Heat conduction control method and device for vapor chamber radiator module

Abstract

The invention provides a heat conduction control method and device for a vapor chamber radiator module, and relates to the technical field of intelligent heat conduction control, and the method comprises the steps: monitoring and obtaining real-time operation parameters of a heat source region, carrying out the thermal load change analysis in a preset time window, and generating predicted temperature rise rate distribution; respectively carrying out regional discrete evaluation on the real-time heat source temperature distribution and the predicted temperature rise rate distribution to obtain a high heat distribution discrete coefficient and a temperature rise distribution discrete coefficient; and setting an adaptive heat conduction strategy based on the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient, and executing heat dissipation control of the heat source area. Through the method, the technical problems of low heat management efficiency and influence on the overall performance and reliability of the server caused by difficulty in realizing rapid and accurate heat conduction adjustment under the dynamic heat load in an existing method can be solved, matched heat conduction control parameters can be set according to the actual heat load state, the heat dissipation efficiency is remarkably improved, and the service life of the server is prolonged. And stable and reliable operation of the server is ensured.

Description

Technical Field

[0001] The present application relates to the field of intelligent thermal conductivity control technology, and in particular to a thermal conductivity control method and device for a vapor chamber heat sink module. Background Art

[0002] With the rapid development of cloud computing, big data and artificial intelligence technologies, modern data centers and high-performance servers are facing increasingly higher computing demands. The core components of these devices (such as CPU, GPU, memory modules, etc.) will generate a lot of heat when running at high load, forming a dynamic and complex heat load distribution. How to efficiently manage this heat has become one of the key technical challenges to ensure device performance and energy efficiency.

[0003] In recent years, vapor chambers have been widely used in servers and data centers as a highly efficient heat conduction technology. Vapor chambers transfer heat quickly by utilizing the phase change principle, and have higher thermal conductivity than traditional heat conduction technologies. However, existing vapor chamber thermal control methods still cannot meet the requirements for fast response and precise adjustment in dynamic heat load scenarios, further limiting their potential for application in high heat density equipment. Summary of the invention

[0004] The purpose of this application is to provide a thermal conductivity control method and device for a vapor chamber heat sink module, so as to solve the technical problem that the existing thermal conductivity control method is difficult to achieve fast and accurate thermal conductivity regulation under dynamic heat load, resulting in low thermal management efficiency and affecting the overall performance and reliability of the server.

[0005] In view of the above problems, the present application provides a thermal conductivity control method and device for a vapor chamber heat sink module.

[0006] In a first aspect, the present application provides a thermal conductivity control method for a vapor chamber heat sink module, which is implemented by a thermal conductivity control device for a vapor chamber heat sink module, comprising: real-time temperature acquisition of a heat source area through a temperature monitoring array deployed on a server to obtain a real-time heat source temperature distribution; monitoring and obtaining real-time operating parameters of the heat source area, performing a heat load change analysis within a predetermined time window based on the real-time operating parameters, and generating a predicted temperature rise rate distribution; performing regional discrete evaluation on the real-time heat source temperature distribution and the predicted temperature rise rate distribution, respectively, to obtain a high heat distribution discrete coefficient and a temperature rise distribution discrete coefficient; setting an adaptive thermal conductivity strategy based on the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient, and controlling the vapor chamber heat sink module to perform heat dissipation control of the heat source area according to the adaptive thermal conductivity strategy.

[0007] In the second aspect, the present application also provides a thermal conductivity control device for a vapor chamber heat sink module, which is used to execute a thermal conductivity control method for a vapor chamber heat sink module as described in the first aspect, including: a heat source temperature distribution acquisition module, which is used to perform real-time temperature acquisition of a heat source area through a temperature monitoring array deployed on a server, and obtain the real-time heat source temperature distribution; a heat load change analysis module, which is used to monitor and obtain real-time operating parameters of the heat source area, and perform heat load change analysis within a predetermined time window according to the real-time operating parameters to generate a predicted temperature rise rate distribution; a regional discrete evaluation module, which is used to perform regional discrete evaluation on the real-time heat source temperature distribution and the predicted temperature rise rate distribution, respectively, to obtain a high heat distribution discrete coefficient and a temperature rise distribution discrete coefficient; an adaptive thermal conductivity strategy setting module, which is used to set an adaptive thermal conductivity strategy based on the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient, and control the vapor chamber heat sink module to perform heat dissipation control of the heat source area according to the adaptive thermal conductivity strategy.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages: Through the temperature monitoring array deployed on the server, the real-time temperature of the heat source area is collected to obtain the real-time heat source temperature distribution; the real-time operating parameters of the heat source area are monitored and obtained, and the heat load change analysis within a predetermined time window is performed according to the real-time operating parameters to generate a predicted temperature rise rate distribution; the real-time heat source temperature distribution and the predicted temperature rise rate distribution are respectively evaluated for regional discreteness to obtain the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient; an adaptive thermal conductivity strategy is set based on the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient, and according to the adaptive thermal conductivity strategy, the temperature vapor chamber radiator module is controlled to perform heat dissipation control of the heat source area; the accuracy and intelligence of the heat load analysis can be improved, and then the matching thermal conductivity control parameters can be set according to the actual heat load state to achieve fast and accurate thermal conductivity regulation, thereby achieving the technical effect of significantly improving the heat dissipation efficiency and ensuring stable and reliable operation of the server.

[0009] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0011] Figure 1 A schematic flow chart of a heat conduction control method for a vapor chamber heat sink module of the present application; Figure 2 This is a schematic structural diagram of a thermal conductivity control device for a vapor chamber heat sink module according to the present application.

[0012] Description of reference numerals: Heat source temperature distribution acquisition module 11, heat load change analysis module 12, regional discrete evaluation module 13, adaptive heat conduction strategy setting module 14. DETAILED DESCRIPTION

[0013] The present application provides a thermal conductivity control method and device for a vapor chamber heat sink module, which solves the technical problem that the existing thermal conductivity control method is difficult to achieve fast and accurate thermal conductivity regulation under dynamic heat load, resulting in low thermal management efficiency and affecting the overall performance and reliability of the server; the accuracy and intelligence of thermal load analysis can be improved, and matching thermal conductivity control parameters can be set according to the actual thermal load state to achieve fast and accurate thermal conductivity regulation, thereby achieving the technical effect of significantly improving the heat dissipation efficiency and ensuring stable and reliable operation of the server.

[0014] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0015] For example, please refer to the attached Figure 1 The present application provides a thermal conductivity control method for a vapor chamber heat sink module, which is applied to a thermal conductivity control device for a vapor chamber heat sink module, and specifically includes the following steps: Step 1: Use the temperature monitoring array deployed on the server to collect real-time temperature data of the heat source area and obtain the real-time temperature distribution of the heat source.

[0016] Specifically, the server heat source area refers to the hardware components or areas inside the server that generate significant heat. The heat in these areas needs to be effectively monitored and managed to ensure the performance, stability and reliability of the server, such as CPU, GPU, memory module, etc. In order to achieve accurate monitoring of the server heat source area, the temperature monitoring array is used as the core monitoring method. The temperature monitoring array is deployed in the server and contains multiple temperature sensors. Each sensor is responsible for collecting temperature data of a predetermined divided area to ensure that all key parts in the heat source area are covered. Then, the temperature of the heat source area is collected in real time through the temperature monitoring array, that is, each sensor records the temperature of its corresponding area in real time to generate local temperature data. Then, the data of all sensors are aggregated to form a real-time temperature distribution map of the entire heat source area. By fine-grained division of the heat source area, the temperature monitoring array can capture the temperature changes of each area and avoid the problem of missing local information in traditional monitoring methods. By obtaining the real-time heat source temperature distribution, a strong foundation can be provided for subsequent dynamic heat load analysis and intelligent heat conduction strategy optimization.

[0017] Step 2: Monitor and obtain real-time operating parameters of the heat source area, perform heat load change analysis within a predetermined time window based on the real-time operating parameters, and generate a predicted temperature rise rate distribution.

[0018] Specifically, by monitoring the real-time operating parameters of the heat source area, the changes in heat load within the future predetermined time window can be predicted, and the temperature rise rate distribution can be generated, providing a basis for precise heat conduction control. First, at the predetermined monitoring time node, the real-time operating parameters of the heat source area are monitored and obtained. The real-time operating parameters of the heat source area are an important source of data that reflects the dynamic changes in the thermal load of the equipment. They can be set according to the type of heat source area, such as CPU / GPU utilization rate, where high utilization rate is usually accompanied by high power consumption and high heat load; power consumption, where power consumption is directly related to heat generation and can be obtained through hardware sensors or energy consumption monitoring modules; frequency and voltage, where high frequency and high voltage state correspond to higher heat load.

[0019] Next, based on the collected real-time operating parameters, the heat load change trend of the heat source area within the predetermined time window is analyzed. For example, by building a temperature rise predictor based on the BP neural network, the heat load change trend of the heat source area within the predetermined time window is analyzed, and accurate temperature rise prediction results are generated. The predicted temperature rise rate of each predetermined area is obtained, and the predicted temperature rise rate distribution is constructed, where the predicted temperature rise rate distribution is used to reflect the temperature change trend of different areas in the future time window. By predicting the temperature rise rate distribution, possible high-temperature areas or abnormal temperature rise trends can be identified in advance to prevent overheating problems. At the same time, targeted local heat dissipation strategies can be formulated based on regional temperature rise rate differences to achieve precise heat dissipation.

[0020] Step three: Perform regional discrete evaluation on the real-time heat source temperature distribution and the predicted temperature rise rate distribution, respectively, to obtain the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient.

[0021] Specifically, firstly, high heat areas are screened in the real-time heat source temperature distribution, that is, areas with temperature values ​​greater than a threshold value are screened out from the real-time heat source temperature distribution and set as high heat areas, and multiple high heat areas are determined; on the other hand, abnormal temperature rise areas are screened in the predicted temperature rise rate distribution, that is, areas with temperature rise rates greater than a threshold value are screened out from the predicted temperature rise rate distribution and set as abnormal temperature rise areas, and multiple abnormal temperature rise areas are determined.

[0022] Then, the regional distribution discrete evaluation of multiple high-heat areas is carried out to obtain the high-heat distribution discrete coefficient, and the regional distribution discrete evaluation of multiple abnormal temperature rise areas is carried out to obtain the temperature rise distribution discrete coefficient. The smaller the high-heat distribution discrete coefficient is, the higher the temperature distribution is, the more uniform the temperature in the high-heat area is, the more stable the heat load is, and global heat dissipation is required; the larger the high-heat distribution discrete coefficient is, the more uneven the temperature distribution is, the more local hot spots exist, and local targeted heat dissipation can be carried out. The smaller the temperature rise distribution discrete coefficient is, the more stable the temperature rise trend is in the abnormal temperature rise area, and the overall heat load change is controllable. It is recommended to appropriately adjust the global heat dissipation power; the smaller and larger the temperature rise distribution discrete coefficient is, the more uneven the temperature rise trend is in the abnormal temperature rise area, and new local hot spots may be formed. It is recommended to cool the areas with faster temperature rise in a targeted manner.

[0023] Step 4: setting an adaptive heat conduction strategy based on the high heat distribution dispersion coefficient and the temperature rise distribution dispersion coefficient, and controlling the vapor chamber heat sink module to perform heat dissipation control of the heat source area according to the adaptive heat conduction strategy.

[0024] Specifically, an adaptive thermal conductivity strategy is set according to the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient, that is, if the high heat distribution discrete coefficient is less than the predetermined high heat discrete scalar and the temperature rise distribution discrete coefficient is less than the predetermined temperature rise discrete scalar, the adaptive thermal conductivity strategy is set to global heat dissipation processing; if the high heat distribution discrete coefficient is greater than or equal to the predetermined high heat discrete scalar and / or the temperature rise distribution discrete coefficient is greater than or equal to the predetermined temperature rise discrete scalar, the adaptive thermal conductivity strategy is set to local heat dissipation processing. Finally, according to the adaptive thermal conductivity strategy, the heat sink module of the temperature vapor chamber is controlled to perform heat dissipation control of the heat source area. By calculating the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient in real time, the adaptive thermal conductivity strategy is dynamically set, and global heat dissipation processing or local heat dissipation processing is selected, thereby achieving accurate and efficient thermal management, and effectively improving the heat dissipation performance and operating efficiency of the server.

[0025] The thermal conductivity control method for a heat spreader module is applied to a thermal conductivity control device for a heat spreader module, which can solve the technical problem that the existing thermal conductivity control method is difficult to achieve fast and accurate thermal conductivity regulation under dynamic heat load, resulting in low thermal management efficiency and affecting the overall performance and reliability of the server. The temperature monitoring array deployed on the server is used to collect the real-time temperature of the heat source area to obtain the real-time heat source temperature distribution; the real-time operating parameters of the heat source area are monitored and obtained, and the heat load change analysis within a predetermined time window is performed according to the real-time operating parameters to generate a predicted temperature rise rate distribution; the real-time heat source temperature distribution and the predicted temperature rise rate distribution are respectively evaluated for regional discreteness to obtain the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient; based on the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient, an adaptive thermal conductivity strategy is set, and according to the adaptive thermal conductivity strategy, the heat spreader module is controlled to perform heat dissipation control of the heat source area; the accuracy and intelligence of the heat load analysis can be improved, and then the matching thermal conductivity control parameters can be set according to the actual heat load state to achieve fast and accurate thermal conductivity regulation, thereby achieving a significant improvement in heat dissipation efficiency and ensuring the technical effect of stable and reliable operation of the server.

[0026] Furthermore, the present application includes: The temperature monitoring array includes N temperature sensors for collecting temperature of N predetermined divided areas of the heat source area, wherein each temperature sensor corresponds to one predetermined divided area, and N is an integer greater than 5.

[0027] Specifically, a temperature monitoring array is constructed, wherein the temperature monitoring array includes N temperature sensors for collecting temperature of N predetermined divided areas of the heat source area, where N is an integer greater than 5, and the value of N can be dynamically set according to the complexity and distribution characteristics of the heat source area to ensure the comprehensiveness and accuracy of the monitoring; the heat source area is divided into N predetermined sub-areas, each of which is configured with a temperature sensor, and each temperature sensor is responsible for real-time monitoring of the temperature of the corresponding sub-area to ensure that the heat changes of each sub-area are independently monitored. Through the design and deployment of the temperature monitoring array, real-time and accurate monitoring of the heat source area can be achieved, providing strong data support for dynamic heat load analysis and heat dissipation strategy optimization.

[0028] Furthermore, the present application includes: N temperature rise prediction branches are constructed based on the BP neural network, wherein the temperature rise prediction branches correspond to the predetermined divided areas one by one; a sample operation parameter set and a plurality of sample temperature rise rate sets are collected, the sample operation parameter is used as input, and the sample temperature rise rate set is used as output, the N temperature rise prediction branches are supervised and trained by using the sample operation parameter set and the plurality of sample temperature rise rate sets, N converged temperature rise prediction branches are obtained, and a temperature rise predictor is constructed; the real-time operation parameters are input into the temperature rise predictor, N predicted temperature rise rates are output, and the predicted temperature rise rate distribution is generated in combination with the N predetermined divided areas.

[0029] Specifically, N temperature rise prediction branches are constructed based on the BP neural network, and each temperature rise prediction branch corresponds to a predetermined divided area, that is, each prediction branch is a sub-model constructed based on the BP neural network, which is responsible for predicting the temperature rise rate of the corresponding predetermined divided area, including an input layer, multiple hidden layers and an output layer, wherein the input data of the input layer is the operating parameters, and the output data of the output layer is the temperature rise rate. Then, a sample operating parameter set and multiple sample temperature rise rate sets are collected as training data, and the sample operating parameters are used as input and the sample temperature rise rate set is used as output. The N temperature rise prediction branches are supervised and trained using the sample operating parameter set and the multiple sample temperature rise rate sets, wherein each prediction branch is trained independently to ensure that it is targetedly adapted to the characteristics of the corresponding area. For each temperature rise prediction branch, a gradient descent algorithm and a loss function are used for supervised training, and the weight and bias parameters are continuously updated until the loss function converges, and N converged temperature rise prediction branches are obtained, and a temperature rise predictor is built based on the N converged temperature rise prediction branches for real-time prediction of the temperature rise rate distribution of the heat source area. By building a temperature rise predictor based on the BP neural network, accurate temperature rise prediction of the heat source area can be achieved, providing key decision support for dynamic heat dissipation control and improving the intelligence level of thermal management.

[0030] Finally, the real-time operating parameters (such as CPU usage, power consumption, etc.) are input into the temperature rise predictor to obtain N predicted temperature rise rates, and the predicted temperature rise rate distribution is generated in combination with the N predetermined divided areas.

[0031] Furthermore, the present application includes: Taking the attribute characteristics of the heat source area as constraints, multiple sample operating parameters are retrieved and obtained, a sample operating parameter set is constructed, and a first sample operating parameter is randomly selected; taking the attribute characteristics of the heat source area and the first sample operating parameter as constraints, a sample temperature rise rate distribution set within a historical predetermined time window is retrieved and obtained; the mode of the sample temperature rise rate distribution set is selected as the first sample temperature rise rate distribution, and the first sample temperature rise rate distribution is divided and marked according to the N predetermined divided areas to obtain a first sample temperature rise rate set, which is added to the multiple sample temperature rise rate sets, wherein the sample operating parameters and the sample temperature rise rate sets have a corresponding relationship.

[0032] Specifically, the attribute characteristics of the heat source area are obtained, including equipment type, area location, etc.; then, information retrieval is performed in the historical operation log with the attribute characteristics of the heat source area as a constraint, multiple sample operation parameters are obtained to construct a sample operation parameter set, and any sample operation parameter is randomly selected from the sample operation parameter set as the first sample operation parameter. Further, information retrieval is performed in the historical operation log with the attribute characteristics of the heat source area and the first sample operation parameter as constraints, that is, historical sample data matching the attribute characteristics of the current heat source area and the first sample operation parameter is obtained, and the temperature rise rate distribution within the predetermined time window corresponding to the sample is extracted to construct a sample temperature rise rate distribution set.

[0033] Further, the mode of the sample temperature rise rate distribution set is selected as the first sample temperature rise rate distribution, the mode is the temperature rise rate distribution that appears most frequently in the sample set, and represents the most common regional temperature rise rate mode; then, the first sample temperature rise rate distribution is divided and identified according to the N predetermined divided areas to obtain a first sample temperature rise rate set; using the same method, a plurality of sample temperature rise rate sets corresponding to a plurality of sample operating parameters are obtained in turn, wherein each sample operating parameter corresponds to a sample temperature rise rate set.

[0034] Furthermore, the present application includes: The real-time heat source temperature distribution is screened according to a predetermined temperature threshold value, and a predetermined divided area greater than the predetermined temperature threshold value is marked as a high heat area, and a plurality of high heat areas are determined; the predicted temperature rise rate distribution is screened according to a predetermined temperature rise rate threshold value, and a predetermined divided area greater than the predetermined temperature rise rate threshold value is marked as an abnormal temperature rise area, and a plurality of abnormal temperature rise areas are determined; and regional distribution discrete evaluation is performed on the plurality of high heat areas and the plurality of abnormal temperature rise areas, respectively, to obtain the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient.

[0035] Specifically, a predetermined temperature threshold value and a predetermined temperature rise rate threshold value are obtained, wherein the threshold value refers to the data standard value for safe operation of the heat source area, and can be set according to the actual scene requirements. Then, the real-time heat source temperature distribution is screened according to the predetermined temperature threshold value, and the predetermined divided area greater than the predetermined temperature threshold value is marked as a high-heat area, and multiple high-heat areas are determined; the predicted temperature rise rate distribution is screened according to the predetermined temperature rise rate threshold value, and the predetermined divided area greater than the predetermined temperature rise rate threshold value is marked as an abnormal temperature rise area, and multiple abnormal temperature rise areas are determined.

[0036] Further, the regional distribution discrete evaluation is performed on the multiple high-heat areas and the multiple abnormal temperature rise areas respectively to quantify the distribution characteristics of the temperature and the temperature rise trend, and obtain the high-heat distribution discrete coefficient and the temperature rise distribution discrete coefficient. Through this evaluation process, the distribution characteristics of the high-heat area and the abnormal temperature rise area can be comprehensively analyzed to provide efficient decision support for dynamic thermal management and intelligent heat dissipation strategies.

[0037] Furthermore, the present application includes: If the high heat distribution dispersion coefficient is less than a predetermined high heat dispersion scalar and the temperature rise distribution dispersion coefficient is less than a predetermined temperature rise dispersion scalar, the adaptive thermal conductivity strategy is set to global heat dissipation processing, wherein global heat dissipation processing refers to performing global heat dissipation processing on all areas of the heat source area according to the first optimized thermal conductivity scheme; if the high heat distribution dispersion coefficient is greater than or equal to the predetermined high heat dispersion scalar and / or the temperature rise distribution dispersion coefficient is greater than or equal to the predetermined temperature rise dispersion scalar, the adaptive thermal conductivity strategy is set to local heat dissipation processing, wherein local heat dissipation processing refers to performing local heat dissipation processing on multiple high heat areas of the heat source area according to the second optimized thermal conductivity scheme.

[0038] Specifically, through the analysis results of the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient, the adaptive thermal conduction strategy is dynamically set, global heat dissipation treatment or local heat dissipation treatment is selected, and the corresponding optimized thermal conduction scheme is executed to ensure efficient heat dissipation and stable operation in the heat source area.

[0039] If the high heat distribution discrete coefficient is less than the predetermined high heat discrete scalar and the temperature rise distribution discrete coefficient is less than the predetermined temperature rise discrete scalar, indicating that the temperature and heating rate of the heat source area are evenly distributed, without significant local hot spots or concentrated heating trends, the adaptive heat conduction strategy is set to global heat dissipation treatment to evenly reduce the temperature of all areas and optimize the overall heat dissipation efficiency, wherein global heat dissipation treatment refers to global heat dissipation treatment of all areas of the heat source area according to the first optimized heat conduction scheme. If the high heat distribution discrete coefficient is greater than or equal to the predetermined high heat discrete scalar and / or the temperature rise distribution discrete coefficient is greater than or equal to the predetermined temperature rise discrete scalar, indicating that there are significant local hot spots or concentrated heating trend areas in the heat source area, the adaptive heat conduction strategy is set to local heat dissipation treatment, that is, concentrated cooling of high heat areas or abnormal heating trend areas to avoid local overheating, wherein local heat dissipation treatment refers to local heat dissipation treatment of multiple high heat areas in the heat source area according to the second optimized heat conduction scheme. By setting a heat conduction strategy based on the discrete coefficient and executing the optimized heat conduction scheme, dynamic and precise thermal management can be achieved, effectively improving the heat dissipation performance and reliability of the equipment.

[0040] Furthermore, the present application includes: A twin heat conduction space of the vapor chamber heat sink module and the heat source area is constructed, and heat conduction control parameters and heat conduction control mechanisms are input into the twin heat conduction space, wherein the heat conduction control parameters include fluid flow path, fluid reflux velocity and fluid phase change rate; multiple temperature overflow values ​​of the multiple high-heat areas are obtained, and a maximum temperature overflow value is selected, wherein the temperature overflow value refers to the difference between the temperature monitoring value of the high-heat area and the predetermined temperature threshold value; multiple temperature rise overflow values ​​of the multiple abnormal temperature rise areas are obtained, and a temperature rise overflow mean is calculated, wherein the temperature rise overflow value refers to the difference between the predicted temperature rise rate of the abnormal temperature rise area and the predetermined temperature rise rate threshold value; temperature rise compensation is performed on the maximum temperature overflow value according to the temperature rise overflow mean to obtain a global temperature overflow value; based on the twin heat conduction space, a heat conduction control parameter optimization analysis is performed to eliminate the global temperature overflow value as a constraint, and the first optimized heat conduction scheme is generated, wherein the first optimized heat conduction scheme is a heat conduction scheme with the largest heat conduction adaptability, and the heat conduction adaptability is negatively correlated with the heat dissipation time and energy consumption.

[0041] Specifically, taking the physical vapor chamber radiator module and heat source area as the objects, simulation and modeling technology are used to construct a digital virtual thermal management system, called a twin thermal conduction space, which maps the state of the physical system in real time and supports thermal conduction parameter optimization and heat dissipation strategy simulation; then, the thermal conduction control parameters and thermal conduction control mechanism are input into the twin thermal conduction space, where the thermal conduction control parameters include fluid flow path, fluid reflux velocity and fluid phase change rate. The fluid flow path refers to the heat transfer path of the working fluid in the vapor chamber from the high heat area to the cold end. The fluid reflux velocity refers to the reflux velocity of the fluid after the heat transfer is completed, which affects the cycle efficiency; the fluid phase change rate refers to the evaporation and condensation rate of the working fluid in the vapor chamber; the thermal conduction control mechanism refers to the dynamic adjustment of the fluid flow path, reflux velocity and phase change rate to adapt to the temperature changes in the heat source area.

[0042] Next, multiple temperature overflow values ​​of the multiple high-heat areas are obtained, and the maximum temperature overflow value is selected, wherein the temperature overflow value refers to the difference between the temperature monitoring value of the high-heat area and the predetermined temperature threshold value; multiple temperature rise overflow values ​​of the multiple abnormal temperature rise areas are obtained, wherein the temperature rise overflow value refers to the difference between the predicted temperature rise rate of the abnormal temperature rise area and the predetermined temperature rise rate threshold value, and then the multiple temperature rise overflow values ​​are averaged to obtain the temperature rise overflow average. Further, the maximum temperature overflow value is compensated for temperature rise according to the temperature rise overflow average, that is, the temperature compensation coefficient is determined according to the temperature rise overflow average, and the maximum temperature overflow value is compensated for temperature rise, wherein the temperature compensation coefficient can be obtained according to historical data analysis, and the global temperature overflow value is obtained.

[0043] Then, in the twin heat conduction space, with the elimination of the global temperature overflow value as a constraint, a heat conduction control parameter optimization analysis is performed to generate multiple first initial heat conduction schemes that meet the constraint conditions, wherein the first initial heat conduction scheme includes simulated heat dissipation time and simulated energy consumption; further, a thermal conductivity fitness evaluation is performed on the multiple first initial heat conduction schemes, wherein the thermal conductivity fitness is negatively correlated with the heat dissipation time and energy consumption, that is, the shorter the heat dissipation time and the smaller the energy consumption, the greater the thermal conductivity fitness, and multiple thermal conductivity fitnesses of the multiple first initial heat conduction schemes are obtained, and the heat conduction scheme with the largest thermal conductivity fitness is selected as the first optimized heat conduction scheme.

[0044] Furthermore, the present application includes: Temperature rise compensation is performed on multiple temperature overflow values ​​of the multiple high-heat areas according to the temperature rise overflow mean value to obtain multiple local temperature overflow values; based on the twin heat conduction space, heat conduction control parameter optimization analysis is performed to eliminate the multiple local temperature overflow values ​​as constraints, and the second optimized heat conduction scheme is generated.

[0045] Specifically, temperature rise compensation is performed on multiple temperature overflow values ​​of the multiple high-heat areas according to the temperature rise overflow mean value, that is, the temperature compensation coefficient is determined according to the temperature rise overflow mean value, and temperature rise compensation is performed on multiple temperature overflow values ​​to obtain multiple local temperature overflow values. Then, in the twin heat conduction space, the heat conduction control parameter optimization analysis is performed to eliminate the multiple local temperature overflow values ​​as constraints, and multiple second initial heat conduction schemes and multiple heat conduction fitnesses are obtained, and the second initial heat conduction scheme with the largest heat conduction fitness is selected as the second optimized heat conduction scheme.

[0046] In summary, the thermal conductivity control method for a vapor chamber heat sink module provided in the present application has the following technical effects: Through the temperature monitoring array deployed on the server, real-time temperature collection is performed on the heat source area to obtain the real-time heat source temperature distribution; the real-time operating parameters of the heat source area are monitored and obtained, and the heat load change analysis within a predetermined time window is performed according to the real-time operating parameters to generate a predicted temperature rise rate distribution; the real-time heat source temperature distribution and the predicted temperature rise rate distribution are respectively evaluated for regional discreteness to obtain the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient; an adaptive thermal conductivity strategy is set based on the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient, and according to the adaptive thermal conductivity strategy, the temperature vapor chamber radiator module is controlled to perform heat dissipation control of the heat source area; the accuracy and intelligence of the heat load analysis can be improved, and then the matching thermal conductivity control parameters can be set according to the actual heat load state to achieve fast and accurate thermal conductivity regulation, thereby achieving the technical effect of significantly improving the heat dissipation efficiency and ensuring stable and reliable operation of the server.

[0047] Embodiment 2: Based on the thermal conductivity control method for a vapor chamber heat sink module in the above embodiment, the present application also provides a thermal conductivity control device for a vapor chamber heat sink module, see the attached Figure 2 , including: a heat source temperature distribution acquisition module 11, which is used to collect real-time temperature of the heat source area through a temperature monitoring array deployed on the server to obtain the real-time heat source temperature distribution; a heat load change analysis module 12, which is used to monitor and obtain the real-time operating parameters of the heat source area, and perform heat load change analysis within a predetermined time window according to the real-time operating parameters to generate a predicted temperature rise rate distribution; a regional discrete evaluation module 13, which is used to perform regional discrete evaluation on the real-time heat source temperature distribution and the predicted temperature rise rate distribution, respectively, to obtain a high heat distribution discrete coefficient and a temperature rise distribution discrete coefficient; an adaptive thermal conductivity strategy setting module 14, which is used to set an adaptive thermal conductivity strategy based on the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient, and control the temperature vapor chamber radiator module to perform heat dissipation control of the heat source area according to the adaptive thermal conductivity strategy.

[0048] Furthermore, the thermal conductivity control device for the temperature vapor chamber heat sink module is also used for: the temperature monitoring array includes N temperature sensors, which are used to collect temperature of N predetermined divided areas in the heat source area, wherein each temperature sensor corresponds to a predetermined divided area, and N is an integer greater than 5.

[0049] Furthermore, the thermal conductivity control device for the temperature vapor chamber heat sink module is also used to: construct N temperature rise prediction branches based on the BP neural network, wherein the temperature rise prediction branches correspond to the predetermined divided areas one by one; collect a sample operating parameter set and a plurality of sample temperature rise rate sets, use the sample operating parameter set as input and the sample temperature rise rate set as output, perform supervised training on the N temperature rise prediction branches using the sample operating parameter set and the plurality of sample temperature rise rate sets, obtain N converged temperature rise prediction branches, and build a temperature rise predictor; input the real-time operating parameters into the temperature rise predictor, output N predicted temperature rise rates, and generate the predicted temperature rise rate distribution in combination with the N predetermined divided areas.

[0050] Furthermore, the thermal conductivity control device for the temperature vapor chamber heat sink module is also used to: retrieve and obtain multiple sample operating parameters based on the attribute characteristics of the heat source area, construct a sample operating parameter set, and randomly select a first sample operating parameter; retrieve and obtain a sample temperature rise rate distribution set within a historical predetermined time window based on the attribute characteristics of the heat source area and the first sample operating parameter; select the mode of the sample temperature rise rate distribution set as the first sample temperature rise rate distribution, and divide and identify the first sample temperature rise rate distribution according to the N predetermined divided areas to obtain a first sample temperature rise rate set, and add it to the multiple sample temperature rise rate sets, wherein the sample operating parameters and the sample temperature rise rate sets have a corresponding relationship.

[0051] Furthermore, the thermal conductivity control device for the temperature vapor chamber heat sink module is also used to: screen the real-time heat source temperature distribution according to a predetermined temperature threshold value, mark the predetermined divided area greater than the predetermined temperature threshold value as a high heat area, and determine a plurality of high heat areas; screen the predicted temperature rise rate distribution according to a predetermined temperature rise rate threshold value, mark the predetermined divided area greater than the predetermined temperature rise rate threshold value as an abnormal temperature rise area, and determine a plurality of abnormal temperature rise areas; perform regional distribution discrete evaluation on the plurality of high heat areas and the plurality of abnormal temperature rise areas, respectively, to obtain the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient.

[0052] Furthermore, the thermal conductivity control device for the temperature vapor chamber heat sink module is also used for: if the high heat distribution discrete coefficient is less than a predetermined high heat discrete scalar and the temperature rise distribution discrete coefficient is less than a predetermined temperature rise discrete scalar, then setting the adaptive thermal conductivity strategy to global heat dissipation processing, wherein global heat dissipation processing refers to performing global heat dissipation processing on all areas of the heat source area according to a first optimized thermal conductivity scheme; if the high heat distribution discrete coefficient is greater than or equal to a predetermined high heat discrete scalar and / or the temperature rise distribution discrete coefficient is greater than or equal to a predetermined temperature rise discrete scalar, then setting the adaptive thermal conductivity strategy to local heat dissipation processing, wherein local heat dissipation processing refers to performing local heat dissipation processing on multiple high heat areas of the heat source area according to a second optimized thermal conductivity scheme.

[0053] Furthermore, the thermal conductivity control device for the vapor chamber heat sink module is also used to: construct a twin thermal conductivity space between the vapor chamber heat sink module and the heat source area, and input thermal conductivity control parameters and thermal conductivity control mechanism into the twin thermal conductivity space, wherein the thermal conductivity control parameters include fluid flow path, fluid reflux velocity and fluid phase change rate; obtain multiple temperature overflow values ​​of the multiple high-heat areas, and select the maximum temperature overflow value, wherein the temperature overflow value refers to the difference between the temperature monitoring value of the high-heat area and the predetermined temperature threshold value; obtain multiple temperature rise values ​​of the multiple abnormal temperature rise areas The overflow value is calculated, and the temperature rise overflow mean is calculated, wherein the temperature rise overflow value refers to the difference between the predicted temperature rise rate of the abnormal temperature rise area and the predetermined temperature rise rate threshold value; the maximum temperature overflow value is compensated for the temperature rise according to the temperature rise overflow mean, and the global temperature overflow value is obtained; based on the twin thermal conductivity space, the thermal conductivity control parameter optimization analysis is performed to eliminate the global temperature overflow value as a constraint, and the first optimized thermal conductivity scheme is generated, wherein the first optimized thermal conductivity scheme is the thermal conductivity scheme with the maximum thermal conductivity fitness, and the thermal conductivity fitness is negatively correlated with the heat dissipation time and energy consumption.

[0054] Furthermore, the thermal conductivity control device for the temperature vapor chamber heat sink module is also used to: perform temperature rise compensation on multiple temperature overflow values ​​of the multiple high-heat areas according to the temperature rise overflow mean value to obtain multiple local temperature overflow values; based on the twin thermal conductivity space, perform thermal conductivity control parameter optimization analysis to eliminate the multiple local temperature overflow values ​​as constraints and generate the second optimized thermal conductivity scheme.

[0055] Each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The thermal conductivity control method and specific examples for a temperature vaporizer heat sink module in the aforementioned embodiment 1 are also applicable to a thermal conductivity control device for a temperature vaporizer heat sink module in this embodiment. Through the aforementioned detailed description of a thermal conductivity control method for a temperature vaporizer heat sink module, those skilled in the art can clearly understand the thermal conductivity control device for a temperature vaporizer heat sink module in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method description.

[0056] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0057] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. A heat conduction control method for a heat spreader module, characterized in that: include: The temperature monitoring array deployed on the server collects the real-time temperature of the heat source area and obtains the real-time heat source temperature distribution; Monitor and obtain real-time operating parameters of the heat source area, analyze the heat load changes within a predetermined time window based on the real-time operating parameters, and generate a predicted temperature rise rate distribution; Performing regional discrete evaluation on the real-time heat source temperature distribution and the predicted temperature rise rate distribution, respectively, to obtain a high heat distribution discrete coefficient and a temperature rise distribution discrete coefficient; An adaptive heat conduction strategy is set based on the high heat distribution dispersion coefficient and the temperature rise distribution dispersion coefficient, and according to the adaptive heat conduction strategy, the vapor chamber heat sink module is controlled to perform heat dissipation control of the heat source area.

2. The heat conduction control method for a vapor chamber heat sink module according to claim 1, characterized in that: The temperature monitoring array includes N temperature sensors for collecting temperature of N predetermined divided areas of the heat source area, wherein each temperature sensor corresponds to one predetermined divided area, and N is an integer greater than 5.

3. The heat conduction control method for a vapor chamber heat sink module according to claim 2, characterized in that: Monitor and obtain real-time operating parameters of the heat source area, perform heat load change analysis within a predetermined time window based on the real-time operating parameters, and generate a predicted temperature rise rate distribution, including: Constructing N temperature rise prediction branches based on the BP neural network, wherein the temperature rise prediction branches correspond to the predetermined divided areas one by one; Collect a sample operation parameter set and a plurality of sample temperature rise rate sets, take the sample operation parameter as input and the sample temperature rise rate set as output, use the sample operation parameter set and the plurality of sample temperature rise rate sets to perform supervised training on the N temperature rise prediction branches, obtain N converged temperature rise prediction branches, and build a temperature rise predictor; The real-time operating parameters are input into the temperature rise predictor, N predicted temperature rise rates are output, and the predicted temperature rise rate distribution is generated in combination with the N predetermined divided areas.

4. The heat conduction control method for a vapor chamber heat sink module according to claim 3, characterized in that: Collect sample run parameter sets and multiple sample temperature rise rate sets, including: Taking the attribute characteristics of the heat source area as a constraint, multiple sample operating parameters are retrieved and obtained, a sample operating parameter set is constructed, and the first sample operating parameter is randomly selected; Taking the attribute characteristics of the heat source area and the first sample operation parameters as constraints, retrieve and obtain a sample temperature rise rate distribution set within a historical predetermined time window; The mode of the sample temperature rise rate distribution set is selected as the first sample temperature rise rate distribution, and the first sample temperature rise rate distribution is divided and marked according to the N predetermined division areas to obtain a first sample temperature rise rate set, which is added to the multiple sample temperature rise rate sets, wherein the sample operation parameters and the sample temperature rise rate set have a corresponding relationship.

5. The heat conduction control method for a vapor chamber heat sink module according to claim 1, characterized in that: The real-time heat source temperature distribution and the predicted temperature rise rate distribution are respectively subjected to regional discrete evaluation to obtain a high heat distribution discrete coefficient and a temperature rise distribution discrete coefficient, including: Screening the real-time heat source temperature distribution according to a predetermined temperature threshold value, marking a predetermined divided area with a temperature greater than the predetermined temperature threshold value as a high-heat area, and determining a plurality of high-heat areas; The predicted temperature rise rate distribution is screened according to a predetermined temperature rise rate threshold value, and a predetermined divided area greater than the predetermined temperature rise rate threshold value is marked as an abnormal temperature rise area, and a plurality of abnormal temperature rise areas are determined; The multiple high heat areas and the multiple abnormal temperature rise areas are respectively evaluated for regional distribution dispersion to obtain the high heat distribution dispersion coefficient and the temperature rise distribution dispersion coefficient.

6. The heat conduction control method for a vapor chamber heat sink module according to claim 5, characterized in that: The adaptive heat conduction strategy is set based on the high heat distribution dispersion coefficient and the temperature rise distribution dispersion coefficient, including: If the high heat distribution dispersion coefficient is less than a predetermined high heat dispersion scalar and the temperature rise distribution dispersion coefficient is less than a predetermined temperature rise dispersion scalar, the adaptive heat conduction strategy is set to global heat dissipation processing, wherein the global heat dissipation processing refers to performing global heat dissipation processing on all areas of the heat source area according to the first optimized heat conduction scheme; If the high heat distribution discrete coefficient is greater than or equal to a predetermined high heat discrete scalar and / or the temperature rise distribution discrete coefficient is greater than or equal to a predetermined temperature rise discrete scalar, the adaptive thermal conductivity strategy is set to local heat dissipation treatment, wherein the local heat dissipation treatment refers to performing local heat dissipation treatment on multiple high heat areas in the heat source area according to the second optimized thermal conductivity scheme.

7. The heat conduction control method for a vapor chamber heat sink module according to claim 6, characterized in that: Generate a first optimized thermal conductivity solution, including: Constructing a twin thermal conduction space of the vapor chamber heat sink module and the heat source area, and inputting thermal conduction control parameters and thermal conduction control mechanisms into the twin thermal conduction space, wherein the thermal conduction control parameters include a fluid flow path, a fluid reflux velocity, and a fluid phase change rate; Acquire multiple temperature overflow values ​​of the multiple high-heat areas, and select a maximum temperature overflow value, wherein the temperature overflow value refers to the difference between the temperature monitoring value of the high-heat area and a predetermined temperature threshold value; Acquire multiple temperature rise overflow values ​​of the multiple abnormal temperature rise areas, and calculate the temperature rise overflow average, wherein the temperature rise overflow value refers to the difference between the predicted temperature rise rate of the abnormal temperature rise area and the predetermined temperature rise rate threshold value; Performing temperature rise compensation on the maximum temperature overflow value according to the temperature rise overflow mean value to obtain a global temperature overflow value; Based on the twin heat conduction space, a heat conduction control parameter optimization analysis is performed to eliminate the global temperature overflow value as a constraint, and the first optimized heat conduction scheme is generated, wherein the first optimized heat conduction scheme is a heat conduction scheme with the maximum heat conduction adaptability, and the heat conduction adaptability is negatively correlated with the heat dissipation time and energy consumption.

8. The heat conduction control method for a vapor chamber heat sink module according to claim 7, characterized in that: Generate a second optimized thermal conductivity solution, including: Performing temperature rise compensation on a plurality of temperature overflow values ​​of the plurality of high-heat areas according to the temperature rise overflow mean value to obtain a plurality of local temperature overflow values; Based on the twin heat conduction space, a heat conduction control parameter optimization analysis is performed to eliminate the multiple local temperature overflow values ​​as a constraint, and the second optimized heat conduction solution is generated.

9. A heat conduction control device for a temperature vapor chamber heat sink module, characterized in that: The steps for implementing the thermal conductivity control method for a vapor chamber heat sink module according to any one of claims 1 to 8 include: The heat source temperature distribution acquisition module is used to collect the real-time temperature of the heat source area through the temperature monitoring array deployed on the server to obtain the real-time heat source temperature distribution; A heat load change analysis module is used to monitor and obtain real-time operating parameters of the heat source area, perform heat load change analysis within a predetermined time window according to the real-time operating parameters, and generate a predicted temperature rise rate distribution; A regional discrete evaluation module is used to perform regional discrete evaluation on the real-time heat source temperature distribution and the predicted temperature rise rate distribution, respectively, to obtain a high heat distribution discrete coefficient and a temperature rise distribution discrete coefficient; The adaptive thermal conductivity strategy setting module is used to set the adaptive thermal conductivity strategy based on the high heat distribution discrete coefficient and the temperature rise distribution discrete coefficient, and control the temperature vapor chamber radiator module to perform heat dissipation control of the heat source area according to the adaptive thermal conductivity strategy.

Images