comparison of the historical data of pressure, temperature, vibration and displacement during the abnormal period will be automatically performed, and the correlation analysis based on the pressure deviation ratio, temperature deviation ratio and vibration adjustment factor will be used 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 will be automatically upgraded, and an abnormal attribution report will be generated, which will clearly mark the main influencing factors that caused the invalid adjustment and push them to the operation and maintenance personnel; an abnormal event archive will be constructed to record all sealing data, sealing abnormality judgment values ​​and sealing compensation adjustment values ​​for this period, and the attribution report will be synchronously recorded in the abnormal event archive; after each adjustment, the sealing abnormality judgment value, sealing compensation adjustment value and adjustment effect will be written into the sealing database; successful adjustment and failure cases will be automatically counted and analyzed, pressure and abnormality judgment parameters will be optimized, and sealing health trend reports, abnormal frequency statistics and maintenance window recommendations will be generated.

[0015] Furthermore, the specific process of receiving the seal compensation adjustment instruction in real time, collecting the adjusted seal data in real time, and making the adjustment deviation judgment 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 module in real time, and the SMA actuator adjusts the current and temperature according to the adjustment command to achieve seal fine-tuning; the hydraulic unit accurately adjusts the pressure according to the adjustment command to compensate for the actual pressure change of the sealing interface; obtaining the displacement data of the adjustment mechanism in the sealing data, obtaining the displacement of the historical adjustment component after adjustment based on the length of the sliding time window, and calculating the average value to obtain the expected displacement of the adjustment component; using the target The absolute pressure deviation ratio is obtained by subtracting the current pressure from the pressure value, taking the absolute value, and then dividing it by the target pressure value; the absolute displacement deviation ratio is obtained by subtracting the current displacement data from the expected displacement of the adjustment component, taking the absolute value, and then dividing it by the expected displacement; the absolute pressure deviation ratio is multiplied by the pressure weight factor to obtain the pressure deviation value, the absolute displacement deviation ratio is multiplied by the displacement weight factor to obtain the displacement deviation value, and the pressure deviation value is added to the displacement deviation value to obtain the comprehensive deviation value; the ratio of the average vibration intensity to the vibration standard value is added to the constant 1 to obtain the vibration adjustment factor; the comprehensive deviation value is multiplied by the vibration adjustment factor to obtain the sealing adjustment deviation correction value.

[0016] Furthermore, the specific process of implementing deviation correction measures based on the adjustment deviation judgment result is: 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 meets the standard and the sealing is normal, and further adjustment operations are automatically stopped, and the normal monitoring mode is switched to resume normal sealing data collection and periodic sealing status self-inspection; and the adjustment records are archived; when the sealing adjustment deviation correction value is greater than or equal to the deviation threshold, it is considered that the sealing compensation adjustment does not meet the standard and 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 sealing data sampling frequency is increased, and the sealing adjustment deviation correction value is cyclically calculated until the sealing adjustment deviation correction value is less than the deviation Threshold; if the seal adjustment deviation correction values ​​for three consecutive times are greater than or equal to the deviation threshold, the dynamic change curves of the sealing data before and after the adjustment are compared through the sliding window analysis method, and the similarity with the historical sealing abnormal events in the abnormal event archive is automatically found. If there is a match, the main cause of the abnormality is automatically marked; start the actuator health detection, and automatically compare the volatility and continuity of the sealing data collected by the sensor. If drift, loss and mutation are found, a maintenance alarm is automatically issued, and after the attribution is clear, it is automatically pushed to the operation and maintenance end, reminding specific attention to the main cause parameters and related components; push the alarm information and the main cause of the abnormality to the operation and maintenance personnel through the information platform, and synchronize all abnormal sealing data, adjustment attempts, execution status and response feedback to the abnormal event archive.

[0017] Furthermore, the specific process of visual intelligent response based on the sealing data and sealing status before and after the sealing compensation adjustment is: based on the sealing data and sealing status before and after the adjustment, the dynamic trend curve of the sealing data is displayed in real time; abnormal points are automatically detected and highlighted, and sound and pop-up alarms are pushed according to the abnormality level; at the same time, historical sealing data and abnormal event queries and report generation are supported, and sealing health trends are intelligently analyzed to provide maintenance recommendations; operation and maintenance personnel are supported to initiate temporary adjustment and maintenance requests through the interface, and all operations and abnormal processes are automatically recorded and archived.

[0018] The second aspect of the present invention provides a sealing system for a two-phase immersion liquid-cooled data center, comprising: a dynamic perception module, a data analysis and decision module, an adaptive execution module and a state visualization module: wherein the dynamic perception module is used to collect sealing data in real time through multiple types of integrated sensors, perform data preprocessing on the real-time sealing data, establish a sealing database, and transmit and standardize the sealing data in real time according to time and location information; the data analysis and decision module is used to receive the preprocessed real-time sealing data and perform structured organization, judge the sealing status in real time according to the received sealing data, and take adaptive sealing compensation adjustment measures according to the sealing status; the adaptive execution module is used to receive the sealing compensation adjustment instruction in real time, and collect the adjusted sealing data in real time, perform adjustment deviation judgment of the adaptive sealing compensation adjustment according to the adjusted sealing data, and implement deviation correction measures based on the adjustment deviation judgment result; the state visualization module is used to perform visual intelligent response based on the sealing data and sealing status before and after the sealing compensation adjustment.

[0019] The third aspect of the present invention provides a sealing device for a two-phase immersion liquid-cooled data center, comprising: an integrated sensor array, a micro-intelligent actuator unit, an adaptive sealing ring assembly and an embedded control and communication terminal: wherein the integrated sensor array is used to integrate multiple sensors, collect sealing data in real time, and provide all-round perception data of the sealing health status; the micro-intelligent actuator unit is used to receive pre-processed real-time sealing data and perform structured organization, drive the sealing ring to tighten and relax, and use intelligent driving components to perform adaptive sealing compensation adjustment measures on the sealing ring; the adaptive sealing ring assembly is used to receive sealing compensation adjustment instructions in real time, and collect adjusted sealing data in real time, and realize real-time tightness adjustment of the sealing surface according to the adjusted sealing data to ensure a long-term healthy sealing 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 sealing data before and after the sealing compensation adjustment and the sealing state, to perform intelligent closed-loop control and remote operation and maintenance interaction.

[0020] Beneficial effects

[0021] The present invention has the following beneficial effects:

[0022] (1) The present invention senses multiple environmental fluctuations such as pressure, temperature, and vibration in real time, automatically identifies the dynamic sealing failure trend from loose to tight and then to loose again through multi-parameter judgment, and independently performs compensation adjustment and deviation correction. This fundamentally overcomes the problem that traditional static seals are difficult to adapt to dynamic working condition changes, and effectively eliminates the risk of leakage caused by seal failure due to fluctuations in a liquid cooling environment.

[0023] (2) The present invention collects, archives and analyzes the health trend of the sealing status and adjustment effect throughout the entire process, supports automatic marking of abnormalities, generation of health reports and maintenance window suggestions, realizes predictive maintenance and intelligent operation and maintenance of the sealing system, and greatly improves operation and maintenance efficiency and sealing reliability.

[0024] (3) The present invention constructs a self-learning optimization mechanism for weight factors and abnormal judgment parameters based on historical sealing data, automatically analyzes the sealing status, adjustment effect and sealing data during long-term operation, and dynamically adjusts the weights of various parameters and judgment thresholds, thereby achieving high adaptability, continuously improving the long-term steady-state performance and self-evolution ability of the sealing system in various environments, and breaking through the technical bottleneck of parameter rigidity and response lag of traditional sealing systems.

[0025] (4) The present invention supports real-time trend display, abnormality highlight alarm, and one-click query of historical data and events through an integrated status visualization module. Operation and maintenance personnel can remotely issue adjustment and maintenance requests at any time. All abnormalities and operation processes are automatically archived, which greatly improves the visual management capability and full life cycle traceability of the data center sealing system.

[0026] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 A flow chart of a sealing method for a two-phase immersion liquid-cooled data center;

[0028] Figure 2 A structural diagram of a sealing system for a two-phase immersion liquid-cooled data center;

[0029] Figure 3 This is a dynamic trend chart of sealing abnormal deviation correction and pressure displacement. DETAILED DESCRIPTION

[0030] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. As those skilled in the art will understand, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0031] See also Figure 1-Figure 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 one, 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 transmitting and standardizing the sealing data in real time according to time and location information; step two, receiving the preprocessed real-time sealing data and performing structured organization, judging the sealing status in real time according to the received sealing data, and taking adaptive sealing compensation adjustment measures according to the sealing status; step three, receiving the sealing compensation adjustment instruction in real time, and collecting the adjusted sealing data in real time, judging the adjustment deviation of the adaptive sealing compensation adjustment according to the adjusted sealing data, and implementing deviation correction measures based on the adjustment deviation judgment result; step four, performing visual intelligent response based on the sealing data and the sealing status before and after the sealing compensation adjustment.

[0032] Specifically, the sealing data is collected in real time through multi-type integrated sensors, the real-time sealing data is pre-processed, a sealing database is established, and the sealing data is transmitted and stored in a standardized manner according to time and location information. The specific process is as follows: using 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 displacement data of the adjustment mechanism; checking whether the sealing data is missing, and for the missing sealing data, using linear interpolation to complete the sealing data, and combining median filtering and mean filtering to perform denoising and filtering operations on the completed sealing data, and performing multi-stage filtering for high-frequency noise and slow-changing interference respectively. Optimize and improve the stability and representativeness of the signal; use the sliding window detection algorithm to identify and eliminate dynamic outliers in the sealing data, and use the parameter setting of multiple window lengths to flexibly adapt the abnormal detection sensitivity under various working conditions, and automatically eliminate abnormal interference caused by short-term extreme abnormal points and sensor failures; further perform preliminary normalization on the sealing data, and standardize the data of different parameters according to the range and working condition history distribution of various sensors to eliminate the numerical influence between different measurement dimensions, and provide a unified scale for subsequent algorithm judgment and multi-parameter analysis; the preprocessed sealing data is uniformly packaged according to timestamps and sensor positions, stored in the sealing database, and sent to the data analysis and decision-making module in real time through the bus.

[0033] In this implementation, multi-dimensional data on temperature, pressure, vibration, and displacement of the sealing system are collected in real time through multi-type integrated sensors. Combined with linear interpolation completion, filtering and denoising, sliding window anomaly removal, and normalization, multiple data preprocessing methods are used to effectively improve the integrity, accuracy, and robustness of the sealing data. The data is uniformly packaged and stored in the sealing database in real time according to timestamps and sensor positions, which not only achieves efficient data standardization management but also provides a solid data foundation for subsequent intelligent analysis, criterion calculation, and closed-loop control. This significantly enhances the sealing system's ability to perceive abnormal conditions and its dynamic response level under variable working conditions, providing strong support for seal health monitoring and intelligent decision-making.

[0034] Specifically, the preprocessed real-time sealing data is received and structured, and the specific process of judging the sealing status in real time based on the received sealing data is as follows: receiving real-time sealing data from the dynamic sensing module, preliminarily decoding and verifying the sealing data, and structurally organizing the sealing data according to the source, acquisition time and data type of the sealing data, establishing a unique identifier for the sealing data of different channels and different measuring points to ensure the accuracy of data traceability and the efficiency of retrieval; obtaining pressure data, and calculating the average value of the pressure data under normal no-leakage conditions within one month to obtain the target pressure value; obtaining temperature data, and calculating the average value of the temperature data under normal no-leakage conditions within one month to obtain the target temperature value; obtaining vibration data, square-summing all vibration data within the time window length, and then calculating 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, calculating the distribution of historical vibration data, taking the 95th percentile value as the vibration standard value, which can effectively distinguish short-term occasional vibrations from systematic abnormal vibrations, and improve early abnormality identification. sensitivity; subtract the current pressure from the target pressure value and divide it by the target pressure value to obtain the pressure deviation ratio; subtract the current temperature from the target temperature value and divide it by the target temperature value to obtain the temperature deviation ratio; square the pressure deviation ratio and the temperature deviation ratio respectively, sum them, 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. The purpose of adding to the constant one is to ensure that even if the vibration risk is zero, the vibration adjustment factor itself is always 1, and the sealing abnormality judgment value will not disappear as a whole due to the vibration term being zero, thereby ensuring the safety, stability and engineering applicability of the judgment criterion; multiply the multi-factor deviation distance by the vibration adjustment factor to obtain the sealing abnormality judgment value, which comprehensively reflects the global deviation degree between the sealing state and the target working condition, and through the dynamic adjustment of the vibration factor, more sensitively reflects the risk amplification effect caused by small leakage, relaxation or mechanical shock; the sealing abnormality judgment value is stored in the sealing database, and the sealing abnormality judgment value is compared with the abnormality threshold in real time, and sealing health classification response measures are taken according to the abnormality threshold comparison results.

[0035] The specific formula for the sealing abnormality judgment value is:

[0036] ;

[0037] 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.

[0038] 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.

[0039] Specifically, the specific process of taking seal health graded response measures based on the abnormality threshold comparison results is as follows: when the seal abnormality judgment value is less than or equal to the abnormality threshold, the seal status is determined to be normal, and only routine monitoring and sealing data recording are performed. Health reports are generated regularly to support predictive maintenance, ensuring that the system can operate efficiently under healthy conditions and avoid ineffective adjustments. Through regular report output, a data basis is provided for equipment health management and trend analysis throughout its life cycle. When the seal abnormality judgment value is greater than the abnormality threshold, the seal status is determined to be abnormal, and the adaptive compensation adjustment step is entered. The collection and monitoring cycle of the seal data is automatically encrypted, the time window is refined, and the dynamic changes in the seal performance are more accurately captured, providing timely and accurate data support for abnormal response. The system automatically identifies continuous small-amplitude and short-term seal abnormal changes, captures early seal abnormality signals, and automatically fine-tunes the seal pressure and displacement, lowers the actuator action threshold, proactively discovers and intervenes in hidden faults and initial degradation phenomena, and effectively delays the process of seal failure. At the same time, the sealing data during the abnormal period is intelligently marked and archived in a graded manner, and automatically synchronized to the abnormal event archive. Based on the characteristics of the abnormal event, a structured abnormality report is automatically generated and pushed to the operation and maintenance personnel.

[0040] In this implementation, an intelligent hierarchical response to the health status of the seal is achieved through a hierarchical comparison of seal anomaly judgment values ​​and thresholds. Under normal conditions, the system automatically enters routine monitoring and data archiving, and regularly generates health reports to support predictive maintenance. Under abnormal conditions, the system not only adaptively adjusts the frequency of data acquisition and monitoring and dynamically refines the abnormal response window, but also has the ability to automatically identify early abnormalities and fine-tune compensation, enabling proactive intervention at the initial stage of hidden dangers to effectively block the spread of risks. This significantly improves the early warning, hierarchical management and operation and maintenance automation level of the sealing system, significantly enhancing the system's reliability and autonomous safety control capabilities.

[0041] 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.

[0042] Among them, the specific formula for the sealing compensation adjustment value is:

[0043] ;

[0044] Where, Indicates the seal compensation adjustment value, which is used to control the seal device to adjust the pressure; Indicates the sealing abnormality judgment value at the current moment, quantitatively reflecting the degree of sealing abnormality; Indicates the abnormal judgment value of the previous moment, reflecting the short-term trend and inertia; 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.

[0045] In this implementation, by calculating and dynamically updating seal anomaly detection values ​​in real time, intelligently issuing pressure increase or pressure reduction adjustment commands achieves adaptive compensation and closed-loop control of the seal status. If repeated adjustments fail, the system automatically attributes the primary cause of the anomaly, escalates the alarm, and generates a detailed report for simultaneous operations and maintenance. All data and adjustment results are automatically archived, supporting parameter self-optimization and health trend analysis. This approach significantly enhances the sealing system's intelligent adjustment, anomaly diagnosis, and autonomous optimization capabilities, strengthening the long-term security of the data center.

[0046] Specifically, the sealing compensation adjustment instruction is received in real time, and the adjusted sealing data is collected in real time. The specific process of adjusting deviation judgment of adaptive sealing compensation adjustment is performed according to the adjusted sealing data: the adjustment command issued by the data analysis and decision module is received in real time, and the SMA actuator adjusts the current and temperature according to the adjustment command to achieve seal fine-tuning; the hydraulic unit accurately adjusts the pressure according to the adjustment command to compensate for the actual pressure change of the sealing interface; the displacement data of the adjustment mechanism in the sealing data is obtained, and based on the length of the sliding time window, the displacement of the historical adjustment component after adjustment is obtained and the average value is calculated to obtain the expected displacement of the adjustment component. This windowing and mean extraction method can adaptively Fluctuations in different working conditions can effectively avoid adjustment errors caused by short-term abnormalities or actuator malfunctions; the absolute pressure deviation ratio is obtained by subtracting the current pressure from the target pressure value, taking the absolute value, and then dividing it by the target pressure value; the absolute displacement deviation ratio is obtained by subtracting the current displacement data from the expected displacement of the adjustment component, taking the absolute value, and then dividing it by the expected displacement; the absolute pressure deviation ratio is multiplied by the pressure weight factor to obtain the pressure deviation value, the absolute displacement deviation ratio is multiplied by the displacement weight factor to obtain the displacement deviation value, and the pressure deviation value is added to the displacement deviation value to obtain the comprehensive deviation value; the ratio of the average vibration intensity to the vibration standard value is added to the constant 1 to obtain the vibration adjustment factor, which 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.

[0047] Among them, the specific formula for the seal adjustment deviation correction value is:

[0048] ;

[0049] 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.

[0050] Set the pressure weight factor to 0.6, the displacement weight factor to 0.4, the target pressure value to 1, the expected displacement value to 5, and the vibration standard value to 0.2. Take five sampling time points and, as the weight factors remain the same, calculate the seal adjustment deviation correction value based on the pressure data, the displacement of the adjustment component after adjustment, and the vibration data at different times over time. This is shown in Table 1, the seal adjustment deviation correction value data table.

[0051] Table 1 Seal adjustment deviation correction value data table

[0052]

[0053] like Figure 3 As shown in the figure, the dynamic trend diagram of the sealing abnormal deviation correction and pressure displacement provided by the embodiment of the present application is shown. The horizontal axis in the figure is the sampling time point, the left vertical axis is the sealing adjustment deviation correction value, the right vertical axis is the pressure and displacement data value, and the deviation threshold is set to 0.15. According to Table 1 and Figure 3 It can be seen that the sealing adjustment deviation correction value before and after adjustment and the pressure and displacement data change over time. At the moment of sealing abnormality, the pressure and displacement both reach relative extreme values, and the vibration data is also high. After the subsequent pressure and displacement recover, the sealing adjustment deviation correction value also drops immediately.

[0054] This implementation dynamically determines seal adjustment deviations by collecting real-time seal data after adjustment, combining target pressure with expected displacement, and calculating pressure and displacement deviations. This data is then combined with vibration effects to dynamically determine seal adjustment deviations. Based on multi-parameter weighting and a vibration-adaptive adjustment factor, the system achieves precise adaptive correction of seal adjustment, significantly improving adjustment response speed and compensation accuracy, effectively enhancing the health, stability, and risk prevention capabilities of the sealing system under complex dynamic conditions.

[0055] Specifically, the specific process of implementing deviation correction measures based on the adjustment deviation judgment result is: 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 meets the standard and the sealing is normal, and further adjustment operations are automatically stopped, and the normal monitoring mode is switched to resume normal sealing data collection and periodic sealing status self-inspection to avoid invalid and excessive adjustment and ensure efficient operation of the system; and archive the adjustment records of this time; when the sealing adjustment deviation correction value is greater than or equal to the deviation threshold, it is considered that the sealing compensation adjustment does not meet the standard and the sealing is abnormal, and the adjustment amplitude is automatically increased and the adjustment interval is shortened; re-collect the sealing data in real time, and increase the sealing data sampling frequency, and cyclically calculate the sealing adjustment deviation correction value until the sealing adjustment deviation correction value is less than the deviation threshold, to ensure that when an abnormality occurs, the system can perform intelligent intervention with a higher frequency and a larger response amplitude, thereby improving the timeliness of the adjustment and the abnormal self-healing ability; if the sealing adjustment is repeated three times in a row If the deviation correction value is greater than or equal to the deviation threshold, a sliding window analysis method is used to compare the dynamic change curves of the sealing data before and after adjustment, automatically searching for similarities with historical sealing anomaly events in the abnormal event archive. If a match is found, the main cause of the anomaly is automatically labeled. Actuator health monitoring is initiated and the volatility and continuity of the sealing data collected by the sensor are automatically compared. If drift, loss, or sudden change is detected, a maintenance alarm is automatically issued. After the cause is clearly attributed, it is automatically pushed to the operation and maintenance end, reminding the operator to pay specific attention to the main parameters and related components. Actuator health monitoring criteria include: whether the response amplitude of the actuator feedback displacement and pressure after the adjustment command is issued deviates from the theoretical value by exceeding the set threshold; whether the response time is abnormally delayed; whether the actuator has three or more consecutive invalid actions or the feedback signal is continuously abnormal; loss of the actuator status signal, excessive fluctuation, or sudden change. If any of these conditions are met, the actuator is automatically determined to have failed and is in an abnormal state, and the corresponding maintenance and alarm process is initiated. The alarm information and the main cause of the anomaly 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 archive.

[0056] In this implementation scheme, accurate closed-loop control and graded response of the sealing adjustment are achieved through real-time comparison of the sealing adjustment deviation correction value with the threshold. When the deviation meets the standard, the system automatically switches to normal monitoring; when the deviation is abnormal, the system automatically increases the adjustment amplitude and increases the sampling frequency, and cyclically corrects until it returns to normal. In the case of continuous adjustment failure, the system can automatically compare historical abnormal events, intelligently mark the main cause of the abnormality, and start the actuator and sensor self-inspection. If an abnormality is found, it will automatically push maintenance alarms and main cause prompts. All abnormal data and response processes are archived in real time to provide a basis for subsequent accurate 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.

[0057] Specifically, the specific process of visual intelligent response based on the sealing data and sealing status before and after the sealing compensation adjustment is: according to the sealing data and sealing status before and after the adjustment, the dynamic trend curve of the sealing data is displayed in real time, which intuitively reflects the whole process of the pressure, displacement and vibration parameters of the sealing system changing over time, helping operation and maintenance personnel to understand the fluctuations in working conditions and the adjustment effects, and timely identify operational anomalies and hidden dangers; automatically detect and highlight abnormal points, and push sound and pop-up alarms according to the abnormality level; at the same time, support historical sealing data and abnormal event query and report generation, intelligently analyze the sealing 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 of sealing health from passive monitoring to active management, intelligent alarm, data-driven operation and maintenance and full life cycle traceability, significantly enhancing the sealing safety and operation and maintenance efficiency of the data center.

[0058] This implementation enables intuitive monitoring of the sealing system's operating status and rapid identification of anomalies through real-time visualization of seal compensation data before and after adjustment. The system not only dynamically displays multi-dimensional trend curves, highlights anomalies, and issues graded intelligent alerts, but also supports efficient retrieval of historical data and anomaly events, automatic report generation, and push notifications for maintenance recommendations. Operations and maintenance personnel can easily initiate adjustments or maintenance operations, with all operations and anomaly processes automatically archived. This significantly enhances the visual management, intelligent early warning, and digitalization and standardization of the sealing system's entire operation and maintenance process.

[0059] Reference Figure 2 As shown, the second aspect of the present invention provides a sealing system 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, including: a dynamic perception module, a data analysis and decision module, an adaptive execution module and a state visualization module: wherein the dynamic perception module is used to collect sealing data in real time through multiple types of integrated sensors, perform data preprocessing on the real-time sealing data, establish a sealing database, and transmit and standardize the sealing data in real time according to time and location information; the data analysis and decision module is used to receive the preprocessed real-time sealing data and perform structured organization, judge the sealing status in real time according to the received sealing data, and take adaptive sealing compensation adjustment measures according to the sealing status; the adaptive execution module is used to receive the sealing compensation adjustment instruction in real time, and collect the adjusted sealing data in real time, perform adjustment deviation judgment of the adaptive sealing compensation adjustment according to the adjusted sealing data, and implement deviation correction measures based on the adjustment deviation judgment result; the state visualization module is used to perform visual intelligent response based on the sealing data and sealing status before and after the sealing compensation adjustment.

[0060] This implementation achieves efficient collection, refined preprocessing, and standardized management of sealing data through the integration of modules such as dynamic perception, intelligent analysis, adaptive execution, and visual linkage. Based on multidimensional real-time data, the system dynamically determines the sealing status and intelligently generates compensation adjustment strategies, enabling closed-loop adaptive adjustment and anomaly self-correction. Full-process data archiving and trend visualization support intuitive monitoring of sealing status, graded alarms, and operational decision-making. This overall enhances the sealing system's intelligent diagnosis, autonomous adjustment, and digital management capabilities in complex environments, significantly strengthening the sealing safety and operational reliability of the data center.

[0061] 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, including: an integrated sensor array, a micro-intelligent actuator unit, an adaptive sealing ring assembly and an embedded control and communication terminal: wherein the integrated sensor array is used to integrate multiple sensors, collect sealing data in real time, and provide all-round perception data of the sealing health status; the micro-intelligent actuator unit is used to receive pre-processed real-time sealing data and perform structured organization, drive the sealing ring to tighten and relax, and use intelligent driving components to perform adaptive sealing compensation adjustment measures on the sealing ring; the adaptive sealing ring assembly is used to receive sealing compensation adjustment instructions in real time, and collect adjusted sealing data in real time, and realize real-time tightness adjustment of the sealing surface according to the adjusted sealing data to ensure a long-term healthy sealing state; the embedded control and communication terminal is used to integrate multiple modules to perform local judgment calculation, adjustment instruction generation, automatic control execution and two-way communication with the cloud based on the sealing data before and after the sealing compensation adjustment and the sealing state, to perform intelligent closed-loop control and remote operation and maintenance interaction.

[0062] In this implementation, an integrated sensor array is used to achieve real-time perception of the multi-dimensional parameters of the sealing system, providing data support for comprehensive monitoring of the health status of the seal. The micro-intelligent actuator unit can accurately drive the sealing ring to implement automatic adjustment, and cooperate with the adaptive sealing ring assembly to achieve dynamic tightness adjustment of the sealing surface, significantly improving the long-term reliability of the seal. The embedded control and communication terminal is responsible for local intelligent judgment calculation, adjustment control and remote two-way interaction with the cloud, ensuring closed-loop intelligent control of the sealing adjustment and efficient operation and maintenance response. Overall, this solution has greatly improved the self-perception, self-adjustment and remote intelligent management capabilities of the sealing system, providing a solid guarantee for the safe and stable operation of the data center.

[0063] It should be noted that, in this document, relational terms such as first and second, etc., are used only 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 "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0064] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. The preferred embodiments do not describe all details in detail, nor do they limit the invention to the specific embodiments described. As those skilled in the art will appreciate, numerous modifications and variations are possible based on the contents of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only 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: include: Step 1: Collect sealing data in real time through multi-type integrated sensors, pre-process the real-time sealing data, establish a sealing database, and transmit and standardize the sealing data in real time according to time and location information; Step 2: Receive pre-processed real-time sealing data and perform structured organization, determine the sealing status in real time based on the received sealing data, and take adaptive sealing compensation adjustment measures based on the sealing status; Step 3: Receive the seal compensation adjustment instruction in real time, collect the adjusted seal data in real time, perform the adjustment deviation judgment of the adaptive seal compensation adjustment according to the adjusted seal data, and implement the deviation correction measures based on the adjustment deviation judgment result; Step 4: Perform visual intelligent response based on the sealing data and sealing status before and after the sealing compensation adjustment.

2. The sealing method for a two-phase immersion liquid cooling data center according to claim 1, characterized in that: The specific process of collecting sealing data in real time through multi-type integrated sensors, preprocessing the real-time sealing data, establishing a sealing database, and transmitting and storing the sealing data in real time according to time and location information is as follows: Use multiple types of sensors to collect the sealing data of the device in real time, including: real-time temperature, pressure, vibration data and displacement data of the adjustment mechanism; Check whether the sealing data is missing. For missing sealing data, use linear interpolation to complete the sealing data. Perform denoising and filtering operations on the completed sealing data by combining median filtering and mean filtering. Use the sliding window detection algorithm to identify and eliminate dynamic outliers in the sealing data, and further perform preliminary normalization on the sealing data. The pre-processed sealing data are packaged uniformly according to timestamp and sensor location, stored in the sealing database, and sent to the data analysis and decision module in real time through the bus.

3. The sealing method for a two-phase immersion liquid cooling data center according to claim 1, characterized in that: The specific process of receiving the pre-processed real-time sealing data and performing structured processing, and determining the sealing status in real time based on the received sealing data is as follows: Receive real-time sealing data from the dynamic sensing module, perform preliminary decoding and verification on the sealing data, and organize the sealing data into structures according to its source, collection time, and data type; Obtain pressure data, and calculate the average value of the pressure data under normal, no-leakage conditions within one month to obtain the target pressure value; obtain temperature data, and calculate the average value of the temperature data under normal, no-leakage conditions within one month to obtain the target temperature value; obtain vibration data, square and sum 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 measurement point as the average vibration intensity; at the same time, calculate the distribution of historical vibration data based on the sliding time window, and take the 95th percentile value as the vibration standard value; Subtract the current pressure from the target pressure value and divide it by the target pressure value to get the pressure deviation ratio; Subtract the current temperature from the target temperature and divide it by the target temperature to get the temperature deviation ratio. The pressure deviation ratio and the temperature deviation ratio are squared and summed, and then the square root is taken to obtain the multi-factor deviation distance. The ratio of the average vibration intensity to the vibration standard value is added to the constant 1 to obtain the vibration adjustment factor. The multi-factor deviation distance is multiplied by the vibration adjustment factor to obtain the sealing abnormality judgment value. The sealing abnormality judgment value is stored in the sealing database, and the sealing abnormality judgment value is compared with the abnormality threshold in real time, and sealing health classification response measures are taken according to the abnormality threshold comparison result.

4. The sealing method for a two-phase immersion liquid cooling data center according to claim 3, characterized in that: The specific process of taking seal health graded response measures based on the abnormal threshold comparison results is as follows: When the sealing abnormality judgment value is less than or equal to the abnormality threshold, the sealing status is determined to be normal, and only routine monitoring and sealing data recording are performed. Health reports are generated regularly to support predictive maintenance; When the sealing abnormality judgment value is greater than the abnormality threshold, it is determined that the sealing state is abnormal, and the adaptive compensation adjustment step is entered. The collection and monitoring cycle of the sealing data is automatically encrypted, and the time window is refined; continuous small-amplitude, short-term sealing abnormal changes are automatically identified, and early sealing abnormality signals are captured. The sealing pressure and displacement are automatically fine-tuned, and the actuator action threshold is lowered; at the same time, the sealing data of the abnormal period are intelligently marked and graded and archived, and automatically synchronized to the abnormal event archive; according to the characteristics of the abnormal event, a structured abnormality report is automatically generated and pushed to the operation and maintenance personnel.

5. The sealing method for a two-phase immersion liquid cooling data center according to claim 1, characterized in that: The specific process of taking the adaptive sealing compensation adjustment measures according to the sealing state is as follows: Calculate the sealing abnormality judgment value in real time, add the sealing abnormality judgment value at the previous moment to the sealing abnormality judgment value at the current moment and calculate the average value to obtain the average sealing state value; subtract the current pressure from the target pressure value to obtain the pressure deviation value; multiply the average sealing state value by the pressure deviation value to obtain the sealing 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 sealing pressure and tighten the sealing 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 sealing adjustment command 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 state is not resolved, the compensation adjustment is automatically cycled and the compensation amount is gradually increased until the sealing state returns to normal; If the sealing status is still abnormal after three consecutive adjustments, the system automatically performs multi-dimensional feature statistics and comparison on the historical data of pressure, temperature, vibration and displacement during the abnormal period, and uses correlation analysis based on the pressure deviation ratio, temperature deviation ratio and vibration adjustment factor 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, which clearly marks the main influencing factors that caused the invalid adjustment and pushes them to the operation and maintenance personnel; an abnormal event archive is constructed to record all sealing data, sealing abnormality judgment values ​​and sealing compensation adjustment values ​​during the period, and the attribution report is synchronously recorded in the abnormal event archive; After each adjustment, the seal abnormality judgment value, seal compensation adjustment value and adjustment effect are written into the seal database; successful adjustment and failure cases are automatically counted and analyzed, pressure and abnormality judgment parameters are optimized, and seal health trend reports, abnormality frequency statistics and maintenance window recommendations are generated.

6. The sealing method for a two-phase immersion liquid cooling data center according to claim 1, characterized in that: The specific process of receiving the sealing compensation adjustment instruction in real time, collecting the adjusted sealing data in real time, and performing the adjustment deviation judgment of the adaptive sealing compensation adjustment according to the adjusted sealing data is as follows: The SMA actuator receives real-time adjustment commands from the data analysis and decision-making module, adjusts the current and temperature according to the adjustment commands, and achieves fine-tuning of the seal. The hydraulic unit accurately adjusts the pressure according to the adjustment commands to compensate for the actual pressure changes at the sealing interface. Obtaining displacement data of the adjustment mechanism in the sealing data, obtaining the displacement of the adjustment component after adjustment based on the length of the sliding time window, and calculating the average value to obtain the expected displacement of the adjustment component; The absolute pressure deviation ratio is obtained by subtracting the current pressure from the target pressure value, taking the absolute value, and then dividing it by the target pressure value; the absolute displacement deviation ratio is obtained by subtracting the current displacement data from the expected displacement of the regulating component, taking the absolute value, and then dividing it by the expected displacement; The pressure deviation value is obtained by multiplying the absolute pressure deviation ratio by the pressure weight factor, the displacement deviation value is obtained by multiplying the absolute displacement deviation ratio by the displacement weight factor, and the pressure deviation value and the displacement deviation value are added to obtain the comprehensive deviation value; the vibration adjustment factor is obtained by adding the ratio of the average vibration intensity to the vibration standard value to the constant 1; Multiply the comprehensive deviation value by the vibration adjustment factor to obtain the seal adjustment deviation correction value.

7. The sealing method for a two-phase immersion liquid cooling data center according to claim 1, characterized in that: The specific process of implementing the deviation correction measures based on the adjustment deviation judgment result is as follows: Compare 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 meets the standard and the seal is normal. Further adjustment operations are automatically stopped and the system switches to normal monitoring mode, resuming normal seal data collection and periodic seal status self-inspection. The adjustment record 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 does not meet the standard and the seal is abnormal, and the adjustment range is automatically increased and the adjustment interval is shortened; the seal data is collected again in real time, and the seal data sampling frequency is increased, and the seal adjustment deviation correction value is cyclically calculated until the seal adjustment deviation correction value is less than the deviation threshold; If the seal adjustment deviation correction values ​​are greater than or equal to the deviation threshold for three consecutive times, the dynamic change curves of the seal data before and after the adjustment are compared through the sliding window analysis method, and the similarity with the historical seal abnormal events in the abnormal event archive is automatically found. If there is a match, the main cause of the abnormality is automatically marked; The actuator health check is initiated, and the volatility and continuity of the sealing data collected by the sensor are automatically compared. If drift, loss or mutation is 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 people to pay 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 sealing data, adjustment attempts, execution status and response feedback are synchronized to the abnormal event archive.

8. The sealing method for a two-phase immersion liquid cooling data center according to claim 1, characterized in that: The specific process of performing visual intelligent response based on the sealing data before and after the sealing compensation adjustment and the sealing status is as follows: Based on the sealing data before and after adjustment and the sealing status, the dynamic trend curve of the sealing data is displayed in real time; abnormal points are automatically detected and highlighted, and sound and pop-up alarms are pushed according to the abnormality level; at the same time, it supports the query and report generation of historical sealing data and abnormal events, intelligently analyzes the health trend of the seal and provides maintenance recommendations; Operation and maintenance personnel are supported 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 cooling data center, using the sealing method for a two-phase immersion liquid cooling data center according to any one of claims 1 to 8, comprising: Dynamic perception module, data analysis and decision module, adaptive execution module and state visualization module, characterized by: The dynamic sensing module is used to collect sealing data in real time through multiple types of integrated sensors, pre-process the real-time sealing data, establish a sealing database, and transmit and standardize the sealing data in real time according to time and location information; The data analysis and decision module is used to receive pre-processed real-time sealing data and perform structured processing, determine the sealing status in real time based on the received sealing data, and take adaptive sealing compensation adjustment measures based on the sealing 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, perform the adjustment deviation judgment of the adaptive seal compensation adjustment according to the adjusted seal data, and implement the deviation correction measures based on the adjustment deviation judgment result; The state visualization module is used to perform visual intelligent response based on the sealing data and the sealing state before and after the sealing compensation adjustment.

10. A sealing device for a two-phase immersion liquid cooling data center, using the sealing method for a two-phase immersion liquid cooling data center according to any one of claims 1 to 8, comprising: The integrated sensor array, micro intelligent actuator unit, adaptive sealing ring assembly and embedded control and communication terminal are characterized by: 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 status; The micro intelligent actuator unit is used to receive pre-processed real-time sealing data and perform structured processing, 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 adaptive sealing ring assembly is used to receive sealing compensation adjustment instructions in real time, collect adjusted sealing data in real time, and adjust the tightness of the sealing surface in real time according to the adjusted sealing data to ensure a long-term healthy sealing state; The embedded control and communication terminal is used to integrate multiple modules to perform local judgment calculation, adjustment instruction generation, automatic control execution and two-way communication with the cloud based on the sealing data and sealing status before and after sealing compensation adjustment, so as to perform intelligent closed-loop control and remote operation and maintenance interaction.

Images