A coolant leakage monitoring method, system and electronic device

Abstract

This application discloses a coolant leakage monitoring method, system, and electronic device, relating to the field of computer technology. By combining pressure sensing data and current monitoring images, the leakage hazard level of the target cabinet with a leakage point is determined. Then, based on the leakage hazard level of the target cabinet and the distance between the target cabinet and adjacent cabinets in the computer room, corresponding coolant leakage safety protection measures are taken for the target cabinet and its adjacent cabinets to avoid safety risks to the cabinet due to coolant leakage.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a method, system and electronic device for monitoring coolant leakage. Background Technology

[0002] As the demand for server computing power continues to increase, the power density of server racks is also increasing when multiple servers are deployed in a rack. Consequently, the heat flux density generated by the rack is also increasing. Therefore, cold plate liquid cooling systems have been widely used in server racks. However, cold plate liquid cooling systems pose a risk of coolant leakage. Therefore, how to monitor coolant leakage in server racks has become a key research topic.

[0003] In related technologies, a coolant leak in the server rack is typically detected when the coolant pipeline pressure reaches a threshold alarm, leading to the closure of the cooling zone valves to stop the leak. However, since coolant leaks pose a safety risk to the server rack, there is an urgent need for a coolant leak monitoring method that can ensure server rack safety, which is of great significance for improving server rack security. Summary of the Invention

[0004] This application provides a method, system, and electronic device for monitoring coolant leakage, in order to at least solve the problem of safety risks to server racks caused by coolant leakage in related technologies.

[0005] This application provides a method for monitoring coolant leakage, including:

[0006] Acquire pressure sensing data of the liquid cooling system of the target cabinet;

[0007] When pressure sensing data indicates a leak in the liquid cooling system, control at least one image acquisition device to acquire the current monitoring image of the area under test where the leak is located.

[0008] Based on the current monitoring images of the area to be tested, determine the coolant diffusion information at the leak point;

[0009] Based on the degree of damage at the leak point as characterized by pressure sensing data and the coolant diffusion information at the leak point, the leakage hazard level of the target cabinet is determined.

[0010] Based on the leakage hazard level of the target cabinet and the distance between the target cabinet and adjacent cabinets in the computer room, determine the safety protection strategy for the target cabinet and adjacent cabinets, and implement corresponding coolant leakage safety protection measures for the target cabinet and adjacent cabinets in accordance with the safety protection strategy for the target cabinet and adjacent cabinets.

[0011] This application also provides a coolant leakage monitoring device, including:

[0012] The acquisition module is used to acquire pressure sensing data of the liquid cooling system of the target cabinet;

[0013] The image acquisition module is used to control at least one image acquisition device to acquire the current monitoring image of the area under test where the leak point is located when the pressure sensing data indicates that there is a leak point in the liquid cooling system.

[0014] The first determining module is used to determine the coolant diffusion information at the leak point based on the current monitoring image of the area to be tested;

[0015] The second determination module is used to determine the leakage hazard level of the target cabinet based on the degree of damage at the leak point characterized by pressure sensing data and the coolant diffusion information at the leak point.

[0016] The protection module is used to determine the safety protection strategy for the target cabinet and adjacent cabinets based on the leakage hazard level of the target cabinet and the distance between the target cabinet and adjacent cabinets in the computer room, so as to implement corresponding coolant leakage safety protection measures for the target cabinet and adjacent cabinets in accordance with the safety protection strategy for the target cabinet and adjacent cabinets.

[0017] This application also provides a coolant leakage monitoring system, including: a control component, a pressure sensor and an image acquisition device, wherein the pressure sensor is deployed at the inlet and outlet positions of multiple test points in the liquid cooling system of the target cabinet;

[0018] The pressure sensor is used to acquire pressure sensing data of the liquid cooling system of the target cabinet;

[0019] Image acquisition equipment is used to acquire monitoring images of the area where each test point is located;

[0020] The control component uses any of the above-mentioned coolant leakage monitoring methods to monitor coolant leakage in the target cabinet.

[0021] This application also provides an electronic device, including: a memory for storing a computer program; and a processor for executing the computer program to implement the steps of any of the above-described coolant leakage monitoring methods.

[0022] This application also provides a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, it implements the steps of any of the above-described coolant leakage monitoring methods.

[0023] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of any of the above-described coolant leakage monitoring methods.

[0024] This application determines the leakage hazard level of the target cabinet with a leakage point by combining pressure sensor data and current monitoring images. Then, based on the leakage hazard level of the target cabinet and the distance between the target cabinet and adjacent cabinets in the computer room, corresponding coolant leakage safety protection measures are taken for the target cabinet and its adjacent cabinets to avoid safety risks to the cabinet due to coolant leakage. Attached Figure Description

[0025] To more clearly illustrate the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 A flowchart illustrating the coolant leakage monitoring method provided in this application embodiment;

[0027] Figure 2 This is a schematic diagram of the overall process of the coolant leakage monitoring method provided in the embodiments of this application;

[0028] Figure 3 This is a schematic diagram of the structure of the coolant leakage monitoring device provided in the embodiments of this application;

[0029] Figure 4 This is a schematic diagram of the structure of the coolant leakage monitoring system provided in the embodiments of this application;

[0030] Figure 5 A schematic diagram of the structure of an exemplary coolant leakage monitoring system provided in an embodiment of this application;

[0031] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0032] The technical solutions of the embodiments 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, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of this application.

[0033] It should be noted that, in the description of this application, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. The terms "first," "second," etc., in this application are used to distinguish similar objects and are not used to describe a specific order or sequence.

[0034] The complex piping structure of liquid cooling systems faces risks such as microchannel rupture in cold plates and joint aging failure during long-term operation. Experimental data shows that when a leak occurs, the interruption of coolant supply leads to localized heat accumulation. The temperature of computing units such as the Graphics Processing Unit (GPU) in the affected rack can rise by more than 15°C within 5 minutes, triggering the GPU's self-protection mechanism to force it to reduce its frequency, which can lead to permanent damage in severe cases.

[0035] Current mainstream solutions only rely on pressure sensor threshold alarms and the closure of regional valves, lacking proactive protection mechanisms for affected computing devices. For example, when a GPU's core temperature exceeds 98°C due to thermal failure, the risk of transistor-level damage increases exponentially, necessitating GPU frequency throttling or server shutdown in specific thermal failure scenarios. More seriously, the thermal diffusion effect caused by leakage can spread to adjacent racks, forming a cascading failure chain. Therefore, a collaborative system integrating precise leakage diagnosis, dynamic thermal risk assessment, and tiered emergency control is urgently needed to maintain maximum continuity of computing services while ensuring hardware safety.

[0036] To address the aforementioned technical problems, embodiments of this application provide a method, system, and electronic device for monitoring coolant leakage. The method includes: acquiring pressure sensing data of the liquid cooling system of a target cabinet; when the pressure sensing data indicates a leak in the liquid cooling system, controlling at least one image acquisition device to acquire a current monitoring image of the area to be monitored where the leak point is located; determining coolant diffusion information of the leak point based on the current monitoring image of the area to be monitored; determining the leakage hazard level of the target cabinet based on the degree of damage to the leak point indicated by the pressure sensing data and the coolant diffusion information of the leak point; and determining a safety protection strategy for the target cabinet and adjacent cabinets based on the leakage hazard level of the target cabinet and the distance between the target cabinet and adjacent cabinets in the computer room, so as to implement corresponding coolant leakage safety protection measures for the target cabinet and adjacent cabinets according to the safety protection strategy for the target cabinet and adjacent cabinets. The method provided by the above solution determines the leakage hazard level of the target cabinet with a leakage point by combining pressure sensor data and current monitoring images. Then, based on the leakage hazard level of the target cabinet and the distance between the target cabinet and adjacent cabinets in the computer room, corresponding coolant leakage safety protection measures are taken for the target cabinet and its adjacent cabinets to avoid safety risks to the cabinet due to coolant leakage.

[0037] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0038] This application provides a method for monitoring coolant leakage, used to monitor coolant leakage in server racks and take corresponding safety precautions against coolant leakage. The execution subject of this application is an electronic device, such as a server, desktop computer, laptop computer, tablet computer, or other electronic devices that can be used for coolant leakage monitoring.

[0039] like Figure 1 The diagram shown is a flowchart illustrating a coolant leakage monitoring method provided in an embodiment of this application. The method includes:

[0040] Step 101: Obtain pressure sensing data of the liquid cooling system of the target cabinet.

[0041] The target rack is the rack where coolant leakage monitoring is performed. This rack houses multiple servers and other electronic devices; the rack can refer to a supernode server cluster. The liquid cooling system is used to cool the target rack and mainly consists of coolant, liquid cooling plates, circulation pipes, pumps, and valves. Pressure sensing data is collected by pressure sensors to reflect the pressure status within the liquid cooling system loop.

[0042] Step 102: When the pressure sensing data indicates that there is a leak in the liquid cooling system, control at least one image acquisition device to acquire the current monitoring image of the area to be tested where the leak is located.

[0043] Specifically, when pressure data triggers a leak warning, image acquisition devices such as near-infrared optical cameras are directed to the area to be tested. The area to be tested is the potential location of the leak, such as pipelines near the leak point. By acquiring current monitoring images of the area to be tested where the leak point is located, visual information is used to supplement the ambiguity of the pressure data, preparing for subsequent assessment of the extent of the leak.

[0044] One approach is to adjust the shooting angle and layout of the image acquisition devices to cover the entire area with as few image acquisition devices as possible.

[0045] Step 103: Determine the coolant diffusion information at the leak point based on the current monitoring image of the area to be tested.

[0046] It should be noted that pressure sensor data only reflects the extent of pipe damage, i.e., whether a leak has occurred, while the current monitoring image can reflect the state of the coolant leaking outside the pipe. Specifically, the larger the liquid stain area in the current monitoring image, the more coolant has leaked. Coolant diffusion information describes the diffusion state of the leaked coolant within the cabinet or server room, such as the amount of coolant diffusion (leakage) and diffusion rate.

[0047] Step 104: Determine the leakage hazard level of the target cabinet based on the degree of damage at the leak point as characterized by pressure sensing data and the coolant diffusion information at the leak point.

[0048] The degree of damage at the leak point refers to the extent of damage to the pipeline at the leak point, such as micro-leakage and pipeline rupture.

[0049] Specifically, by comprehensively considering the degree of damage at the leak point, the amount of coolant diffusion (leakage), and the diffusion rate, a precise assessment of the leakage hazard level of the target cabinet where the leak point is located was achieved, laying the foundation for improving the accuracy of subsequent safety protection strategy determination results.

[0050] In cases where multiple leak points exist in the target cabinet, the leakage hazard level of the target cabinet is determined by combining the pressure sensing data of multiple leak points to characterize the degree of damage at the leak points and the coolant diffusion information at the leak points.

[0051] Step 105: Based on the leakage hazard level of the target cabinet and the distance between the target cabinet and adjacent cabinets in the computer room, determine the safety protection strategy for the target cabinet and adjacent cabinets, and implement corresponding coolant leakage safety protection measures for the target cabinet and adjacent cabinets in accordance with the safety protection strategy for the target cabinet and adjacent cabinets.

[0052] Specifically, since there are multiple server racks in the computer room, after determining the leakage hazard level of any server rack, if the target server rack has a serious leakage and the adjacent server racks are close together, the leaked coolant may flow to the adjacent server racks. In fact, due to the weakened heat dissipation capacity of the target server rack, the heat generated by the target server rack may spread to the adjacent server racks, causing thermal interference to the target server rack. Therefore, it is necessary to take corresponding coolant leakage safety protection measures for the target server rack and adjacent server racks to ensure the safety of the target server rack and its adjacent server racks.

[0053] Based on the above embodiments, as an implementable approach, in one embodiment, determining the coolant diffusion information at the leak point based on the current monitoring image of the area to be tested includes:

[0054] Step 1031: Identify the liquid stains in the current monitoring image of the area to be tested, and obtain the liquid stain identification result of the current monitoring image;

[0055] Step 1032: Determine the coolant diffusion area at the leak point based on the liquid stain recognition results of the current monitoring image;

[0056] Step 1033: Determine the amount of coolant leaked at the leak point based on the coolant diffusion area at the leak point.

[0057] The information on coolant diffusion at the leak point includes at least the amount of coolant leaked at the leak point.

[0058] It should be noted that when the liquid cooling system of the server rack leaks, the coolant will form liquid stains inside the rack (around the pipes, on the equipment surface, or on the ground). These liquid stains are clearly different from other parts of the rack in the image, such as different colors and textures. Therefore, image recognition technology can be used to identify the liquid stains in the current monitoring image, and then determine the amount of coolant leaked at the leak point based on the coolant diffusion area represented by the liquid stain identification results.

[0059] In a fixed environment such as a server rack, where the surface is a flat metal plate without absorbent materials and the room temperature and humidity are stable, the coolant will not spread indefinitely after leakage. Instead, it will form a liquid film on the surface. The area of ​​the liquid film and the total amount of leaked liquid have a fixed corresponding relationship, which can be calibrated in advance through experiments.

[0060] Specifically, this application embodiment transforms visually observed leakage traces into quantifiable leakage data, providing a precise basis for subsequent safety decisions and laying the foundation for improving the accuracy of safety protection strategy determination results.

[0061] Accordingly, in one embodiment, if the current monitoring image includes multiple time-series images captured consecutively, the liquid stains in the multiple time-series images can be identified according to the capturing order of the multiple time-series images to obtain the liquid stain identification result of each time-series image; based on the liquid stain identification result of each time-series image, the temporal change information of the coolant diffusion area at the leak point is determined; based on the temporal change information of the coolant diffusion area at the leak point, the coolant diffusion rate at the leak point is determined; and based on the coolant diffusion area represented by the liquid stain identification result of the last time-series image among the multiple time-series images, the coolant leakage amount at the leak point is determined.

[0062] The coolant diffusion information at the leak point includes at least the coolant diffusion rate and the amount of coolant leaked. In scenarios where the current monitoring images are multiple time-series images captured continuously, the dynamic process of coolant diffusion is tracked through the time-dimensional image sequence, rather than relying on a single static image, thus achieving accurate quantification of the coolant diffusion rate and the amount of coolant leaked.

[0063] Specifically, the sequence of diffusion area changes over time can be determined based on the liquid stain area value obtained from each time-series image. Then, by calculating the area increase (leakage increment) and the image sampling interval, the trend of area change can be determined to obtain the coolant diffusion rate at the leak point.

[0064] Based on the above embodiments, as an implementable approach, if the pressure sensing data includes the current inlet and outlet pressures of multiple test points in the liquid cooling system and the historical inlet and outlet pressures of the previous moment, in one embodiment, the method further includes:

[0065] Step 201: For any point to be measured, determine the pressure drop rate of the point to be measured based on the current inlet and outlet pressures and the historical inlet and outlet pressures of the previous moment.

[0066] Step 202: If the pressure drop rate of the test point reaches the preset pressure drop rate threshold, the test point is designated as the leak point.

[0067] Step 203: Determine the degree of damage to the leak point based on the ratio between the pressure drop rate at the leak point and the preset upper limit of the pressure drop rate.

[0068] Specifically, high-precision pressure sensors are deployed in pairs at several key locations in the liquid cooling system (such as the cold plate inlet, cold plate outlet, and branch inlets and outlets of the liquid cooling pipes inside the cabinet) to obtain the current inlet and outlet pressures of the measured points and the historical inlet and outlet pressures at the previous moment. The sampling frequency of the pressure sensors can be set to above 10Hz, thereby achieving high sensitivity to capture transient pressure drops.

[0069] Specifically, an inlet pressure sensor and an outlet pressure sensor are installed at each measurement point. The system acquires the current inlet and outlet pressures at time t and compares them with the previous time t- By comparing the inlet and outlet pressures of t, the pressure drop rate at the test point is obtained, and then it is determined whether a leak has occurred at the test point.

[0070] The formula for calculating the pressure drop rate is as follows:

[0071]

[0072] in, Indicates the rate of voltage reduction. This indicates the current inlet pressure. This indicates current export pressure. This indicates the historical entry pressure at the previous moment. This indicates the historical export pressure at the previous moment. This indicates the inlet and outlet pressure collection cycle.

[0073] Furthermore, the preset pressure reduction rate upper limit is a normalized threshold, representing the maximum pressure change rate that can be withstood. Exceeding this value indicates a burst leak. Therefore, the degree of damage at the leak point can be determined based on the ratio of the pressure reduction rate to the preset pressure reduction rate upper limit.

[0074] Based on the above embodiments, as an implementable approach, in one embodiment, the leakage hazard level of the target cabinet is determined according to the degree of damage at the leak point characterized by pressure sensing data and the coolant diffusion information at the leak point, including:

[0075] Step 1041: Determine the leakage severity coefficient of the leak point based on the degree of damage at the leak point and the coolant diffusion rate and coolant leakage amount characterized by the coolant diffusion information.

[0076] Step 1042: Determine the leakage hazard level of the target cabinet based on the leakage severity coefficient of the leakage point.

[0077] For example, Table 1 below shows the leakage hazard level classification table for the target cabinet provided in the embodiments of this application:

[0078] Table 1. Leakage Hazard Classification of Target Cabinets

[0079]

[0080] Specifically, in one embodiment, the leakage severity coefficient of the leak point can be determined based on the following formula:

[0081]

[0082] in, This indicates the severity coefficient of the leak at the leak point. Indicates the extent of damage at the leak point. This indicates the rate of pressure drop at the leak point. This represents the difference between the current inlet / outlet pressure at the leak point and the historical inlet / outlet pressure at the previous moment. This indicates the inlet and outlet pressure sampling period at the leak point. This indicates the preset upper limit of the buck rate. Indicates the severity of coolant leakage. Indicates the amount of coolant leakage. This indicates the preset upper limit for coolant leakage. This indicates the severity of the coolant diffusion rate at the leak point. This indicates the rate at which the coolant diffuses at the leak point. Information on the temporal variation of the coolant diffusion area at the leak point. This indicates the image acquisition cycle of the current monitored image. This indicates the preset upper limit of coolant diffusion rate. , and This indicates the preset weighting coefficient. , and The cumulative total is 1. , express Coolant diffusion area at any given time express Coolant diffusion area at any given time express Time and The time difference between them is also the image acquisition cycle of the current monitoring image.

[0083] Among them, the maximum expected diffusion rate This is a preset normalization constant, representing the limit of diffusion rate that the system can identify and handle. When the diffusion rate exceeds this value, it indicates that the leakage is in a critical state of being out of control.

[0084] Specifically, in one embodiment, when there are multiple leakage points in the target cabinet, the average leakage severity coefficient of the multiple leakage points can be calculated to obtain the leakage severity coefficient of the target cabinet.

[0085] Accordingly, in one embodiment, the leakage severity coefficient of the leak point can also be determined solely based on the degree of damage at the leak point and the amount of coolant leakage characterized by coolant diffusion information. The specific calculation formula is as follows:

[0086]

[0087] At this time and The cumulative total is 1.

[0088] Based on the above embodiments, as an implementable approach, in one embodiment, a security protection strategy for the target cabinet and adjacent cabinets is determined according to the leakage hazard level of the target cabinet and the distance between the target cabinet and adjacent cabinets in the data center, including:

[0089] Step 1051: Based on the leakage hazard level of the target cabinet, determine the first safety protection sub-strategy and the liquid cooling system valve closure strategy for the target cabinet;

[0090] Step 1052: Based on the leakage hazard level of the target cabinet and the distance between the target cabinet and adjacent cabinets in the data center, determine the second security protection sub-strategy for each adjacent cabinet.

[0091] The safety protection strategy includes a first safety protection sub-strategy, a second safety protection sub-strategy, and a liquid cooling system valve closing strategy. The first and second safety protection sub-strategies include at least emergency power outage and frequency limiting operation.

[0092] Specifically, the distance between the target server rack and adjacent server racks in the data center refers to the actual physical distance between them. This distance can be obtained through methods such as 3D modeling of the data center.

[0093] Specifically, a dynamic risk assessment model can be created to determine the risk to nearby equipment after a leak occurs, and to determine the corresponding handling strategy.

[0094] The principle of the dynamic risk assessment model is as follows: distance is a key factor affecting risk; that is, the closer to the leak point, the higher the risk; the more severe the leak, the larger the affected area. The formula for assessing the distance risk value is:

[0095]

[0096] Where D represents the physical distance from the adjacent rack to the target rack (leak point). The unit is meters, which can be obtained and converted in real time through the rack positioning system; To maximize the effective impact distance, a value of 5 meters is recommended, but this can be adjusted based on the actual scenario. This application's embodiment specifies a distance risk value. The grading rules are shown in Table 2 below:

[0097] Table 2 Distance Risk Values Classification rule table

[0098]

[0099] Furthermore, by combining the leak diagnosis and grading results (leak severity coefficient) with the risk assessment results (distance risk value), a multi-level emergency action chain is generated, as shown in Table 3 below:

[0100] Table 3 Emergency Action Chain Rules

[0101]

[0102] Furthermore, the security protection strategies corresponding to different response levels are shown in Table 4 below:

[0103] Table 4 Security Protection Strategy Rules

[0104]

[0105] The security protection sub-strategy includes a first security protection sub-strategy and a second security protection sub-strategy. Specifically, the first security protection sub-strategy for the target cabinet is determined based on R < 0.1. The valve closing operation of the liquid cooling system in Table 4 depends on the level of control granularity of the liquid cooling system. If the granularity cannot reach the equipment or cabinet level, then the maximum supported range of liquid cooling valves will be closed. The target response time represents the time the management system should take to complete the operation; this time is a recommended value.

[0106] Specifically, based on the leakage classification results (leakage severity coefficient) and risk assessment conclusions (distance risk value), graded collaborative control can be initiated. On the liquid cooling side, for minor leaks, the branch valves in the leak area are closed and the water pump flow is adjusted. For severe leaks, the main branch valve in the leak area is closed and the circulating water pump is shut down to quickly prevent the leakage from spreading. On the GPU side, according to the risk level, gradient frequency reduction is performed on the server GPUs in the leak area, and emergency shutdown is triggered for high-risk devices. If there are running tasks, the platform is linked to migrate to a safe node. At the same time, the management platform sends alarm information of the corresponding level, including the location of the leak and the risk situation, to notify the operation and maintenance personnel to intervene.

[0107] Specifically, in one embodiment, the thermal conductivity of the material of each adjacent cabinet and the air velocity of the computer room can also be obtained; for any adjacent cabinet, the target response time limit of the second security protection sub-strategy of the adjacent cabinet is determined based on the thermal conductivity of the material of the adjacent cabinet and the air velocity of the computer room.

[0108] Among them, the target response time limit includes at least whether the peak-shifting power-off conditions are met, whether the power outage is executed immediately, and whether a short-term frequency-limiting operation is allowed before the power outage.

[0109] Specifically, to further improve the protection accuracy of adjacent cabinets, this application embodiment not only determines the second security protection sub-strategy based on the severity of the leakage and the distance between the cabinets, but also additionally acquires environmental parameters affecting the heat diffusion rate, including the thermal conductivity of the adjacent cabinet materials and the airflow velocity inside the data center, and calculates the target response time limit for the security protection strategy to be implemented for the adjacent cabinet. Among these measures, staggered power-off can prevent over-response during a leakage event from paralyzing the entire row of cabinets, thereby improving the service continuity of the intelligent computing center.

[0110] For example, such as Figure 2 The diagram illustrates the overall flow of the coolant leakage monitoring method provided in this embodiment. First, system preparation and parameter presets are performed, including pressure thresholds, image recognition model parameters, leakage classification rules, and risk assessment models. Subsequently, the system collects multimodal data in real time, including liquid cooling link pressure and camera monitoring images, and performs leakage diagnosis and classification processing, as well as dynamic risk assessment: the former determines the severity of the leak based on pressure drop rate, liquid stain area, and diffusion speed; the latter calculates the thermal diffusion risk by combining the leakage level and rack distance. Based on these two results, the system performs a two-factor decision, generating a graded security protection strategy for the target rack and adjacent racks. An automatic collaborative control module simultaneously drives the liquid cooling system control unit and the GPU cluster control unit to perform corresponding actions (such as valve closure, frequency limiting, or emergency power failure). After automatic control is completed, corresponding alarms are sent to maintenance personnel, achieving closed-loop linkage between liquid cooling system control and computing equipment protection, realizing rapid diagnosis, risk quantification, and cross-system collaborative protection of leakage events.

[0111] Specifically, in one embodiment, before determining the security protection strategy, a heat diffusion time series prediction model can be constructed based on historical operating data, cabinet material characteristics, airflow velocity, equipment power load, and other factors. This model is used to predict the possible temperature change trends of each cabinet within several future time windows, thereby enabling adaptive adjustment of the security protection strategy. The prediction model obtains the estimated temperature rise that a certain adjacent cabinet may reach in the future, and dynamically adjusts its second security protection sub-strategy based on the predicted value. For example, if the predicted temperature rise will exceed the equipment safety threshold within a preset temperature rise cycle (i.e., the temperature rise is rapid), the cabinet will be upgraded to a higher priority protection strategy (such as early power cut-off); if the predicted temperature rise is slow, the response time of the strategy execution can be extended to reduce unnecessary performance loss.

[0112] The coolant leakage monitoring method provided in this application acquires pressure sensing data of the liquid cooling system of a target cabinet; when the pressure sensing data indicates a leak in the liquid cooling system, it controls at least one image acquisition device to acquire the current monitoring image of the area to be monitored where the leak point is located; based on the current monitoring image of the area to be monitored, it determines the coolant diffusion information of the leak point; based on the degree of damage to the leak point indicated by the pressure sensing data and the coolant diffusion information of the leak point, it determines the leakage hazard level of the target cabinet; based on the leakage hazard level of the target cabinet and the distance between the target cabinet and adjacent cabinets in the computer room, it determines the safety protection strategy for the target cabinet and adjacent cabinets, so as to implement corresponding coolant leakage safety protection measures for the target cabinet and adjacent cabinets according to the safety protection strategy for the target cabinet and adjacent cabinets. The method described above determines the leakage hazard level of the target cabinet with a leak by combining pressure sensor data and current monitoring images. Then, based on the target cabinet's leakage hazard level and the distance between the target cabinet and adjacent cabinets in the data center, appropriate coolant leakage safety protection measures are implemented for the target cabinet and its neighboring cabinets to avoid safety risks caused by coolant leakage. Furthermore, through dual input-driven mechanisms of leak diagnosis results and dynamic risk assessment values, a multi-level response mechanism and closed-loop verification system are constructed, achieving a systematic leap in liquid cooling fault protection effectiveness. Compared to traditional single isolation strategies, this significantly improves the timeliness of emergency response in high-density computing scenarios, ensuring rapid isolation of the fault area while effectively suppressing the impact of heat diffusion on adjacent equipment. Based on the combination and matching of different leakage levels and risk thresholds, gradient protection actions are precisely triggered, avoiding computing power loss due to over-protection and preventing hardware damage caused by insufficient response. This provides highly reliable and adaptive leak safety protection for server clusters such as intelligent computing centers.

[0113] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method.

[0114] The embodiments of this application also provide a coolant leakage monitoring device for performing the coolant leakage monitoring method provided in the above embodiments.

[0115] like Figure 3 The diagram shown is a structural schematic of the coolant leakage monitoring device provided in an embodiment of this application. The coolant leakage monitoring device 30 includes: an acquisition module 301, an image acquisition module 302, a first determination module 303, a second determination module 304, and a protection module 305.

[0116] The system includes: an acquisition module for acquiring pressure sensing data of the liquid cooling system of the target cabinet; an image acquisition module for controlling at least one image acquisition device to acquire the current monitoring image of the area to be tested where the leak point is located when the pressure sensing data indicates a leak in the liquid cooling system; a first determination module for determining the coolant diffusion information of the leak point based on the current monitoring image of the area to be tested; a second determination module for determining the leakage hazard level of the target cabinet based on the degree of damage to the leak point indicated by the pressure sensing data and the coolant diffusion information of the leak point; and a protection module for determining the safety protection strategy for the target cabinet and adjacent cabinets based on the leakage hazard level of the target cabinet and the distance between the target cabinet and adjacent cabinets in the computer room, so as to implement corresponding coolant leakage safety protection measures for the target cabinet and adjacent cabinets according to the safety protection strategy for the target cabinet and adjacent cabinets.

[0117] For a description of the features in the embodiment corresponding to the coolant leakage monitoring device, please refer to the relevant description of the embodiment corresponding to the coolant leakage monitoring method, which will not be repeated here.

[0118] The embodiments of this application also provide a coolant leakage monitoring system for performing the coolant leakage monitoring method provided in the above embodiments.

[0119] like Figure 4 The diagram shown is a structural schematic of a coolant leakage monitoring system provided in an embodiment of this application. The coolant leakage monitoring system includes a control component, a pressure sensor, and an image acquisition device.

[0120] The pressure sensors are deployed at the inlet and outlet positions of multiple test points in the liquid cooling system of the target cabinet; the pressure sensors are used to acquire pressure sensing data of the liquid cooling system of the target cabinet; the image acquisition device is used to acquire monitoring images of the area where each test point is located; the control component uses the coolant leakage monitoring method provided in the above embodiment to monitor coolant leakage in the target cabinet.

[0121] The control components include a central control unit, a liquid-cooled actuator, and a GPU management interface. The central control unit is used to execute the method provided in the above embodiments to determine the security protection strategy for the target cabinet and adjacent cabinets. The liquid-cooled actuator is used to control the valves / pumps of the liquid cooling system of the target cabinet. The GPU management interface is used to limit the frequency of the cabinet. The central control unit includes a device management interface, which is used to shut down the cabinet (emergency power failure).

[0122] For example, such as Figure 5The diagram shows an exemplary coolant leakage monitoring system provided in this application embodiment. This system comprises a data center management platform software, a central control unit, core functional modules, a sensing and detection layer, and an execution and linkage layer. The data center management platform provides the central control unit with basic information such as rack location, equipment status, and environmental parameters through 3D server room modeling and cluster management functions. The central control unit integrates three core functions: a precise leakage diagnosis and grading module, a dynamic risk assessment module, and a tiered collaborative control module. These modules are used to determine the severity of leakage based on pressure changes and image recognition data, quantify risk based on rack distance and environmental characteristics, and generate safety protection strategies for the target rack and adjacent racks, respectively. In the sensing and detection layer, pressure sensors collect real-time pressure data at various test points in the liquid cooling system, and near-infrared optical cameras acquire liquid stain images to identify the leakage area and diffusion rate. The execution and linkage layer includes liquid cooling actuators (valves / pumps), a GPU management interface (frequency limiting), and a device management interface (power-down), used to automatically execute corresponding emergency actions based on control decisions. This system achieves real-time monitoring, risk prediction, and cross-system linkage control of liquid cooling leakage events through the collaborative work of the above modules, ensuring the safe operation of the intelligent computing center equipment.

[0123] For a description of the features in the embodiment corresponding to the coolant leakage monitoring system, please refer to the relevant description of the embodiment corresponding to the coolant leakage monitoring method, which will not be repeated here.

[0124] Embodiments of this application also provide an electronic device, such as... Figure 6 The diagram shown is a schematic diagram of the structure of an electronic device provided in an embodiment of this application, including a processor 10 and a memory 20. The memory 20 stores a computer program, and the processor 10 is configured to run the computer program to execute the steps in any of the above-described embodiments of the coolant leakage monitoring method.

[0125] Embodiments of this application also provide a computer-readable storage medium storing a computer program, wherein the computer program is configured to execute the steps in any of the above embodiments of the coolant leakage monitoring method when it is run.

[0126] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard disk, magnetic disk, or optical disk.

[0127] Embodiments of this application also provide a computer program product, which includes a computer program that, when executed by a processor, implements the steps in any of the above embodiments of the coolant leakage monitoring method.

[0128] Embodiments of this application also provide another computer program product, including a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in any of the above embodiments of the coolant leakage monitoring method.

[0129] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0130] The above provides a detailed description of a coolant leakage monitoring method, system, and electronic device provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are only intended to help understand the method and its core ideas. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the protection scope of the claims of this application.

Claims

1. A method for monitoring coolant leakage, characterized in that, include: Acquire pressure sensing data of the liquid cooling system of the target cabinet; When the pressure sensing data indicates that there is a leak in the liquid cooling system, at least one image acquisition device is controlled to acquire the current monitoring image of the area to be tested where the leak is located. Based on the current monitoring image of the area to be tested, determine the coolant diffusion information at the leak point; Based on the degree of damage at the leak point as characterized by the pressure sensing data and the coolant diffusion information at the leak point, the leakage hazard level of the target cabinet is determined; Based on the leakage hazard level of the target cabinet and the distance between the target cabinet and adjacent cabinets in the computer room, a safety protection strategy for the target cabinet and the adjacent cabinets is determined, and corresponding coolant leakage safety protection measures are implemented for the target cabinet and the adjacent cabinets in accordance with the safety protection strategy for the target cabinet and the adjacent cabinets. The step of determining the leakage hazard level of the target cabinet based on the degree of damage at the leak point characterized by the pressure sensing data and the coolant diffusion information at the leak point includes: The leakage severity coefficient of the leak point is determined based on the degree of damage at the leak point and the coolant diffusion rate and leakage amount characterized by the coolant diffusion information. The leakage hazard level of the target cabinet is determined based on the leakage severity coefficient of the leakage point. The determination of the leakage severity coefficient at the leak point based on the degree of damage at the leak point and the coolant diffusion rate and leakage amount characterized by the coolant diffusion information includes: The leakage severity coefficient of the leak point is determined based on the following formula: in, This indicates the severity coefficient of the leakage at the leak point. This indicates the degree of damage at the leak point. This indicates the rate of pressure drop at the leak point. This represents the difference between the current inlet / outlet pressure at the leak point and the historical inlet / outlet pressure at the previous moment. This indicates the inlet and outlet pressure sampling period for the leak point. This indicates the preset upper limit of the buck rate. Indicates the severity of coolant leakage. Indicates the amount of coolant leakage. This indicates the preset upper limit for coolant leakage. This indicates the severity of the coolant diffusion rate at the leak point. This indicates the coolant diffusion rate at the leak point. Information indicating the temporal variation of the coolant diffusion area at the leak point. This indicates the image acquisition period of the currently monitored image. This indicates the preset upper limit of coolant diffusion rate. , and This indicates the preset weighting coefficient.

2. The coolant leakage monitoring method according to claim 1, characterized in that, The step of determining the coolant diffusion information at the leak point based on the current monitoring image of the area to be tested includes: The liquid stains in the current monitoring image of the area to be tested are identified to obtain the liquid stain identification result of the current monitoring image; Based on the liquid stain recognition results of the current monitoring image, determine the coolant diffusion area of ​​the leak point; The amount of coolant leaked at the leak point is determined based on the coolant diffusion area at the leak point. The coolant diffusion information at the leak point includes at least the amount of coolant leaked at the leak point.

3. The coolant leakage monitoring method according to claim 1, characterized in that, The current monitoring image includes multiple time-series images captured consecutively. Determining the coolant diffusion information at the leak point based on the current monitoring image of the area to be tested includes: According to the shooting order of the multiple time-series images, the liquid stains in the multiple time-series images are identified to obtain the liquid stain identification result for each time-series image; Based on the liquid stain identification results of each time-series image, determine the time-series change information of the coolant diffusion area at the leak point; Based on the temporal change information of the coolant diffusion area at the leak point, the coolant diffusion rate at the leak point is determined; The amount of coolant leakage at the leak point is determined by the coolant diffusion area characterized by the liquid stain identification result of the last time series image among the multiple time series images. The coolant diffusion information at the leak point includes at least the coolant diffusion rate and the amount of coolant leaked at the leak point.

4. The coolant leakage monitoring method according to claim 1, characterized in that, The pressure sensing data includes the current inlet and outlet pressures of multiple test points in the liquid cooling system and the historical inlet and outlet pressures of the previous moment. The method further includes: For any of the measured points, the pressure drop rate of the measured point is determined based on the current inlet / outlet pressure and the historical inlet / outlet pressure at the previous moment. If the pressure drop rate at the test point reaches a preset pressure drop rate threshold, the test point is considered a leak point. The degree of damage to the leak point is determined by calculating the ratio between the pressure drop rate at the leak point and the preset upper limit of the pressure drop rate.

5. The coolant leakage monitoring method according to claim 1, characterized in that, The step of determining the security protection strategy for the target cabinet and the adjacent cabinets based on the leakage hazard level of the target cabinet and the distance between the target cabinet and adjacent cabinets in the data center includes: Based on the leakage hazard level of the target cabinet, determine the first safety protection sub-strategy and the liquid cooling system valve closure strategy for the target cabinet; Based on the leakage hazard level of the target cabinet and the distance between the target cabinet and adjacent cabinets in the computer room, a second security protection sub-strategy is determined for each of the adjacent cabinets; The safety protection strategy includes a first safety protection sub-strategy, a second safety protection sub-strategy, and a liquid cooling system valve closing strategy. The first and second safety protection sub-strategies include at least emergency power outage and frequency limiting operation.

6. The coolant leakage monitoring method according to claim 5, characterized in that, The method further includes: The thermal conductivity of the materials of each of the adjacent cabinets and the air velocity in the computer room are obtained; For any of the adjacent cabinets, the target response time limit of the second security protection sub-strategy for the adjacent cabinets is determined based on the thermal conductivity of the material of the adjacent cabinets and the air velocity of the computer room. The target response time limit includes at least whether the peak power-off conditions are met.

7. A coolant leakage monitoring system, characterized in that, include: The system includes control components, pressure sensors, and image acquisition equipment. The pressure sensors are deployed at the inlet and outlet positions of multiple test points in the liquid cooling system of the target cabinet. The pressure sensor is used to acquire pressure sensing data of the liquid cooling system of the target cabinet; The image acquisition device is used to acquire monitoring images of the area where each of the test points is located. The control component employs the coolant leakage monitoring method as described in any one of claims 1 to 6 to monitor coolant leakage in the target cabinet.

8. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor, configured to implement the steps of the coolant leakage monitoring method as described in any one of claims 1 to 6 when executing the computer program.

Images