A server immersion liquid cooling control system and method

Abstract

The invention relates to the field of liquid cooling control technology, and specifically discloses a server immersion liquid cooling control system and method, wherein the system comprises: a temperature control standard analysis module determines reference temperature limits of all reference monitoring points on the inner wall of a server cabinet; a temperature model building module builds an immersion liquid cooling dynamic temperature model of the server cabinet; a temperature control equation analysis module determines a heat dissipation temperature control equation for each reference monitoring point based on the immersion liquid cooling dynamic temperature model of the server cabinet; an optimal temperature determination module determines the optimal temperature of the coolant at the inlet of the liquid cooling cycle in real time based on the real-time temperatures of all monitoring points located on the inner wall of the server cabinet and the reference temperature limits and heat dissipation temperature control equations of all reference monitoring points on the inner wall of the server cabinet; and realizes precise control of the server immersion liquid cooling system and real-time monitoring and dynamic heat dissipation of the server cabinet temperature, thereby ensuring that the server can operate stably at an appropriate temperature.

Description

Technical Field

[0001] The present invention relates to the technical field of liquid cooling control, and in particular to a server immersion liquid cooling control system and method. Background Art

[0002] At present, distributed data centers are becoming increasingly important due to the demand for local data storage and computing and the requirement to reduce data transmission delays. Compared with traditional large data centers, distributed data centers can make more effective use of existing resources and can be quickly deployed to meet temporary needs. The heat dissipation technology of the server cabinets that carry distributed data centers is a major research difficulty for data servers. Traditional data center server cabinets use air cooling systems to achieve effective heat dissipation, and air cooling systems are indirect heat dissipation, which has problems such as low heat exchange efficiency, high noise, and dust in the air that reduces the life of the server. Therefore, for distributed data centers, it is necessary to find more efficient and environmentally friendly cooling methods. In recent years, liquid cooling systems have begun to be widely used in some fields, such as cloud computing and artificial intelligence. Liquid cooling systems are more efficient than traditional air cooling systems because they can directly transfer heat to water and then discharge heat through water pipes, thus avoiding energy loss during air circulation. At the same time, liquid cooling systems are also quieter than air cooling systems because they do not require fans or other mechanical equipment to cool the server.

[0003] However, since data servers contain many components that generate concentrated heat, the liquid cooling circulation system used in data servers must have a high-precision circulating liquid temperature control strategy to ensure that many components in the data server can stably operate within the safe temperature limit to meet the server's heat dissipation requirements. However, the existing circulating liquid temperature control strategy in the liquid cooling circulation system used in the server is relatively simple, and it is difficult to cope with the irregular heating of components caused by the data server when processing data loads of different magnitudes, making the heat dissipation effect of the server immersion liquid cooling control technology poor and difficult to meet the actual heat dissipation needs of the server.

[0004] Therefore, the present invention proposes a server immersion liquid cooling control system and method. Summary of the invention

[0005] The present invention provides a server immersion liquid cooling control system and method, which is used to analyze the heat dissipation temperature control equation of each reference monitoring point of each component to be monitored through real-time monitoring and joint analysis of the temperatures of multiple monitoring points on the inner wall of the server cabinet and the temperatures of the liquid cooling cycle inlet and outlet, and accurately control the optimal temperature of the coolant at the liquid cooling cycle inlet according to the heat dissipation temperature control equation, thereby realizing dynamic and precise control of the server immersion liquid cooling system, coping with the irregular heating of components caused by the data server when processing data loads of different magnitudes, and ensuring that the server can run stably at an appropriate temperature.

[0006] The present invention provides a server immersion liquid cooling control system, which is characterized by comprising:

[0007] A temperature control standard analysis module is used to determine reference temperature limits of all reference monitoring points on the inner wall of the server cabinet based on the heat dissipation standard of the server cabinet;

[0008] A temperature model building module is used to build an immersion liquid cooling dynamic temperature model of the server cabinet based on the real-time temperature of multiple monitoring points on the inner wall of the server cabinet, the real-time temperature of the coolant at the inlet of the liquid cooling cycle, and the real-time temperature of the cooling liquid at the outlet of the liquid cooling cycle;

[0009] A temperature control equation analysis module is used to determine the heat dissipation temperature control equation for each reference monitoring point based on the immersion liquid cooling dynamic temperature model of the server cabinet;

[0010] The optimal temperature determination module is used to determine the optimal temperature of the coolant at the inlet of the liquid cooling cycle in real time based on the real-time temperature of all monitoring points located on the inner wall of the server cabinet and the reference temperature limit of all reference monitoring points on the inner wall of the server cabinet and the heat dissipation temperature control equation.

[0011] Preferably, the temperature control standard analysis module includes:

[0012] The temperature limit reading submodule is used to determine the temperature limit values ​​in the temperature control standards of all components to be monitored in the server cabinet in the heat dissipation standards of the server cabinet;

[0013] A reference point determination submodule, used to determine a reference monitoring point of each component to be monitored among all monitoring points on the inner wall of the server cabinet based on the distribution position of each component to be monitored in the server cabinet;

[0014] The thermal conductivity conversion submodule is used to perform thermal conductivity conversion on the temperature limit of each component to be monitored, and determine the reference temperature limit of the reference monitoring point of each component to be monitored.

[0015] Preferably, the reference point determination submodule includes:

[0016] A projection spacing determination unit is used to determine the projection area of ​​the three-dimensional shapes of all components to be monitored on each shape surface of the server cabinet, and determine the spacing between two adjacent physical center positions of the projection area of ​​the three-dimensional shapes of all components to be monitored on each shape surface of the server cabinet as the projection spacing of the corresponding two adjacent components to be monitored on the corresponding shape surface of the server cabinet;

[0017] A spacing distance determination unit, used to determine the most suitable spacing distance of each shape surface of the server cabinet based on all projection distances of two adjacent components to be monitored on each shape surface of the server cabinet;

[0018] A monitoring point position determination unit, configured to determine the positions of all monitoring points on the inner wall of the server cabinet based on the optimum spacing distance of each shape surface of the server cabinet, wherein the positions of all monitoring points on each shape surface of the inner wall of the server cabinet are distributed with the optimum spacing distance of the corresponding shape surface as the spacing distance between two adjacent monitoring points;

[0019] The reference point determination unit is used to determine the reference monitoring point of each component to be monitored among all the monitoring points on the inner wall of the server cabinet based on the positions of all the monitoring points on the inner wall of the server cabinet and the distribution positions of the three-dimensional shapes of all the components to be monitored in the server cabinet.

[0020] Preferably, the interval distance determining unit comprises:

[0021] An exclusive setting subunit, used for treating all projection spacings of the two adjacent components to be monitored on each shape surface of the server cabinet except the current projection spacing as exclusive projection spacings of the current projection spacing when the total number of all components to be monitored is greater than 3;

[0022] A compatibility calculation subunit is used to calculate the similarity between each projection interval and the average of all corresponding exclusive projection intervals as the compatibility of each projection interval;

[0023] The first analysis subunit is used to determine whether the minimum compatibility among the compatibilities of all projection spacings of the two adjacent components to be monitored on each shape surface of the server cabinet is less than the compatibility threshold. If so, the average of all exclusive projection spacings of the projection spacing corresponding to the minimum compatibility is used as the optimum spacing distance of the corresponding shape surface. Otherwise, the spacing of all projection spacings of the two adjacent components to be monitored on the corresponding shape surface of the server cabinet is used as the optimum spacing distance of the corresponding shape surface.

[0024] The second analysis subunit is used to regard the distance between all projections of the two adjacent components to be monitored on the corresponding shape surface of the server cabinet as the most suitable spacing distance of the corresponding shape surface when the total number of all components to be monitored is not greater than 3.

[0025] Preferably, the reference point determination unit comprises:

[0026] A first selection subunit is used to determine, based on the positions of all monitoring points on the inner wall of the server cabinet and the distribution positions of the three-dimensional shapes of all components to be monitored in the server cabinet, a monitoring point closest to each component to be monitored among all monitoring points on the inner wall of the server cabinet as a first selected monitoring point for each component to be monitored;

[0027] A second selection subunit is used for, when the first selected monitoring points of all components to be monitored overlap, determining the second largest value between the distance between each component to be monitored of at least two components to be monitored that overlap with the corresponding first selected monitoring points and the distance between all monitoring points as the first distance, and treating the corresponding overlapping monitoring points as the reference monitoring points of the component to be monitored with the largest first distance among the at least two components to be monitored that overlap with the corresponding first selected monitoring points, and treating the first selected monitoring points of all components to be monitored that do not overlap with the corresponding first selected monitoring points as the reference monitoring points of the corresponding components to be monitored, and determining the reference monitoring points of each component to be monitored that has not been selected as the reference monitoring point among all monitoring points except the monitoring points currently selected as the reference monitoring points;

[0028] The reference point determination subunit is used to use the first selected monitoring point of each component to be monitored as the reference monitoring point of each component to be monitored when the first selected monitoring points of all components to be monitored do not overlap.

[0029] Preferably, the thermal conductivity conversion submodule includes:

[0030] A thermal conductivity parameter determination unit, used to determine all thermal conductivity media and the conduction path of each thermal conductivity medium between each component to be monitored and the corresponding reference monitoring point;

[0031] The thermal conductivity conversion unit is used to perform thermal conductivity conversion on the temperature limit of each component to be monitored based on all thermal conductivity media between each component to be monitored and the corresponding reference monitoring point and the conduction path of each thermal conductivity medium, so as to obtain the reference temperature limit of the reference monitoring point of each component to be monitored.

[0032] Preferably, the temperature model building module includes:

[0033] The data alignment submodule is used to perform time-series alignment on all the real-time temperatures of multiple monitoring points on the inner wall of the server cabinet, all the real-time temperatures of the coolant at the inlet of the liquid cooling cycle, and all the real-time temperatures of the cooling liquid at the outlet of the liquid cooling cycle, so as to obtain the real-time temperatures of multiple monitoring points on the inner wall of the server cabinet, the real-time temperatures of the coolant at the inlet of the liquid cooling cycle, and the real-time temperatures of the cooling liquid at the outlet of the liquid cooling cycle at each moment;

[0034] The temperature model generation submodule is used to mark the real-time temperatures of multiple monitoring points on the inner wall of the server cabinet at each moment, the real-time temperature of the coolant at the inlet of the liquid cooling cycle, and the real-time temperature of the cooling liquid at the outlet of the liquid cooling cycle, respectively, into the three-dimensional model of the immersion liquid cooling system of the server cabinet, and obtain the immersion liquid cooling temperature model of the server cabinet at each moment;

[0035] The dynamic model generation submodule is used to sort and connect the immersion liquid cooling temperature models of the server cabinet at all times in time sequence to obtain the immersion liquid cooling dynamic temperature model of the server cabinet.

[0036] Preferably, the temperature control equation analysis module includes:

[0037] Model building submodule, used to build a heat dissipation temperature control equation extraction model;

[0038] The equation output submodule is used to input the immersion liquid cooling dynamic temperature model of the server cabinet with all reference monitoring point locations marked into the heat dissipation temperature control equation extraction model to obtain the heat dissipation temperature control equation for each reference monitoring point.

[0039] Preferably, the optimal temperature determination module includes:

[0040] A real-time temperature determination submodule, used to determine the real-time temperatures of all reference monitoring points on the inner wall of the server cabinet based on the real-time temperatures of all monitoring points located on the inner wall of the server cabinet;

[0041] The first optimal temperature determination submodule is used to input the real-time temperatures of all monitoring points on the inner wall of the server cabinet, the reference temperature limits of all reference monitoring points on the inner wall of the server cabinet, and the heat dissipation temperature control equations of all reference monitoring points into a preset temperature rise temperature control model in real time when the real-time temperatures of all reference monitoring points do not exceed the reference temperature limits of the corresponding reference monitoring points, so as to obtain the optimal temperature of the coolant at the inlet of the liquid cooling cycle;

[0042] The second optimal temperature determination submodule is used to input the real-time temperatures of all monitoring points on the inner wall of the server cabinet, the reference temperature limits of all reference monitoring points on the inner wall of the server cabinet, and the heat dissipation temperature control equations of all reference monitoring points into the preset cooling temperature control model in real time when the real-time temperature of at least one reference monitoring point exceeds the reference temperature limit of the corresponding reference monitoring point, so as to obtain the optimal temperature of the coolant at the inlet of the liquid cooling cycle.

[0043] The present invention provides a server immersion liquid cooling control method, which is applied to the server immersion liquid cooling control system described in any one of Embodiments 1 to 9, comprising:

[0044] S1: Determine the reference temperature limit values ​​of all reference monitoring points on the inner wall of the server cabinet based on the heat dissipation standard of the server cabinet;

[0045] S2: Based on the real-time temperature of multiple monitoring points on the inner wall of the server cabinet, the real-time temperature of the coolant at the inlet of the liquid cooling cycle, and the real-time temperature of the coolant at the outlet of the liquid cooling cycle, an immersion liquid cooling dynamic temperature model of the server cabinet is constructed;

[0046] S3: Determine the heat dissipation temperature control equation for each reference monitoring point based on the immersion liquid cooling dynamic temperature model of the server cabinet;

[0047] S4: Based on the real-time temperatures of all monitoring points located on the inner wall of the server cabinet, the reference temperature limits of all reference monitoring points on the inner wall of the server cabinet, and the heat dissipation temperature control equation, the optimal temperature of the coolant at the inlet of the liquid cooling cycle is determined in real time.

[0048] The beneficial effects of the present invention compared to the prior art are as follows: through real-time monitoring and joint analysis of the temperatures of multiple monitoring points on the inner wall of the server cabinet and the temperatures of the liquid cooling cycle inlet and outlet, the heat dissipation temperature control equation of the reference monitoring point of each component to be monitored is analyzed, and the optimal temperature of the coolant at the liquid cooling cycle inlet is accurately controlled according to the heat dissipation temperature control equation, thereby achieving dynamic and precise control of the server immersion liquid cooling system, and being able to cope with the irregular heating of components caused by the data server when processing data loads of different magnitudes, ensuring that the server can operate stably at an appropriate temperature.

[0049] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures specifically pointed out in this application document.

[0050] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0052] Figure 1 It is a schematic diagram of the internal functional modules of the server immersion liquid cooling control system in an embodiment of the present invention;

[0053] Figure 2 Schematic diagram of internal functional submodules of the temperature control standard analysis module in an embodiment of the present invention;

[0054] Figure 3A schematic diagram of internal functional submodules of a temperature model building module in an embodiment of the present invention;

[0055] Figure 4 Schematic diagram of internal functional submodules of the temperature control equation analysis module in an embodiment of the present invention;

[0056] Figure 5 Schematic diagram of internal functional submodules of the optimal temperature determination module in an embodiment of the present invention;

[0057] Figure 6 This is a flow chart of a server immersion liquid cooling control method in an embodiment of the present invention. DETAILED DESCRIPTION

[0058] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0059] Embodiment 1:

[0060] The present invention provides a server immersion liquid cooling control system, referring to Figure 1 ,include:

[0061] A temperature control standard analysis module is used to determine reference temperature limits of all reference monitoring points on the inner wall of the server cabinet based on the heat dissipation standard of the server cabinet;

[0062] A temperature model building module is used to build an immersion liquid cooling dynamic temperature model of the server cabinet based on the real-time temperature of multiple monitoring points on the inner wall of the server cabinet, the real-time temperature of the coolant at the inlet of the liquid cooling cycle, and the real-time temperature of the cooling liquid at the outlet of the liquid cooling cycle;

[0063] A temperature control equation analysis module is used to determine the heat dissipation temperature control equation for each reference monitoring point based on the immersion liquid cooling dynamic temperature model of the server cabinet;

[0064] The optimal temperature determination module is used to determine the optimal temperature of the coolant at the inlet of the liquid cooling cycle in real time based on the real-time temperature of all monitoring points located on the inner wall of the server cabinet and the reference temperature limit of all reference monitoring points on the inner wall of the server cabinet and the heat dissipation temperature control equation.

[0065] In this embodiment, the server cabinet is a device specially used for storing and storing servers. It is usually made of metal or non-metal materials, and its shape and size can be customized according to different server sizes and requirements. The main function of the server cabinet is to provide a stable, safe and reliable environment for the server to ensure the normal operation of the server and the security of data.

[0066] In addition to providing physical support for the server, the server cabinet also has some other functions. For example, it can easily install and manage various hardware devices and software tools, such as power adapters, fan controllers, temperature sensors, etc.; it can effectively manage and organize the cables and connectors inside the server to avoid a messy state; it can protect the server from external factors, such as vibration, humidity changes, temperature fluctuations, etc.

[0067] In general, the server cabinet is a very important component that can provide a stable, safe and reliable working environment for the server, thereby ensuring the accuracy and security of the data.

[0068] In this embodiment, the heat dissipation standard is preset information including the maximum temperature allowed for all components to be monitored in the server cabinet when they are working safely.

[0069] In this embodiment, the reference monitoring point is used to replace the monitoring point of the corresponding component to be monitored for real-time temperature monitoring, and the real-time monitoring of the heat dissipation state of the component to be monitored can be achieved through thermal conductivity conversion.

[0070] In this embodiment, the reference temperature limit of the reference monitoring point is obtained by converting the temperature limit of the component to be monitored by thermal conductivity, and indicates that the temperature value that the corresponding reference monitoring point will reach when the component to be monitored is at the corresponding temperature limit.

[0071] In this embodiment, the real-time temperature of multiple monitoring points on the inner wall of the server cabinet, the real-time temperature of the coolant at the liquid cooling cycle inlet, and the real-time temperature of the cooling liquid at the liquid cooling cycle outlet are obtained in real time through temperature sensors respectively arranged at each monitoring point, the liquid cooling cycle inlet, and the liquid cooling cycle outlet.

[0072] In this embodiment, the heat dissipation temperature control equation of the reference monitoring point is an equation that represents the numerical relationship between the real-time temperature of the coolant at the inlet of the liquid cooling cycle and the real-time temperature of other monitoring points in the server cabinet and the real-time temperature of the reference monitoring point.

[0073] In this embodiment, the optimal temperature of the coolant at the inlet of the liquid cooling cycle is a value that the coolant at the inlet of the liquid cooling cycle needs to be equal to at the corresponding moment when all the components to be monitored in the server cabinet can operate stably below the temperature limit. If the real-time temperature of the coolant at the inlet of the liquid cooling cycle at this moment is actually greater than the optimal temperature of the coolant at the inlet of the liquid cooling cycle, the heat dissipation demand of the server cabinet at this moment will not be met. If the real-time temperature of the coolant at the inlet of the liquid cooling cycle at this moment is actually less than the optimal temperature of the coolant at the inlet of the liquid cooling cycle, more energy consumption will be required to provide a lower temperature for the coolant, thereby increasing energy consumption.

[0074] The beneficial effects of the above technology are: through real-time monitoring and joint analysis of the temperatures of multiple monitoring points on the inner wall of the server cabinet and the temperatures of the liquid cooling cycle inlet and outlet, the heat dissipation temperature control equation of the reference monitoring point of each component to be monitored is analyzed, and the optimal temperature of the coolant at the liquid cooling cycle inlet is accurately controlled according to the heat dissipation temperature control equation, thereby realizing dynamic and precise control of the server immersion liquid cooling system, and being able to cope with the irregular heating of components caused by the data server when processing data loads of different magnitudes, ensuring that the server can operate stably at an appropriate temperature.

[0075] Embodiment 2:

[0076] Based on Example 1, the temperature control standard analysis module, refer to Figure 2 ,include:

[0077] The temperature limit reading submodule is used to determine the temperature limit values ​​in the temperature control standards of all components to be monitored in the server cabinet in the heat dissipation standards of the server cabinet;

[0078] A reference point determination submodule, used to determine a reference monitoring point of each component to be monitored among all monitoring points on the inner wall of the server cabinet based on the distribution position of each component to be monitored in the server cabinet;

[0079] The thermal conductivity conversion submodule is used to perform thermal conductivity conversion on the temperature limit of each component to be monitored, and determine the reference temperature limit of the reference monitoring point of each component to be monitored.

[0080] In this embodiment, the components to be monitored are components installed in the server cabinet whose temperature status needs to be monitored to avoid local overheating that may cause damage or failure of the server cabinet, such as processors, motherboards, memory sticks, hard disk drives, etc.

[0081] In this embodiment, the temperature control standard of the component to be monitored includes information on the maximum temperature allowed when the component to be monitored is working safely.

[0082] In this embodiment, the distribution positions of the components to be monitored in the server cabinet are represented by a large number of three-dimensional coordinates obtained by representing all position points on the surface of the components to be monitored using three-dimensional coordinates in a preset three-dimensional coordinate system.

[0083] The beneficial effects of the above technology are: by interpreting the heat dissipation standards of the server cabinet, the temperature limits in the temperature control standards of all components to be monitored in the server cabinet are obtained, and based on the distribution positions of the components to be monitored in the server cabinet, the reference monitoring points of the components to be monitored are reasonably determined, and the temperature limit of each component to be monitored is converted by thermal conductivity to obtain the reference temperature limit of the reference monitoring point of the component to be monitored, which is convenient for indirect monitoring and reference of the real-time temperature and heat dissipation status of the components to be monitored.

[0084] Embodiment 3:

[0085] Based on Example 2, the reference point determination submodule includes:

[0086] A projection spacing determination unit is used to determine the projection area of ​​the three-dimensional shapes of all components to be monitored on each shape surface of the server cabinet, and determine the spacing between two adjacent physical center positions of the projection area of ​​the three-dimensional shapes of all components to be monitored on each shape surface of the server cabinet as the projection spacing of the corresponding two adjacent components to be monitored on the corresponding shape surface of the server cabinet;

[0087] A spacing distance determination unit, used to determine the most suitable spacing distance of each shape surface of the server cabinet based on all projection distances of two adjacent components to be monitored on each shape surface of the server cabinet;

[0088] A monitoring point position determination unit, configured to determine the positions of all monitoring points on the inner wall of the server cabinet based on the optimum spacing distance of each shape surface of the server cabinet, wherein the positions of all monitoring points on each shape surface of the inner wall of the server cabinet are distributed with the optimum spacing distance of the corresponding shape surface as the spacing distance between two adjacent monitoring points;

[0089] The reference point determination unit is used to determine the reference monitoring point of each component to be monitored among all the monitoring points on the inner wall of the server cabinet based on the positions of all the monitoring points on the inner wall of the server cabinet and the distribution positions of the three-dimensional shapes of all the components to be monitored in the server cabinet.

[0090] In this embodiment, the projection area of ​​the three-dimensional shape of the component to be monitored on each shape surface of the server cabinet is the projection area formed by the three-dimensional shape of the component to be monitored on the plane corresponding to each shape surface of the server cabinet.

[0091] In this embodiment, each shape surface of the server cabinet is a complete plane contained in the server cabinet. For example, when the server cabinet is a rectangular parallelepiped, it includes six shape surfaces of rectangular or square shapes.

[0092] In this embodiment, the optimum spacing distance of each shape surface of the server cabinet is the optimal spacing between adjacent monitoring points in the evenly distributed monitoring points in each shape surface of the server cabinet, determined under the premise of taking into account the spacing density between the distribution positions of the components to be monitored in the server cabinet. If the spacing distance is too large, it will increase the error in the process of referencing or indirectly monitoring the temperature or heat dissipation status of the components to be monitored by monitoring the temperature of the referenced monitoring points, or it will cause the referenced monitoring points of the components to be monitored to overlap. If the spacing distance is too small, it will increase the setting cost of the temperature sensor and the subsequent processing cost of the temperature data. Therefore, it is necessary to determine the optimum spacing distance.

[0093] In this embodiment, all monitoring points of each shape surface exist in a uniformly distributed state.

[0094] In this embodiment, the position of the monitoring point is represented by a three-dimensional coordinate value in a preset three-dimensional coordinate system.

[0095] The beneficial effects of the above technology are as follows: by analyzing the intervals between the adjacent physical centers of the projection areas of each shape surface of all components to be monitored on the server cabinet, the most suitable interval distance of each shape surface of the server cabinet is determined, avoiding the occurrence of two situations: increasing the error in the process of referencing or indirectly monitoring the temperature or heat dissipation status of the components to be monitored by monitoring the temperature of the referenced monitoring points, or causing the referenced monitoring points of the components to be monitored to overlap. It also reasonably controls the setting cost of the temperature sensor and the subsequent processing cost of the temperature data, and based on the positions of all monitoring points on the inner wall of the server cabinet determined in this way, the referenced monitoring points of the components to be monitored are determined, so that the temperature or heat dissipation status of the components to be monitored can be indirectly monitored or referenced through the referenced monitoring points in the future.

[0096] Embodiment 4:

[0097] On the basis of Embodiment 3, the interval distance determining unit includes:

[0098] An exclusive setting subunit, used for treating all projection spacings of the two adjacent components to be monitored on each shape surface of the server cabinet except the current projection spacing as exclusive projection spacings of the current projection spacing when the total number of all components to be monitored is greater than 3;

[0099] A compatibility calculation subunit is used to calculate the similarity between each projection interval and the average of all corresponding exclusive projection intervals as the compatibility of each projection interval;

[0100] The first analysis subunit is used to determine whether the minimum compatibility among the compatibilities of all projection spacings of the two adjacent components to be monitored on each shape surface of the server cabinet is less than the compatibility threshold. If so, the average of all exclusive projection spacings of the projection spacing corresponding to the minimum compatibility is used as the optimum spacing distance of the corresponding shape surface. Otherwise, the spacing of all projection spacings of the two adjacent components to be monitored on the corresponding shape surface of the server cabinet is used as the optimum spacing distance of the corresponding shape surface.

[0101] The second analysis subunit is used to regard the distance between all projections of the two adjacent components to be monitored on the corresponding shape surface of the server cabinet as the most suitable spacing distance of the corresponding shape surface when the total number of all components to be monitored is not greater than 3.

[0102] In this embodiment, calculating the similarity of each projection distance to the mean of all corresponding exclusive projection distances includes:

[0103] The ratio of the minimum value to the maximum value of a single projection distance and the corresponding average value of all exclusive projection distances is regarded as the similarity between the corresponding projection distance and the corresponding average value of all exclusive projection distances.

[0104] In this embodiment, the compatibility of the projection intervals indicates the degree to which the projection intervals can be added to all corresponding exclusive projection intervals to form a value set with highly similar characteristics.

[0105] In this embodiment, the compatibility threshold is the value that the minimum compatibility among the compatibilities of all projection spacings of pairwise adjacent components to be monitored on each shape surface of the server cabinet needs to be less than when the average of all exclusive projection spacings corresponding to the minimum compatibility is taken as the optimum spacing distance of the corresponding shape surface.

[0106] The beneficial effect of the above technology is: by considering the total number of components to be monitored, it is determined whether it is necessary to perform an exclusive analysis on all projection spacings of pairwise adjacent components to be monitored on each shape surface of the server cabinet, and a corresponding exclusive analysis is performed based on the judgment result to achieve deletion and filtering of extreme deviation values ​​in all projection spacings of pairwise adjacent components to be monitored on each shape surface of the server cabinet, thereby ensuring that the optimal spacing distance of each shape surface determined subsequently is sufficiently appropriate.

[0107] Embodiment 5:

[0108] On the basis of Embodiment 3, the reference point determination unit includes:

[0109] A first selection subunit is used to determine, based on the positions of all monitoring points on the inner wall of the server cabinet and the distribution positions of the three-dimensional shapes of all components to be monitored in the server cabinet, a monitoring point closest to each component to be monitored among all monitoring points on the inner wall of the server cabinet as a first selected monitoring point for each component to be monitored;

[0110] A second selection subunit is used for, when the first selected monitoring points of all components to be monitored overlap, determining the second largest value between the distance between each component to be monitored of at least two components to be monitored that overlap with the corresponding first selected monitoring points and the distance between all monitoring points as the first distance, and treating the corresponding overlapping monitoring points as the reference monitoring points of the component to be monitored with the largest first distance among the at least two components to be monitored that overlap with the corresponding first selected monitoring points, and treating the first selected monitoring points of all components to be monitored that do not overlap with the corresponding first selected monitoring points as the reference monitoring points of the corresponding components to be monitored, and determining the reference monitoring points of each component to be monitored that has not been selected as the reference monitoring point among all monitoring points except the monitoring points currently selected as the reference monitoring points;

[0111] The reference point determination subunit is used to use the first selected monitoring point of each component to be monitored as the reference monitoring point of each component to be monitored when the first selected monitoring points of all components to be monitored do not overlap.

[0112] In this embodiment, among all monitoring points except the currently selected reference monitoring point, determining the reference monitoring point of each component to be monitored that has not been selected as the reference monitoring point includes:

[0113] Among all monitoring points except the monitoring point currently selected as the reference monitoring point, determine the monitoring point closest to each component to be monitored that is not selected as the reference monitoring point, and use it as the new first selected monitoring point for each monitoring point that is not selected as the reference monitoring point;

[0114] Determine whether there is overlap between the new first selected monitoring points of all the components to be monitored that have not been selected as reference monitoring points, and if so, determine the second largest value between the distance between each component to be monitored of at least two components to be monitored that overlap with the corresponding new first selected monitoring point and the distance between all monitoring points that have not been selected as reference monitoring points as the new first distance, and use the corresponding overlapping monitoring point as the reference monitoring point of the component to be monitored with the largest new first distance among the at least two components to be monitored that overlap with the corresponding new first selected monitoring point, and use the first selected monitoring points of all components to be monitored that have not been selected as reference monitoring points and do not overlap with the corresponding new first selected monitoring point as the reference monitoring points of the corresponding components to be monitored;

[0115] Otherwise, the first selected monitoring point of each component to be monitored that has not been selected as a reference monitoring point is used as the reference monitoring point of each component to be monitored;

[0116] The above scheme is executed repeatedly until the reference monitoring points of all components to be monitored are determined.

[0117] The beneficial effects of the above technology are: it implements the principle of the closest possible spacing, and allocates non-overlapping reference monitoring points to each component to be monitored, thereby controlling the error in the process of referencing or indirectly monitoring the temperature or heat dissipation status of the component to be monitored by monitoring the temperature of the reference monitoring point to the greatest extent.

[0118] Embodiment 6:

[0119] Based on Example 2, the thermal conductivity conversion submodule includes:

[0120] A thermal conductivity parameter determination unit, used to determine all thermal conductivity media and the conduction path of each thermal conductivity medium between each component to be monitored and the corresponding reference monitoring point;

[0121] The thermal conductivity conversion unit is used to perform thermal conductivity conversion on the temperature limit of each component to be monitored based on all thermal conductivity media between each component to be monitored and the corresponding reference monitoring point and the conduction path of each thermal conductivity medium, so as to obtain the reference temperature limit of the reference monitoring point of each component to be monitored.

[0122] In this embodiment, all heat-conducting media between the component to be monitored and the corresponding reference monitoring point are: the physical medium that needs to be passed through in the server cabinet along the path from the position point with the smallest distance between the corresponding reference monitoring point and the surface of the center to be monitored to the corresponding reference monitoring point, such as air or materials used by other components.

[0123] In this embodiment, the conduction path of the heat conductive medium is a path from the point on the central surface to be monitored where the distance between the corresponding reference monitoring point is the smallest to the corresponding reference monitoring point, and a portion of the path that needs to pass through the corresponding heat conductive medium.

[0124] In this embodiment, based on all heat-conducting media between each component to be monitored and the corresponding reference monitoring point and the conduction path of each heat-conducting medium, the temperature limit of each component to be monitored is converted by thermal conductivity to obtain the reference temperature limit of the reference monitoring point of each component to be monitored, including:

[0125] Determine the conduction paths of all heat-conducting media traversed successively in the path from the position point with the smallest distance between the central surface to be monitored and the corresponding reference monitoring point to the corresponding reference monitoring point;

[0126] Substitute the temperature limit of the component to be monitored and the first conductive path passed into the thermal conductivity formula of the first conductive medium passed (i.e., the thermal conductivity formula corresponding to the thermal conductive medium in the prior art), and calculate the temperature value conducted to the end point of the first conductive path passed when the real-time temperature of the component to be monitored is the temperature limit;

[0127] The temperature value of the end point of the first conductive path is used as the temperature value of the starting point of the second conductive path, and the temperature value of the starting point of the second conductive path and the second conductive path are substituted into the thermal conductivity formula of the second conductive medium to calculate the temperature value of the conductive path conducted to the end point of the second conductive path when the real-time temperature of the component to be monitored is the temperature limit value;

[0128] And so on, until the path from the position point with the smallest distance between the corresponding reference monitoring point and the central surface to be monitored is traversed to the corresponding reference monitoring point, the reference temperature limit value of the reference monitoring point of each component to be monitored is obtained.

[0129] The beneficial effect of the above technology is: based on all heat-conducting media between each component to be monitored and the corresponding reference monitoring point and the conduction path of each heat-conducting medium, thermal conductivity conversion is performed on the temperature limit of each component to be monitored, thereby obtaining the accurate reference temperature limit of the reference monitoring point of each component to be monitored.

[0130] Embodiment 7:

[0131] Based on Example 1, the temperature model building module is referenced Figure 3 ,include:

[0132] The data alignment submodule is used to perform time alignment on all the real-time temperatures of multiple monitoring points on the inner wall of the server cabinet currently acquired (from the time when acquisition starts to a period of time ending at the current time), all the real-time temperatures of the coolant at the inlet of the liquid cooling cycle, and all the real-time temperatures of the cooling liquid at the outlet of the liquid cooling cycle, to obtain the real-time temperatures of multiple monitoring points on the inner wall of the server cabinet currently acquired at each moment, the real-time temperatures of the coolant at the inlet of the liquid cooling cycle, and the real-time temperatures of the cooling liquid at the outlet of the liquid cooling cycle;

[0133] The temperature model generation submodule is used to mark the real-time temperatures of multiple monitoring points on the inner wall of the server cabinet at each moment, the real-time temperature of the coolant at the inlet of the liquid cooling cycle, and the real-time temperature of the cooling liquid at the outlet of the liquid cooling cycle, respectively, into the three-dimensional model of the immersion liquid cooling system of the server cabinet, and obtain the immersion liquid cooling temperature model of the server cabinet at each moment;

[0134] The dynamic model generation submodule is used to sort and connect the immersion liquid cooling temperature models of the server cabinet at all times in time sequence to obtain the immersion liquid cooling dynamic temperature model of the server cabinet.

[0135] In this embodiment, the three-dimensional model of the immersion liquid cooling system is a model representing the three-dimensional shape and size of the device containing the cooling liquid that completely immerses the server cabinet and the three-dimensional shape and size of the server cabinet.

[0136] The beneficial effect of the above technology is: it realizes the construction of an immersion liquid cooling temperature model of the server cabinet at each moment.

[0137] Embodiment 8:

[0138] Based on Example 1, the temperature control equation analysis module refers to Figure 4 ,include:

[0139] Model building submodule, used to build a heat dissipation temperature control equation extraction model;

[0140] The equation output submodule is used to input the immersion liquid cooling dynamic temperature model of the server cabinet with all reference monitoring point locations marked into the heat dissipation temperature control equation extraction model to obtain the heat dissipation temperature control equation for each reference monitoring point.

[0141] In this embodiment, a heat dissipation temperature control equation extraction model is constructed, including:

[0142] The heat dissipation temperature control equations of different monitoring points on the inner wall of the server cabinet immersed in a large number of other server immersion liquid cooling control systems that are artificially inferred, and the immersion liquid cooling dynamic temperature model of the server cabinet that includes the real-time temperature values ​​of the different monitoring points over a period of time and the real-time temperature of the liquid cooling cycle inlet of the server immersion liquid cooling control system over the same period of time are used as the model input and model output of the training samples, respectively, and the heat dissipation temperature control equation extraction model is obtained after training the neural network model.

[0143] The beneficial effect of the above technology is: using the neural network model, accurate analysis of the immersion liquid cooling dynamic temperature model of the server cabinet with all reference monitoring point locations marked can be achieved, and then the heat dissipation temperature control equation of each reference monitoring point can be accurately analyzed.

[0144] Embodiment 9:

[0145] Based on Example 1, the optimal temperature determination module refers to Figure 5 ,include:

[0146] A real-time temperature determination submodule, used to determine the real-time temperatures of all reference monitoring points on the inner wall of the server cabinet based on the real-time temperatures of all monitoring points located on the inner wall of the server cabinet;

[0147] The first optimal temperature determination submodule is used to input the real-time temperatures of all monitoring points on the inner wall of the server cabinet, the reference temperature limits of all reference monitoring points on the inner wall of the server cabinet, and the heat dissipation temperature control equations of all reference monitoring points into a preset temperature rise temperature control model in real time when the real-time temperatures of all reference monitoring points do not exceed the reference temperature limits of the corresponding reference monitoring points, so as to obtain the optimal temperature of the coolant at the inlet of the liquid cooling cycle;

[0148] The second optimal temperature determination submodule is used to input the real-time temperatures of all monitoring points on the inner wall of the server cabinet, the reference temperature limits of all reference monitoring points on the inner wall of the server cabinet, and the heat dissipation temperature control equations of all reference monitoring points into the preset cooling temperature control model in real time when the real-time temperature of at least one reference monitoring point exceeds the reference temperature limit of the corresponding reference monitoring point, so as to obtain the optimal temperature of the coolant at the inlet of the liquid cooling cycle.

[0149] In this embodiment, the preset temperature rise temperature control model is a pre-trained model for determining the optimal coolant temperature at the inlet of the liquid cooling cycle that is higher than the current coolant temperature at the inlet of the liquid cooling cycle according to input data, and the construction process includes:

[0150] In a manual control instance of a server immersion liquid cooling control system (i.e., information on corresponding temperature input and response in the process of manually controlling the server immersion liquid cooling control system), multiple groups of real-time temperatures of liquid cooling cycle inlets corresponding to the time when the earlier real-time temperature of all reference monitoring points in two adjacent moments is less than the reference temperature limit of the corresponding reference monitoring point, and the real-time temperatures of liquid cooling cycle inlets corresponding to the time when the later real-time temperature of all reference monitoring points in two adjacent moments is equal to the reference temperature limit of the corresponding reference monitoring point are extracted as training samples;

[0151] In the training samples, multiple groups of real-time temperatures of all referenced monitoring points at the earlier of two adjacent moments, the referenced temperature limits of all referenced monitoring points and the heat dissipation temperature control equations of all referenced monitoring points are used as model inputs, and the real-time temperature of the liquid cooling cycle inlet at the later of the corresponding two adjacent moments is used as the model output. Neural network model training is performed to obtain the preset heating temperature control model.

[0152] In this embodiment, the preset cooling temperature control model is a pre-trained model for determining the optimal coolant temperature at the inlet of the liquid cooling cycle that is lower than the current coolant temperature at the inlet of the liquid cooling cycle according to input data, and the construction process includes:

[0153] In a manual control instance of a server immersion liquid cooling control system (i.e., information including corresponding temperature input and response in the process of manually controlling the server immersion liquid cooling control system), multiple groups of real-time temperatures of liquid cooling cycle inlets corresponding to at least one reference monitoring point at the earlier of two adjacent time periods are greater than the reference temperature limit of the corresponding reference monitoring point, and real-time temperatures of liquid cooling cycle inlets corresponding to all reference monitoring points at the later of two adjacent time periods are equal to the reference temperature limit of the corresponding reference monitoring point are extracted as training samples;

[0154] In the training samples, multiple groups of real-time temperatures of all referenced monitoring points at the earlier of two adjacent moments, the referenced temperature limits of all referenced monitoring points and the heat dissipation temperature control equations of all referenced monitoring points are used as model inputs, and the real-time temperature of the liquid cooling cycle inlet at the later of the corresponding two adjacent moments is used as the model output. Neural network model training is performed to obtain the preset cooling temperature control model.

[0155] The beneficial effect of the above technology is: based on whether the real-time temperatures of all reference monitoring points do not exceed the reference temperature limit of the corresponding reference monitoring points, according to different situations and combined with the neural network model, it is realized that when the real-time temperatures of all reference monitoring points do not exceed the reference temperature limit of the corresponding reference monitoring points, the real-time temperatures of all monitoring points on the inner wall of the server cabinet and the reference temperature limits of all reference monitoring points on the inner wall of the server cabinet and the heat dissipation temperature control equations of all reference monitoring points are used to reasonably determine the optimal temperature of the coolant at the inlet of the liquid cooling cycle, so that the energy consumption of the server immersion liquid cooling system is minimized while meeting the heat dissipation conditions of all components to be monitored in the server cabinet.

[0156] Embodiment 10:

[0157] The present invention provides a server immersion liquid cooling control method, which is applied to the server immersion liquid cooling control system described in any one of embodiments 1 to 9, referring to Figure 6 ,include:

[0158] S1: Determine the reference temperature limit values ​​of all reference monitoring points on the inner wall of the server cabinet based on the heat dissipation standard of the server cabinet;

[0159] S2: Based on the real-time temperature of multiple monitoring points on the inner wall of the server cabinet, the real-time temperature of the coolant at the inlet of the liquid cooling cycle, and the real-time temperature of the coolant at the outlet of the liquid cooling cycle, an immersion liquid cooling dynamic temperature model of the server cabinet is constructed;

[0160] S3: Determine the heat dissipation temperature control equation for each reference monitoring point based on the immersion liquid cooling dynamic temperature model of the server cabinet;

[0161] S4: Based on the real-time temperatures of all monitoring points located on the inner wall of the server cabinet, the reference temperature limits of all reference monitoring points on the inner wall of the server cabinet, and the heat dissipation temperature control equation, the optimal temperature of the coolant at the inlet of the liquid cooling cycle is determined in real time.

[0162] The beneficial effects of the above technology are: through real-time monitoring and joint analysis of the temperatures of multiple monitoring points on the inner wall of the server cabinet and the temperatures of the liquid cooling cycle inlet and outlet, the heat dissipation temperature control equation of the reference monitoring point of each component to be monitored is analyzed, and the optimal temperature of the coolant at the liquid cooling cycle inlet is accurately controlled according to the heat dissipation temperature control equation, thereby realizing dynamic and precise control of the server immersion liquid cooling system, and being able to cope with the irregular heating of components caused by the data server when processing data loads of different magnitudes, ensuring that the server can operate stably at an appropriate temperature.

[0163] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A server immersion liquid cooling control system, characterized in that: include: A temperature control standard analysis module is used to determine reference temperature limits of all reference monitoring points on the inner wall of the server cabinet based on the heat dissipation standard of the server cabinet; The temperature control standard analysis module includes: a temperature limit reading submodule, which is used to determine the temperature limit values ​​in the temperature control standards of all components to be monitored in the server cabinet in the heat dissipation standard of the server cabinet; a reference point determination submodule, which is used to determine the reference monitoring point of each component to be monitored among all monitoring points on the inner wall of the server cabinet based on the distribution position of each component to be monitored in the server cabinet; a thermal conductivity conversion submodule, which is used to perform thermal conductivity conversion on the temperature limit value of each component to be monitored, and determine the reference temperature limit value of the reference monitoring point of each component to be monitored; wherein the reference monitoring point is a monitoring point used to replace the corresponding component to be monitored for real-time temperature monitoring, and real-time monitoring of the heat dissipation state of the component to be monitored can be achieved through thermal conductivity conversion; A temperature model building module is used to build an immersion liquid cooling dynamic temperature model of the server cabinet based on the real-time temperature of multiple monitoring points on the inner wall of the server cabinet, the real-time temperature of the coolant at the inlet of the liquid cooling cycle, and the real-time temperature of the cooling liquid at the outlet of the liquid cooling cycle; A temperature control equation analysis module is used to determine the heat dissipation temperature control equation for each reference monitoring point based on the immersion liquid cooling dynamic temperature model of the server cabinet; The optimal temperature determination module is used to determine the optimal temperature of the coolant at the inlet of the liquid cooling cycle in real time based on the real-time temperature of all monitoring points located on the inner wall of the server cabinet and the reference temperature limit of all reference monitoring points on the inner wall of the server cabinet and the heat dissipation temperature control equation.

2. The server immersion liquid cooling control system according to claim 1, characterized in that: The reference point determines the submodule, including: A projection spacing determination unit is used to determine the projection area of ​​the three-dimensional shapes of all components to be monitored on each shape surface of the server cabinet, and determine the spacing between two adjacent physical center positions of the projection area of ​​the three-dimensional shapes of all components to be monitored on each shape surface of the server cabinet as the projection spacing of the corresponding two adjacent components to be monitored on the corresponding shape surface of the server cabinet; A spacing distance determination unit, used to determine the most suitable spacing distance of each shape surface of the server cabinet based on all projection distances of two adjacent components to be monitored on each shape surface of the server cabinet; A monitoring point position determination unit, configured to determine the positions of all monitoring points on the inner wall of the server cabinet based on the optimum spacing distance of each shape surface of the server cabinet, wherein the positions of all monitoring points on each shape surface of the inner wall of the server cabinet are distributed with the optimum spacing distance of the corresponding shape surface as the spacing distance between two adjacent monitoring points; The reference point determination unit is used to determine the reference monitoring point of each component to be monitored among all the monitoring points on the inner wall of the server cabinet based on the positions of all the monitoring points on the inner wall of the server cabinet and the distribution positions of the three-dimensional shapes of all the components to be monitored in the server cabinet.

3. The server immersion liquid cooling control system according to claim 2, characterized in that: A separation distance determination unit, comprising: An exclusive setting subunit, used for treating all projection spacings of the two adjacent components to be monitored on each shape surface of the server cabinet except the current projection spacing as exclusive projection spacings of the current projection spacing when the total number of all components to be monitored is greater than 3; A compatibility calculation subunit is used to calculate the similarity between each projection interval and the average of all corresponding exclusive projection intervals as the compatibility of each projection interval; The first analysis subunit is used to determine whether the minimum compatibility among the compatibilities of all projection spacings of the two adjacent components to be monitored on each shape surface of the server cabinet is less than the compatibility threshold. If so, the average of all exclusive projection spacings of the projection spacing corresponding to the minimum compatibility is used as the optimum spacing distance of the corresponding shape surface. Otherwise, the spacing of all projection spacings of the two adjacent components to be monitored on the corresponding shape surface of the server cabinet is used as the optimum spacing distance of the corresponding shape surface. The second analysis subunit is used to regard the distance between all projections of the two adjacent components to be monitored on the corresponding shape surface of the server cabinet as the most suitable spacing distance of the corresponding shape surface when the total number of all components to be monitored is not greater than 3.

4. The server immersion liquid cooling control system according to claim 2, characterized in that: The reference point determination unit includes: A first selection subunit is used to determine, based on the positions of all monitoring points on the inner wall of the server cabinet and the distribution positions of the three-dimensional shapes of all components to be monitored in the server cabinet, a monitoring point closest to each component to be monitored among all monitoring points on the inner wall of the server cabinet as a first selected monitoring point for each component to be monitored; A second selection subunit is used for, when the first selected monitoring points of all components to be monitored overlap, determining the second largest value between the distance between each component to be monitored of at least two components to be monitored that overlap with the corresponding first selected monitoring points and the distance between all monitoring points as the first distance, and treating the corresponding overlapping monitoring points as the reference monitoring points of the component to be monitored with the largest first distance among the at least two components to be monitored that overlap with the corresponding first selected monitoring points, and treating the first selected monitoring points of all components to be monitored that do not overlap with the corresponding first selected monitoring points as the reference monitoring points of the corresponding components to be monitored, and determining the reference monitoring points of each component to be monitored that has not been selected as the reference monitoring point among all monitoring points except the monitoring points currently selected as the reference monitoring points; The reference point determination subunit is used to use the first selected monitoring point of each component to be monitored as the reference monitoring point of each component to be monitored when the first selected monitoring points of all components to be monitored do not overlap.

5. The server immersion liquid cooling control system according to claim 1, characterized in that: Temperature control equation analysis module, including: Model building submodule, used to build a heat dissipation temperature control equation extraction model; The equation output submodule is used to input the immersion liquid cooling dynamic temperature model of the server cabinet with all reference monitoring point locations marked into the heat dissipation temperature control equation extraction model to obtain the heat dissipation temperature control equation for each reference monitoring point.

6. A server immersion liquid cooling control method, characterized in that: The server immersion liquid cooling control system applied to any one of claims 1 to 5 comprises: S1: Determine the reference temperature limit values ​​of all reference monitoring points on the inner wall of the server cabinet based on the heat dissipation standard of the server cabinet; S2: Based on the real-time temperature of multiple monitoring points on the inner wall of the server cabinet, the real-time temperature of the coolant at the inlet of the liquid cooling cycle, and the real-time temperature of the coolant at the outlet of the liquid cooling cycle, an immersion liquid cooling dynamic temperature model of the server cabinet is constructed; S3: Determine the heat dissipation temperature control equation for each reference monitoring point based on the immersion liquid cooling dynamic temperature model of the server cabinet; S4: Based on the real-time temperatures of all monitoring points located on the inner wall of the server cabinet, the reference temperature limits of all reference monitoring points on the inner wall of the server cabinet, and the heat dissipation temperature control equation, the optimal temperature of the coolant at the inlet of the liquid cooling cycle is determined in real time.

Images