Liquid leakage monitoring method and device based on immersion liquid cooling system

Abstract

The application provides a liquid leakage monitoring method and device based on an immersion liquid cooling system, and relates to the technical field of monitoring. The method comprises the following steps: acquiring a signal time delay value of a differential pair, wherein the differential pair is immersed in cooling liquid of the immersion liquid cooling system; determining a predicted time delay value according to the signal time delay value and a preset grey model; and determining a liquid leakage monitoring result for the immersion liquid cooling system based on the predicted time delay value. The preset grey model is used to predictively determine whether the immersion liquid cooling system has a liquid leakage condition, so that liquid leakage monitoring prediction can be achieved, an alarm mechanism can be triggered in advance, and system or board failure caused by liquid leakage can be effectively avoided.

Description

Technical Field

[0001] This application relates to the field of monitoring technology, and in particular to a method and device for monitoring leakage based on an immersion liquid cooling system. Background Technology

[0002] Like any powerful computer hardware, the CPU (Central Processing Unit / Processor) and GPU (Graphics Processing Unit) generate heat during operation and require proper cooling to achieve maximum performance. During normal operation, the transistors inside the CPU / GPU convert electricity into heat, which raises the CPU / GPU temperature. If this heat does not have an effective way to dissipate, the CPU / GPU will exceed its safe operating temperature and malfunction.

[0003] To address the heat dissipation problem, immersion liquid cooling technology was developed. Because heat is dissipated through contact with a non-conductive liquid medium, the overall structure of immersion liquid cooling is directly close to the entire heat source to provide heat dissipation, resulting in better heat dissipation capabilities.

[0004] However, the dielectric fluid (coolant) used in immersion liquid cooling technology poses a risk of leakage, preventing it from effectively dissipating heat from the system. Once a leak occurs, it is irreversible, as the dielectric coolant will spread outside the sealed chassis's piping. Therefore, an effective method for leak detection is urgently needed. Summary of the Invention

[0005] The purpose of this application is to provide a method and apparatus for monitoring leakage in an immersion liquid cooling system, for monitoring leakage in the immersion liquid cooling system.

[0006] This application provides a leakage monitoring method based on an immersion liquid cooling system, including:

[0007] Obtain the signal delay value of the differential pair, which is immersed in the coolant of the immersion liquid cooling system;

[0008] The predicted delay value is determined based on the signal delay value and the preset gray model;

[0009] Based on the predicted time delay value, the leakage monitoring results for the immersion liquid cooling system are determined.

[0010] Optionally, obtaining the signal delay value of the differential pair includes:

[0011] Based on the connector, the differential pairs in the path are identified, and the signal delay values ​​of the differential pairs are obtained.

[0012] Optionally, obtaining the signal delay value of the differential pair includes:

[0013] The signal delay value of the differential pair is periodically acquired, with an acquisition period of less than or equal to 10 seconds.

[0014] Optionally, determining the predicted delay value based on the signal delay value and a preset gray model includes:

[0015] Based on the signal delay values, a first delay value sample set is obtained, which includes at least four signal delay values.

[0016] The first time delay numerical sample set is corrected to obtain the second time delay numerical sample set;

[0017] Based on the second time delay value sample set and the preset gray model, the predicted time delay value is obtained.

[0018] Optionally, obtaining the predicted time delay value based on the second time delay value sample set and the preset gray model includes:

[0019] Sort the sample delay values ​​in the second delay value sample set to obtain the first delay value sequence;

[0020] The first time-delayed numerical sequence is accumulated and transformed to obtain the second time-delayed numerical sequence.

[0021] The second time delay value sequence is input into the preset gray model for prediction to obtain the predicted time delay value.

[0022] Optionally, the preset gray model includes parameters to be determined;

[0023] The step of inputting the second time delay numerical sequence into the preset gray model for prediction to obtain the predicted time delay value includes:

[0024] The preset gray model is modified and adjusted to obtain the target gray model;

[0025] The second time delay numerical sequence is input into the preset gray model, and the preset gray model is deformed to obtain the matrix representation corresponding to the preset gray model;

[0026] The matrix representation is simplified to obtain the matrix equation;

[0027] The error matrix is ​​determined based on the least squares method and the matrix equation.

[0028] The error matrix is ​​solved to obtain the estimated values ​​of the undetermined parameters;

[0029] Substituting the estimated value into the target gray model yields the backup delay value;

[0030] The backup delay value is inversely accumulated to generate a restoration process, resulting in the predicted delay value.

[0031] Optionally, determining the leakage monitoring result for the immersion liquid cooling system based on the predicted time delay value includes:

[0032] The probability of leakage is determined based on the predicted delay value;

[0033] If the leakage probability is greater than a preset probability threshold, the leakage monitoring result for the immersion liquid cooling system is determined to be a leakage.

[0034] If the leakage probability is less than or equal to the preset probability threshold, the leakage monitoring result for the immersion liquid cooling system is determined to be no leakage.

[0035] Optionally, determining the leakage monitoring result for the immersion liquid cooling system based on the predicted time delay value includes:

[0036] If the predicted delay value is less than the preset delay value, the leakage monitoring result for the immersion liquid cooling system is determined to be leakage.

[0037] If the predicted delay value is greater than or equal to less than the preset delay value, the leakage monitoring result for the immersion liquid cooling system is determined to be no leakage.

[0038] Optionally, after determining that the leakage monitoring result for the immersion liquid cooling system is a leakage, the method further includes:

[0039] Leakage alerts are sent via network serial port to notify the administrator to perform equipment maintenance.

[0040] This application also provides a leakage monitoring device based on an immersion liquid cooling system, comprising:

[0041] The acquisition module is configured to acquire the signal delay value of the differential pair, which is immersed in the coolant of the immersion liquid cooling system;

[0042] The first determining module is configured to determine the predicted delay value based on the signal delay value and a preset gray model.

[0043] The second determining module is configured to determine the leakage monitoring result for the immersion liquid cooling system based on the predicted time delay value.

[0044] This application also provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of the leakage monitoring method based on an immersion liquid cooling system as described above.

[0045] This application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the leakage monitoring method based on the immersion liquid cooling system described above.

[0046] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the leakage monitoring method based on an immersion liquid cooling system as described above.

[0047] The leakage monitoring method and apparatus based on an immersion liquid cooling system provided in this application acquires the signal delay value of a differential pair, wherein the differential pair is immersed in the coolant of the immersion liquid cooling system; determines a predicted delay value based on the signal delay value and a preset gray model; and determines the leakage monitoring result for the immersion liquid cooling system based on the predicted delay value. By using a preset gray model to predict in advance whether a leakage has occurred in an immersion heat dissipation server system or an immersion liquid cooling system, leakage detection prediction can be achieved, and an alarm mechanism can be triggered in advance, effectively preventing system or board failure due to leakage. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 This is a schematic diagram of the single-phase immersion liquid cooling system provided in this application;

[0050] Figure 2 This is a schematic diagram of the two-phase immersion liquid cooling system provided in this application;

[0051] Figure 3 This is one of the flowcharts of the leakage monitoring method based on an immersion liquid cooling system provided in this application;

[0052] Figure 4 This is a schematic diagram of the structure of the curved high-speed backplane connector provided in this application;

[0053] Figure 5 This is a flowchart illustrating the process for determining the predicted delay value provided in this application;

[0054] Figure 6 This is the second flowchart of the leakage monitoring method based on an immersion liquid cooling system provided in this application:

[0055] Figure 7 This is a schematic diagram of the leakage monitoring device based on an immersion liquid cooling system provided in this application;

[0056] Figure 8 This is a schematic diagram of the structure of the electronic device provided in this application. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0058] The terms "first," "second," etc., used in this application's specification are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class, without limiting the number of objects; for example, a first object can be one or more. Furthermore, in the specification, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects have an "or" relationship.

[0059] First, the relevant content involved in this invention will be explained.

[0060] ESG (Environmental, Social and Governance) refers to environmental, social and corporate governance.

[0061] IT (Information Technology) refers to information technology.

[0062] Immersion liquid cooling, also known as submersion liquid cooling.

[0063] Like any powerful computer hardware, CPUs and GPUs generate heat during operation and require proper cooling to achieve maximum performance. During normal operation, the transistors inside the CPU / GPU convert electricity into heat, which raises the CPU / GPU's temperature. If this heat does not have an effective way to dissipate, the CPU / GPU will exceed its safe operating temperature and malfunction.

[0064] But what is the best way to keep CPUs / GPUs operating at ideal temperatures? There are many ways to cool processors, and most desktop computers and laptops use air-cooled or liquid-cooled radiators. However, as the temperatures of critical components in IT equipment continue to rise, data center operators are scrambling to find energy-efficient cooling solutions that align with sustainable investment in ESG (Environmental, Social, and Governance) principles. Immersion liquid cooling technology offers better cooling efficiency compared to air conduction. By contacting the non-conductive liquid medium for heat dissipation, immersion liquid cooling's overall architecture directly contacts the entire heat source for better heat dissipation.

[0065] Immersion liquid cooling involves directly submerging IT equipment in a coolant (liquid medium), relying on the coolant to absorb the heat generated by the equipment. Depending on whether the coolant undergoes a phase change during the circulation process, it can be divided into single-phase immersion liquid cooling systems and two-phase immersion liquid cooling systems.

[0066] See Figure 1 , Figure 1 This is a schematic diagram of the single-phase immersion liquid cooling system provided in this application: In a single-phase immersion liquid cooling system, IT equipment (servers) are also installed in a non-conductive liquid medium tank (coolant). This non-conductive medium uses hydrocarbon fluids similar to mineral oil or vegetable oil. The heat generated by the server comes into direct contact with the heat-generating components through the cooling medium. Unlike two-phase immersion liquid cooling, a single-phase immersion liquid cooling system requires a coolant pump in the cooling device (coolant distribution device) to draw the medium into the cooling device for heat dissipation circulation and heat exchange. If the coolant temperature cannot be reduced in time, an additional set of cooling equipment (such as an evaporative cooling tower, coolant pump, and existing cooling water circulation) is needed to complete the entire heat conversion cycle. Due to the high boiling point and non-volatile properties of this type of liquid, the liquid tank of a single-phase immersion liquid cooling system usually does not require stringent closed-loop design and environmental control requirements.

[0067] See Figure 2 , Figure 2This is a schematic diagram of the two-phase immersion liquid cooling system provided in this application: In the two-phase immersion liquid cooling system, IT equipment (servers) are installed in a specially designed sealed chassis within a liquid water tank, which contains non-conductive fluorocarbons. Liquid fluorocarbons have a low boiling point (typically around 50°C). The heat generated by the server causes the liquid around the heat-generating components to boil and undergo a phase change, producing bubbles. This naturally creates a flow of hot and cold water, carrying away heat—this is the origin of "two-phase" immersion liquid cooling. The heat dissipation causes the bubbles from the phase change to rise and transform into vapor. The vapor is then collected and cooled through condenser coils in the sealed tank, returning to a liquid state and falling back into a backup liquid pool. The sealed water tank uses an environmental controller to maintain the continuous phase change process of the coolant, cyclically circulating for heat exchange.

[0068] Existing immersion liquid cooling systems, whether single-phase or two-phase, require a dielectric fluid in direct contact with the electronic components. Generally, dielectric fluids used for immersion cooling fall into two categories: hydrocarbons (i.e., mineral oil, synthetic oil, or bio-oil) and fluorocarbons (engineered fluids). The dielectric fluid is categorized as either single-phase or two-phase, depending on whether the cooling fluid becomes a gas during the cooling cycle. Single-phase immersion liquid cooling systems utilize heat exchange by circulating the dielectric fluid between high-temperature electronic components. Two-phase immersion liquid cooling systems use fluorocarbons that boil at low temperatures, transferring heat from the components in gaseous form. This gas is recovered, condensed through a heat exchanger, and returned to the components. Therefore, if the dielectric fluid used in single-phase or two-phase immersion liquid cooling systems is at risk of leakage, the immersion cooling may not effectively dissipate heat from the system.

[0069] There are two main categories of possible causes for leakage: one is assembly problems, such as inaccurate positioning of the interface or poor connector manufacturing tolerances, which can lead to leakage; the other is aging of the pipelines used for circulating coolant in immersion cooling systems, as the pipelines become brittle due to long-term use, which can also cause leakage.

[0070] Once a leak occurs, it is irreversible because the dielectric coolant will spread to the outside of the sealed enclosure's piping.

[0071] Therefore, regarding how to perform dielectric fluid leakage detection and prevent dielectric fluid coolant from spreading outside the chassis system, this application provides a leakage monitoring method and device based on an immersion liquid cooling system. This method acquires the signal delay value of a differential pair, which is immersed in the coolant of the immersion liquid cooling system. Based on the signal delay value and a preset gray model, a predicted delay value is determined. Based on the predicted delay value, the leakage monitoring result for the immersion liquid cooling system is determined. By using a preset gray model to pre-determine whether a leakage has occurred in the immersion cooling server system or the immersion liquid cooling system, leakage detection and prediction can be achieved, triggering an alarm mechanism in advance. This can effectively prevent system or board cooling failure due to leakage.

[0072] The leakage monitoring method based on an immersion liquid cooling system provided in this application will be described in detail below with reference to the accompanying drawings, through specific embodiments and application scenarios.

[0073] like Figure 3 As shown in the embodiment of this application, a leakage monitoring method based on an immersion liquid cooling system is provided. This method may include the following steps 301 and 303:

[0074] Step 1301: Obtain the signal delay value of the differential pair, wherein the differential pair is immersed in the coolant of the immersion liquid cooling system.

[0075] Specifically, the signal delay value refers to the delay value of the differential pair signal. An immersion liquid cooling system can be a single-phase immersion liquid cooling system or a two-phase immersion liquid cooling system.

[0076] Coolant is a node fluid or liquid medium. In the case of a single-phase immersion liquid cooling system, the coolant can be a hydrocarbon, such as mineral oil, synthetic oil, or bio-oil. In the case of a two-phase immersion liquid cooling system, the coolant can be a fluorocarbon, such as a whole-engineering fluid or fluorinated liquid.

[0077] In practical applications, a dielectric coolant level detection mechanism can be used. Due to the non-conductive physical property of dielectric coolants, there are different dielectric coefficient formulations. Currently, the most commonly used dielectric coolants on the market are fluorinated liquid formulations, such as the four fluorinated liquids shown in Table 1: FC-3284, FC-72, FC-3283, and FC-40.

[0078] Table 1

[0079]

[0080]

[0081] The speed at which a signal propagates on a transmission line is not the speed at which electrons move in the transmission line, but rather the speed at which electromagnetic waves propagate on the transmission line. The speed at which the electromagnetic field changes determines the speed of the signal, while the speed at which the electric and magnetic fields change is related to the properties of the dielectric material and some constants, as shown in formula (1):

[0082]

[0083] Where V is the speed of signal propagation on the transmission line or the speed of electromagnetic wave propagation on the transmission line, ε0 is the dielectric constant in vacuum, and ε r μ is the relative permittivity, μ0 is the permeability in vacuum, and μ r is the relative permeability.

[0084] For example, see Figure 4 , Figure 4 This is a schematic diagram of the structure of the curved high-speed backplane connector provided in this application: In the differential pair wiring area within the wafer, the trace length on the inner side of the differential pair is generally shorter than the trace length on the outer side. Therefore, it is necessary to adjust the delay by changing the medium of the transmission line. Generally, delay compensation is achieved by adding plastic or air.

[0085] After delay compensation, the delay of connectors under normal air-cooled operation can generally be adjusted to within 1ps. However, after immersion in fluorinated liquid, the reference medium around the unequal length sections of the differential pair traces changes, leading to varying degrees of delay degradation under liquid-cooled operation. This is an unavoidable situation for bend-type high-speed connectors during use. For fully shielded high-speed connectors, if the connector trace area is completely immersed in liquid, it will result in a delay degradation of approximately 5ps. For high-speed transmission systems, the delay changes under air-cooled and liquid-cooled conditions need to be considered in advance. Table 2 shows the delay difference within the differential pair.

[0086] Table 2

[0087] Path 1 0 5.1 Path 2 -0.5 4.4 Path 3 -0.7 6.2

[0088] Among them, pathways 1-3 are the pathways from the outside to the inside of the difference pair.

[0089] The difference in signal transmission delay caused by the different dielectric constants of air and fluorinated liquid is used to detect and determine if leakage has occurred. In the case of a server system using immersion liquid cooling, if the differential pair signal delay difference is reduced, it indicates that the high-speed connector is exposed to the air, thus indicating that fluorinated liquid has leaked.

[0090] In other words, the shorter the delay value, the better in the event of leakage. Therefore, leakage can be detected by measuring the signal delay value of the differential pair.

[0091] Because the connector is immersed in the coolant of the immersion liquid cooling system, the differential pair in the connector is also immersed in the coolant of the immersion liquid cooling system.

[0092] Specifically, the BMC (Baseboard Management Controller) reads the delay values ​​of the differential pair signals in the connector path, i.e., the signal delay values ​​of the differential pair, to predict whether leakage has occurred.

[0093] Step 302: Determine the predicted delay value based on the signal delay value and the preset gray model.

[0094] Specifically, the pre-defined grey model is based on grey theory, which encompasses various application areas, among which grey theory prediction is an extremely important part.

[0095] The presupposed grey model is the GM(1,1) grey model, which is the most widely used predictive model in grey theory. Past research has shown promising results in fields such as economics, finance, and engineering. Compared to traditional predictive models that require a large amount of historical data to build, the GM(1,1) grey model can be built quickly without a large amount of data. According to its definition, a minimum of only four data points are needed to build the model and make predictions, which is its greatest advantage.

[0096] The basis of the pre-defined grey model is that the data of the past sequence of events in the system are processed by grey successive accumulation generation (AGO) calculation to change the data level, thereby discovering and revealing potential patterns, and predicting possible future situations from the patterns.

[0097] Therefore, based on the obtained signal delay value, the signal delay value can be input into a preset gray model to obtain the predicted delay value.

[0098] Step 303: Based on the predicted time delay value, determine the leakage monitoring result for the immersion liquid cooling system.

[0099] Furthermore, based on the predicted time delay value, the monitoring results of whether the immersion liquid cooling system has leaked are determined, that is, the leakage monitoring results are determined.

[0100] The leakage detection method for an immersion liquid cooling system provided in this application acquires the signal delay value of a differential pair immersed in the coolant of the immersion liquid cooling system; determines a predicted delay value based on the signal delay value and a preset gray model; and determines the leakage detection result for the immersion liquid cooling system based on the predicted delay value. By using a preset gray model to predict whether leakage has occurred in the immersion heat dissipation server system or the immersion liquid cooling system, leakage detection prediction can be achieved, and an alarm mechanism can be triggered in advance, effectively preventing system or board cooling failure due to leakage.

[0101] In one or more optional embodiments of this application, the specific implementation process of obtaining the signal delay value of the differential pair can be as follows:

[0102] Based on the connector, the differential pairs in the path are identified, and the signal delay values ​​of the differential pairs are obtained.

[0103] In practical applications, the BMC can read the differential pair of a certain path in the connector and detect the time delay of the differential pair signal to obtain the signal delay value of the differential pair. This improves the accuracy and speed of obtaining the signal delay value.

[0104] In one or more optional embodiments of this application, the specific implementation process of obtaining the signal delay value of the differential pair can be as follows:

[0105] The signal delay value of the differential pair is periodically acquired, with an acquisition period of less than or equal to 10 seconds.

[0106] In practical applications, since the preset gray model requires four data points, multiple signal delay values ​​need to be acquired. Therefore, the signal delay values ​​of the differential pairs can be acquired periodically according to the acquisition cycle. This ensures the accuracy and reliability of the predicted delay values.

[0107] For example, the initial value of the signal delay value sampling of the differential pair is 4, and the sampling interval can be set to 10 seconds, that is, the acquisition period.

[0108] In one or more optional embodiments of this application, the specific implementation process of determining the predicted delay value based on the signal delay value and the preset gray model is as follows:

[0109] Based on the signal delay values, a first delay value sample set is obtained, which includes at least four signal delay values.

[0110] The first time delay numerical sample set is corrected to obtain the second time delay numerical sample set;

[0111] Based on the second time delay value sample set and the preset gray model, the predicted time delay value is obtained.

[0112] In practical applications, at least four signal delay values ​​can be combined to form a first delay value sample set. Then, the number of samples in the first delay value sample set, i.e., the number of signal delay values, is adjusted to obtain a second delay value sample set, thereby achieving an optimized average error value. The second delay value sample set contains sample delay values.

[0113] Furthermore, the second time delay numerical sample set can be input into the preset gray model to obtain the predicted time delay value.

[0114] This can reduce the error in the predicted delay value, thereby improving the accuracy of the predicted delay value.

[0115] In one or more optional embodiments of this application, the specific implementation process of obtaining the predicted time delay value based on the second time delay value sample set and the preset gray model can be as follows:

[0116] Sort the sample delay values ​​in the second delay value sample set to obtain the first delay value sequence;

[0117] The first time-delayed numerical sequence is accumulated and transformed to obtain the second time-delayed numerical sequence.

[0118] The second time delay value sequence is input into the preset gray model for prediction to obtain the predicted time delay value.

[0119] Specifically, see Figure 5 , Figure 5 This application provides a flowchart illustrating the process of determining the predicted time delay value: The sample time delay values ​​in the second time delay value sample set are sorted to obtain the first time delay value sequence, i.e., the sample time delay values ​​(original data) are arranged in sequence. Wherein, the first time delay value sequence X... (0) As shown in formula (2):

[0120] X (0) ={X (0) (1), X (0) (2), ..., X (0) (i), ..., X (0) (n)} (2)

[0121] Where X (0) (i) represents the sample delay value at time order i; n represents the last time order, and its value should not be less than 4.

[0122] Furthermore, using the first time-delay numerical sequence X (0)The cumulative generation operation transforms the data into a second-delay numerical sequence X. (1) That is, to accumulate and generate X (1) In this way, sequences that were originally highly variable and irregular can be transformed to exhibit a smoother and more obvious regularity. The second time-delay numerical sequence X after AGO transformation... (1) Compared to the first time-delayed numerical sequence X (0) It is easier to build a model to approximate it. Among them, the second time-delay numerical sequence X (1) As shown in formula (3):

[0123] X (1) ={X (1) (1), X (1) (2), ..., X (1) (k), ..., X (1) (n)} (3)

[0124] Among them, the second time-delay numerical sequence X (1) With the first time-delay numerical sequence X (0) The transformation relationship is shown in equation (4):

[0125]

[0126] In one or more optional embodiments of this application, the preset gray model includes undetermined parameters; the specific implementation process of inputting the second time delay value sequence into the preset gray model for prediction to obtain the predicted time delay value can be as follows:

[0127] The preset gray model is modified and adjusted to obtain the target gray model;

[0128] The second time delay numerical sequence is input into the preset gray model, and the preset gray model is deformed to obtain the matrix representation corresponding to the preset gray model;

[0129] The matrix representation is simplified to obtain the matrix equation;

[0130] The error matrix is ​​determined based on the least squares method and the matrix equation.

[0131] The error matrix is ​​solved to obtain the estimated values ​​of the undetermined parameters;

[0132] Substituting the estimated value into the target gray model yields the backup delay value;

[0133] The backup delay value is inversely accumulated to generate a restoration process, resulting in the predicted delay value.

[0134] Specifically, see Figure 5Establish the GM(1,1) model, which is a preset gray model, and express it in terms of gray differential equations, as shown in formula (5).

[0135] X (0) (k)+aZ (1) (k)=b, k=2, 3,...,n (5)

[0136] Among them, Z (1) (k)=αX (1) (k)+(1-α)X (1) (k-1), where a and b are parameters to be determined, and α is the adjusting factor, which is usually set to 0.5 and is modified only when the data structure is special.

[0137] By modifying and adjusting the gray differential equation of formula (5), the target gray model can be solved, as shown in the well-known (6).

[0138]

[0139] Input the second time delay numerical sequence into formula (5) to solve for the undetermined parameters a and b, and obtain the matrix representation corresponding to the preset gray model, as shown in formula (7).

[0140]

[0141] make The matrix representation is simplified to obtain the matrix equation, as shown in formula (8).

[0142] Y = BA (8)

[0143] Then, the value of matrix A can be estimated using the ordinary least-squares method. Therefore, the error matrix is ​​first set as shown in formula (9).

[0144] E = Y - BA

[0145] Q = (Y - BA) T (Y-BA) (9)

[0146] Furthermore, the parameter A that minimizes the Q value can be estimated according to formula (10).

[0147]

[0148]

[0149] Solving equation (10) yields estimated values ​​of the undetermined parameters. and Will and Substitute into formula (6) to solve for X (1) (k+1) represents the spare delay value.

[0150] Then obtain X (1) (k+1) is inversely accumulated to generate and restore X. (0) (k+1), the resulting value is the desired prediction value, i.e. the prediction delay value.

[0151] Specifically, formula (11) can be derived from formula (4).

[0152]

[0153] The predicted value at the future time sequence number ξ, i.e., the prediction delay value. The predicted time delay value can be obtained by formula (12).

[0154]

[0155] In one or more optional embodiments of this application, the specific implementation process of determining the leakage monitoring result for the immersion liquid cooling system based on the predicted time delay value can be as follows:

[0156] The probability of leakage is determined based on the predicted delay value;

[0157] If the leakage probability is greater than a preset probability threshold, the leakage monitoring result for the immersion liquid cooling system is determined to be a leakage.

[0158] If the leakage probability is less than or equal to the preset probability threshold, the leakage monitoring result for the immersion liquid cooling system is determined to be no leakage.

[0159] Specifically, the presence of leakage can be predicted based on the predicted delay value: the probability of leakage can be determined based on the predicted delay value and a preset probability determination rule.

[0160] Furthermore, if the leakage probability is greater than a preset probability threshold, the leakage monitoring result for the immersion liquid cooling system is determined to be a leakage; if the leakage probability is less than or equal to the preset probability threshold, the leakage monitoring result for the immersion liquid cooling system is determined to be no leakage. This further improves the accuracy of leakage monitoring.

[0161] In one or more optional embodiments of this application, the specific implementation process of determining the leakage monitoring result for the immersion liquid cooling system based on the predicted time delay value can be as follows:

[0162] If the predicted delay value is less than the preset delay value, the leakage monitoring result for the immersion liquid cooling system is determined to be leakage.

[0163] If the predicted delay value is greater than or equal to less than the preset delay value, the leakage monitoring result for the immersion liquid cooling system is determined to be no leakage.

[0164] Specifically, the presence of leakage can be pre-determined based on the predicted time delay value: if the predicted time delay value is less than a preset time delay value, the leakage monitoring result for the immersion liquid cooling system is determined to be leakage; if the predicted time delay value is greater than or equal to less than the preset time delay value, the leakage monitoring result for the immersion liquid cooling system is determined to be no leakage. This improves the speed and timeliness of leakage monitoring.

[0165] In one or more optional embodiments of this application, after determining that the leakage monitoring result for the immersion liquid cooling system is a leakage, the method further includes:

[0166] Leakage alerts are sent via network serial port to notify the administrator to perform equipment maintenance.

[0167] Specifically, when a leak is detected, the BMC can notify the administrator to perform equipment maintenance via network serial port alert.

[0168] The following is combined Figure 6 The leakage monitoring method based on an immersion liquid cooling system provided in this application is further explained. Figure 6 This is the second flowchart of the leakage monitoring method based on an immersion liquid cooling system provided in this application, including:

[0169] BMC predicts whether leakage has occurred by reading the signal delay values ​​of differential pairs in the connector path and using a preset gray model. In other words, BMC can detect the delay of differential pair signals by reading the differential pair (differential pair information of high-speed connector) in a certain path of the connector, thereby obtaining the signal delay value of the differential pair.

[0170] The initial value of the signal delay values ​​of the differential pair of the preset gray model is 4. The sampling interval can be set to 10 seconds to obtain the first delay value sample set.

[0171] The number of samples in the first time-delay numerical sample set is adjusted to achieve an optimized average error value, thus obtaining the second time-delay numerical sample set. The second time-delay numerical samples are then processed according to... Figure 5 The process is performed to obtain the predicted delay value, which is based on the preset gray model.

[0172] Based on the obtained time delay value, the presence or absence of liquid leakage can be predicted. When a leakage is detected, the BMC can notify the administrator to perform equipment maintenance through the network serial port.

[0173] It should be noted that the leakage monitoring method based on an immersion liquid cooling system provided in this application embodiment can be executed by a leakage monitoring device based on an immersion liquid cooling system, or by the leakage monitoring device itself. The device includes a control module for executing the leakage monitoring method based on an immersion liquid cooling system. This application embodiment uses the example of a leakage monitoring device based on an immersion liquid cooling system executing the leakage monitoring method to illustrate the leakage monitoring device based on an immersion liquid cooling system provided in this application embodiment.

[0174] It should be noted that, in the embodiments of this application, the leakage monitoring methods based on immersion liquid cooling systems shown in the accompanying drawings are all illustrated by way of example with reference to one of the accompanying drawings in the embodiments of this application. In specific implementation, the leakage monitoring methods based on immersion liquid cooling systems shown in the accompanying drawings can also be implemented in conjunction with any other accompanying drawings shown in the above embodiments, which will not be elaborated here.

[0175] The leakage monitoring device based on an immersion liquid cooling system provided in this application is described below. The leakage monitoring method based on an immersion liquid cooling system described below can be referred to in correspondence with the leakage monitoring method described above.

[0176] Figure 7 A schematic diagram of the leakage monitoring device based on an immersion liquid cooling system provided in this application embodiment is shown below. Figure 7 As shown, it specifically includes:

[0177] The acquisition module 701 is configured to acquire the signal delay value of the differential pair, which is immersed in the coolant of the immersion liquid cooling system;

[0178] The first determining module 702 is configured to determine the predicted delay value based on the signal delay value and a preset gray model.

[0179] The second determining module 703 is configured to determine the leakage monitoring result for the immersion liquid cooling system based on the predicted time delay value.

[0180] The leakage monitoring device based on an immersion liquid cooling system provided in this application acquires the signal delay value of a differential pair immersed in the coolant of the immersion liquid cooling system; determines a predicted delay value based on the signal delay value and a preset gray model; and determines the leakage monitoring result for the immersion liquid cooling system based on the predicted delay value. By using a preset gray model to predict whether leakage has occurred in the immersion heat dissipation server system or the immersion liquid cooling system, leakage detection and prediction can be achieved, triggering an alarm mechanism in advance, and effectively preventing system or board cooling failure due to leakage.

[0181] In one or more optional embodiments of this application, the acquisition module 701 is further configured to:

[0182] Based on the connector, the differential pairs in the path are identified, and the signal delay values ​​of the differential pairs are obtained.

[0183] In one or more optional embodiments of this application, the acquisition module 701 is further configured to:

[0184] The signal delay value of the differential pair is periodically acquired, with an acquisition period of less than or equal to 10 seconds.

[0185] In one or more optional embodiments of this application, the first determining module 702 is further configured to:

[0186] Based on the signal delay values, a first delay value sample set is obtained, which includes at least four signal delay values.

[0187] The first time delay numerical sample set is corrected to obtain the second time delay numerical sample set;

[0188] Based on the second time delay value sample set and the preset gray model, the predicted time delay value is obtained.

[0189] In one or more optional embodiments of this application, the first determining module 702 is further configured to:

[0190] Sort the sample delay values ​​in the second delay value sample set to obtain the first delay value sequence;

[0191] The first time-delayed numerical sequence is accumulated and transformed to obtain the second time-delayed numerical sequence.

[0192] The second time delay value sequence is input into the preset gray model for prediction to obtain the predicted time delay value.

[0193] In one or more optional embodiments of this application, the preset gray model includes parameters to be determined;

[0194] The first determining module 702 is further configured to:

[0195] The preset gray model is modified and adjusted to obtain the target gray model;

[0196] The second time delay numerical sequence is input into the preset gray model, and the preset gray model is deformed to obtain the matrix representation corresponding to the preset gray model;

[0197] The matrix representation is simplified to obtain the matrix equation;

[0198] The error matrix is ​​determined based on the least squares method and the matrix equation.

[0199] The error matrix is ​​solved to obtain the estimated values ​​of the undetermined parameters;

[0200] Substituting the estimated value into the target gray model yields the backup delay value;

[0201] The backup delay value is inversely accumulated to generate a restoration process, resulting in the predicted delay value.

[0202] In one or more optional embodiments of this application, the second determining module 703 is further configured to:

[0203] The probability of leakage is determined based on the predicted delay value;

[0204] If the leakage probability is greater than a preset probability threshold, the leakage monitoring result for the immersion liquid cooling system is determined to be a leakage.

[0205] If the leakage probability is less than or equal to the preset probability threshold, the leakage monitoring result for the immersion liquid cooling system is determined to be no leakage.

[0206] In one or more optional embodiments of this application, the second determining module 703 is further configured to:

[0207] If the predicted delay value is less than the preset delay value, the leakage monitoring result for the immersion liquid cooling system is determined to be leakage.

[0208] If the predicted delay value is greater than or equal to less than the preset delay value, the leakage monitoring result for the immersion liquid cooling system is determined to be no leakage.

[0209] In one or more optional embodiments of this application, the leakage monitoring device based on the immersion liquid cooling system further includes an alarm module configured to:

[0210] Leakage alerts are sent via network serial port to notify the administrator to perform equipment maintenance.

[0211] Figure 8 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 8 As shown, the electronic device may include a processor 810, a communication interface 820, a memory 830, and a communication bus 840, wherein the processor 810, the communication interface 820, and the memory 830 communicate with each other through the communication bus 840. The processor 810 can call logic instructions in the memory 830 to execute a leakage monitoring method based on an immersion liquid cooling system. The method includes: acquiring the signal delay value of a differential pair, wherein the differential pair is immersed in the coolant of the immersion liquid cooling system; determining a predicted delay value based on the signal delay value and a preset gray model; and determining a leakage monitoring result for the immersion liquid cooling system based on the predicted delay value.

[0212] Furthermore, the logical instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0213] On the other hand, this application also provides a computer program product, which includes a computer program stored on a computer-readable storage medium. The computer program includes program instructions, and when the program instructions are executed by a computer, the computer is able to execute the leakage monitoring method based on the immersion liquid cooling system provided by the above methods. The method includes: acquiring the signal delay value of a differential pair, wherein the differential pair is immersed in the coolant of the immersion liquid cooling system; determining a predicted delay value based on the signal delay value and a preset gray model; and determining a leakage monitoring result for the immersion liquid cooling system based on the predicted delay value.

[0214] In another aspect, this application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, is implemented to perform the above-described methods for monitoring leakage in an immersion liquid cooling system. The method includes: acquiring signal delay values ​​of differential pairs, the differential pairs being immersed in the coolant of the immersion liquid cooling system; determining a predicted delay value based on the signal delay value and a preset gray model; and determining a leakage monitoring result for the immersion liquid cooling system based on the predicted delay value.

[0215] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0216] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0217] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for monitoring leakage in an immersion liquid cooling system, characterized in that, include: Obtain the signal delay value of the differential pair, which is immersed in the coolant of the immersion liquid cooling system; The predicted delay value is determined based on the signal delay value and the preset gray model; Based on the predicted time delay value, the leakage monitoring results for the immersion liquid cooling system are determined; The step of determining the predicted delay value based on the signal delay value and the preset gray model includes: Based on the signal delay values, a first delay value sample set is obtained, which includes at least four signal delay values. The first time delay numerical sample set is corrected to obtain the second time delay numerical sample set; Sort the sample delay values ​​in the second delay value sample set to obtain the first delay value sequence; The first time-delayed numerical sequence is accumulated and transformed to obtain the second time-delayed numerical sequence. The second time delay value sequence is input into the preset gray model for prediction to obtain the predicted time delay value.

2. The leakage monitoring method based on an immersion liquid cooling system according to claim 1, characterized in that, The step of obtaining the signal delay value of the differential pair includes: Based on the connector, the differential pairs in the path are identified, and the signal delay values ​​of the differential pairs are obtained.

3. The leakage monitoring method based on an immersion liquid cooling system according to claim 1 or 2, characterized in that, The step of obtaining the signal delay value of the differential pair includes: The signal delay value of the differential pair is periodically acquired, with an acquisition period of less than or equal to 10 seconds.

4. The leakage monitoring method based on an immersion liquid cooling system according to claim 1, characterized in that, The preset gray model includes parameters to be determined; The step of inputting the second time delay numerical sequence into the preset gray model for prediction to obtain the predicted time delay value includes: The preset gray model is modified and adjusted to obtain the target gray model; The second time delay numerical sequence is input into the preset gray model, and the preset gray model is deformed to obtain the matrix representation corresponding to the preset gray model; The matrix representation is simplified to obtain the matrix equation; The error matrix is ​​determined based on the least squares method and the matrix equation. The error matrix is ​​solved to obtain the estimated values ​​of the undetermined parameters; Substituting the estimated value into the target gray model yields the backup delay value; The backup delay value is inversely accumulated to generate a restoration process, resulting in the predicted delay value.

5. The leakage monitoring method based on an immersion liquid cooling system according to claim 1, characterized in that, The step of determining the leakage monitoring results for the immersion liquid cooling system based on the predicted time delay value includes: The probability of leakage is determined based on the predicted delay value; If the leakage probability is greater than a preset probability threshold, the leakage monitoring result for the immersion liquid cooling system is determined to be a leakage. If the leakage probability is less than or equal to the preset probability threshold, the leakage monitoring result for the immersion liquid cooling system is determined to be no leakage.

6. The leakage monitoring method based on an immersion liquid cooling system according to claim 1, characterized in that, The step of determining the leakage monitoring results for the immersion liquid cooling system based on the predicted time delay value includes: If the predicted delay value is less than the preset delay value, the leakage monitoring result for the immersion liquid cooling system is determined to be leakage. If the predicted delay value is greater than or equal to less than the preset delay value, the leakage monitoring result for the immersion liquid cooling system is determined to be no leakage.

7. The leakage monitoring method based on an immersion liquid cooling system according to claim 5 or 6, characterized in that, After determining that the leakage monitoring result for the immersion liquid cooling system is a leakage, the process further includes: Leakage alerts are sent via network serial port to notify the administrator to perform equipment maintenance.

8. A leakage monitoring device based on an immersion liquid cooling system, characterized in that, include: The acquisition module is configured to acquire the signal delay value of the differential pair, which is immersed in the coolant of the immersion liquid cooling system; The first determining module is configured to determine the predicted delay value based on the signal delay value and a preset gray model. The second determining module is configured to determine the leakage monitoring result for the immersion liquid cooling system based on the predicted time delay value. The first determining module is specifically configured to obtain a first delay value sample set based on the signal delay values, the first delay value sample set including at least four signal delay values; correct the first delay value sample set to obtain a second delay value sample set; and sort the sample delay values ​​in the second delay value sample set to obtain a first delay value sequence. The first time delay numerical sequence is accumulated and transformed to obtain a second time delay numerical sequence; the second time delay numerical sequence is input into the preset gray model for prediction to obtain the predicted time delay value.

9. An electronic device, characterized in that, The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the steps of the leakage monitoring method based on an immersion liquid cooling system as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the steps of the leakage monitoring method based on an immersion liquid cooling system as described in any one of claims 1 to 7.

Images