Control method and device

Abstract

The invention provides a control method and device which can be applied to the technical field of server heat dissipation. The control method comprises the following steps: processing a plurality of pieces of obtained initial temperature data of a plurality of detection points in at least one power supply unit, and determining target temperature data in the plurality of pieces of initial temperature data; processing the target temperature data, the environmental data of the power supply unit and the related load data to obtain temperature features, environmental features and load features for the plurality of fluctuation types; obtaining a temperature prediction result of the power supply unit at the later moment based on the temperature characteristics, the environment characteristics and the load characteristics; and controlling the flow of the liquid cooling device in the power supply unit based on the temperature prediction result.

Description

Technical Field

[0001] This application relates to the technical field of server cooling, and particularly to a control method and device. Background Art

[0002] Servers and their components (such as power supply units) generate a large amount of heat during operation. Good heat dissipation can enable the server to operate at an appropriate temperature, avoiding performance degradation or hardware failures caused by overheating. In related technologies, temperature sensors, basic input / output systems, and server management software are mainly used to monitor the temperature of servers and their components in real time, and corresponding temperature alarm functions are set. However, the above methods have time lag and are limited to simple temperature monitoring and alarm, making it difficult to meet the actual monitoring requirements. Summary of the Invention

[0003] In view of the above problems, this application provides a control method, device, equipment, medium, and program product.

[0004] According to the first aspect of this application, a control method is provided, including: processing multiple initial temperature data of multiple detection points in at least one power supply unit obtained, to determine target temperature data among the multiple initial temperature data; respectively processing the target temperature data, the environmental data of the power supply unit, and the relevant load data, to obtain temperature characteristics, environmental characteristics, and load characteristics for multiple fluctuation types; based on the temperature characteristics, environmental characteristics, and load characteristics, obtaining a temperature prediction result of the power supply unit at a later time; controlling the flow rate of the liquid cooling device in the power supply unit based on the temperature prediction result.

[0005] The second aspect of this application provides a control device, including: a first data processing module, configured to process multiple initial temperature data of multiple detection points in at least one power supply unit obtained, to determine target temperature data among the multiple initial temperature data; a second data processing module, configured to respectively process the target temperature data, the environmental data of the power supply unit, and the relevant load data, to obtain temperature characteristics, environmental characteristics, and load characteristics for multiple fluctuation types; a result obtaining module, configured to obtain a temperature prediction result of the power supply unit at a later time based on the temperature characteristics, environmental characteristics, and load characteristics; a control module, configured to control the flow rate of the liquid cooling device in the power supply unit based on the temperature prediction result.

[0006] The third aspect of this application provides an electronic device, including: one or more processors; a memory, configured to store one or more computer programs, wherein the above one or more processors execute the above one or more computer programs to implement the steps of the above method.

[0007] The fourth aspect of the present application further provides a computer-readable storage medium, on which a computer program or instructions are stored, and when the computer program or instructions are executed by a processor, the steps of the above method are implemented.

[0008] The fifth aspect of the present application further provides a computer program product, including a computer program or instructions, and when the computer program or instructions are executed by a processor, the steps of the above method are implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Through the following description of the embodiments of the present application with reference to the drawings, the above content and other objects, features, and advantages of the present application will become clearer. In the drawings:

[0010] Figure 1A Schematically shows an application scenario diagram of a control method, device, equipment, medium, and program product according to an embodiment of the present application;

[0011] Figure 1B Schematically shows a top view of a liquid-cooled power supply rack according to an embodiment of the present application;

[0012] Figure 1C Schematically shows a blind plug-in schematic diagram of a cabinet front direct current input bus of a liquid-cooled power supply rack according to an embodiment of the present application;

[0013] Figure 2 Schematically shows a flowchart of a control method according to an embodiment of the present application;

[0014] Figure 3 Schematically shows a flowchart of another control method according to an embodiment of the present application;

[0015] Figure 4 Schematically shows a structural block diagram of a control device according to an embodiment of the present application;

[0016] Figure 5 Schematically shows a block diagram of an electronic device suitable for implementing a control method according to an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] Hereinafter, embodiments of the present application will be described with reference to the drawings. However, it should be understood that these descriptions are exemplary and are not intended to limit the scope of the present application. In the following detailed description, for the sake of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present application. However, it is obvious that one or more embodiments can also be implemented without these specific details. In addition, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessarily confusing the concepts of the present application.

[0018] The terms used herein are merely for describing specific embodiments and are not intended to limit the present application. The terms "including", "comprising" and the like used herein indicate the presence of the described features, steps, operations and / or components, but do not exclude the presence or addition of one or more other features, steps, operations or components.

[0019] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0020] In the case of using expressions such as "at least one of A, B, and C", generally, it should be interpreted according to the meaning commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include, but is not limited to, a system having only A, only B, only C, having A and B, having A and C, having B and C, and / or having A, B, and C, etc.).

[0021] In the technical solution of the present application, the user information involved (including but not limited to user personal information, user image information, user device information, such as location information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) are all information and data authorized by the user or fully authorized by all parties. And the processing of the relevant data, such as collection, storage, use, processing, transmission, provision, disclosure and application, etc., all comply with relevant laws, regulations and standards, take necessary confidentiality measures, do not violate public order and good customs, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0022] In some examples, a temperature monitoring function is provided by using server management software (such as built-in management software and third-party management software) to detect the temperature of the server power supply. For example, various parameters of the server and its components can be displayed through a dashboard, and an alarm notification is sent when the temperature exceeds a set threshold.

[0023] However, the sampling rate of some management software is low and cannot track the rapid change of temperature in real time. The management software usually can only issue an alarm after the temperature exceeds the set threshold and cannot predict potential temperature problems in advance.

[0024] In some examples, a temperature monitoring function is provided in the settings of the Basic Input / Output System (BIOS) of some servers. By entering the BIOS setup interface of the server, the monitoring option can be found and the corresponding temperature alarm threshold can be set. When the power supply temperature exceeds the threshold, the server will send an alarm signal.

[0025] However, the BIOS depends on the performance of hardware sensors. If the sensors are poorly connected or have insufficient accuracy, the BIOS cannot provide accurate monitoring either. Moreover, the alarm function of the BIOS is usually relatively simple, only issuing an alarm when the temperature exceeds the set threshold, and it cannot provide a detailed analysis of the temperature change trend.

[0026] Based on the above problems, the present application provides a control method, including: processing multiple initial temperature data of multiple detection points in at least one power supply unit obtained, and determining the target temperature data among the multiple initial temperature data; respectively processing the target temperature data, the environmental data of the power supply unit, and the relevant load data to obtain temperature characteristics, environmental characteristics, and load characteristics for multiple fluctuation types; based on the temperature characteristics, environmental characteristics, and load characteristics, obtaining the temperature prediction result of the power supply unit at a later time; and controlling the flow rate of the liquid cooling device in the power supply unit based on the temperature prediction result.

[0027] According to the embodiments of the present application, by processing temperature data, environmental data, and load data for different fluctuation types, temperature characteristics, environmental characteristics, and load characteristics on different time scales are obtained, improving the flexibility of processing different types of data. Since the temperature prediction result of the power supply unit at a later time is predicted in advance based on multi-dimensional characteristic data at the current time, the flow rate control process of the liquid cooling device is made more scientific and reasonable, providing more comprehensive decision-making support for the operation and maintenance of the power supply unit, optimizing resource utilization, and reducing the operating cost of the system.

[0028] Figure 1A The application scenario diagrams of the control method, device, equipment, medium, and program product according to the embodiments of the present application are schematically shown.

[0029] As Figure 1A shown, the application scenario according to this embodiment may include a terminal device 101, a network 102, and a server 103. The network 102 is used to provide a medium for a communication link between the terminal device 101 and the server 103. The network 102 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0030] Users can use the terminal device 101 to interact with the server 103 through the network 102 to receive or send messages, etc. For example, users use the terminal device 101 to send requests for temperature prediction or setting alarm thresholds to the server 103. The terminal device 101 may be various electronic devices with a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop portable computers, and desktop computers, etc.

[0031] Server 103 can be a server that provides various services. For example, server 103 obtains relevant data from a database or a real-time data source, such as historical temperature data, environmental data, and server load data.

[0032] For example, server 103 sends the prediction results and alarm information back to the terminal device 101 through the network 102. The terminal device 101 receives the prediction results and alarm information from server 103.

[0033] For example, the terminal device 101 displays the prediction results and alarm information, and the user can take corresponding measures based on this information. The user can perform further operations according to the result display, such as adjusting the alarm threshold or requesting a new prediction.

[0034] It should be noted that the control method provided in the embodiments of the present application can generally be executed by server 103. Correspondingly, the control device provided in the embodiments of the present application can generally be set in server 103. The control method provided in the embodiments of the present application can also be executed by a server or a server cluster different from server 103 and capable of communicating with the terminal device 101 and / or server 103. Correspondingly, the control device provided in the embodiments of the present application can also be set in a server or a server cluster different from server 103 and capable of communicating with the terminal device 101 and / or server 103.

[0035] It should be understood that Figure 1A the numbers of the terminal devices, networks, and servers in

[0036] Figure 1B The top view of the liquid-cooled power supply rack according to the embodiment of the present application is schematically shown.

[0037] As Figure 1B shown, the liquid-cooled power supply rack can be a carrier for multiple power supply units 111 in a whole cabinet server, and can be used to realize the transfer between the power supply unit 111 and the cabinet power busbar 112 and the management of the power supply unit 111. Multiple detection points are set in each power supply unit 111, and the detection points can obtain the temperature data of the power supply unit 111 in real time. The liquid-cooled power supply rack 110 can also be used for the heat dissipation realization and management of the power supply unit 111, and has an input blind plug design.

[0038] The liquid-cooled power supply rack integrates a liquid-cooled water inlet 113, a liquid-cooled water outlet 114, and a DC output interface 115 at the rear window, all of which support blind insertion. The cooling refrigerant can enter from the liquid-cooled water inlet 113. Through a first water distributor 116 with a solenoid valve, the water flow rate of the liquid-cooled device in each power supply unit 111 can be independently controlled, thereby realizing independent heat dissipation management for each power supply unit 111. The cooling refrigerant flowing out of each power supply unit 111 converges to the liquid-cooled water outlet 114 through the second water distributor 117, and the liquid-cooled water outlet 114 and the whole cabinet server form a heat dissipation cycle.

[0039] Figure 1C Schematically shows a blind insertion schematic diagram of the front DC input bus of the cabinet of the liquid-cooled power supply rack according to an embodiment of the present application.

[0040] The power density of the liquid-cooled power supply rack is relatively high, and the input current can reach 1600A. To reduce the internal space occupied by the input cables, the input bus 118 can be designed on both sides of the front window of the cabinet. As Figure 1C shown in the perspective view, connection blind pluggers 119 are provided on both sides of the liquid-cooled power supply rack 110 to achieve blind insertion power taking with the DC input bus contacts 120 in Figure 1B . Inside the two sides of the liquid-cooled power supply rack, the input cables connect the blind plug connectors 119 to the power transfer board 121 at the rear end of the chassis as shown in Figure 1B . It should be noted that the power management module 122 can independently execute the control method provided by the embodiments of the present application, and the control method provided by the embodiments of the present application can also be executed by a server 103 different from the power management module 122 and capable of communicating with the power management module 122 as shown in Figure 1A to realize the water and electricity control of the liquid-cooled power supply rack.

[0041] Figure 2 Schematically shows a flowchart of the control method according to an embodiment of the present application.

[0042] As Figure 2 shown, the control method of this embodiment includes operations S210 to S240.

[0043] In operation S210, process the multiple initial temperature data of multiple detection points in at least one power supply unit obtained to determine the target temperature data among the multiple initial temperature data.

[0044] In an embodiment of the present application, the power supply unit can convert an external power supply into a stable DC power supply required by components inside the server. As one of the core components of the whole cabinet server, multiple power supply units can be set according to actual requirements. The detection points can be the positions of multiple temperature sensors arranged inside the power supply unit. The initial temperature data can be the temperature data detected in real time by the sensors. The target temperature data is the temperature data that meets the preset conditions after processing the initial temperature data.

[0045] For example, the target temperature data can be determined based on the weight value of the target detection point among multiple detection points and the initial temperature data. For example, by comparing the initial temperature data between the target detection point and other detection points, the weight value of the target detection point is determined, and then based on the initial temperature data, the weight value, and the preset conditions, the target temperature data is obtained.

[0046] In operation S220, the target temperature data, the environmental data of the power supply unit, and the relevant load data are processed respectively to obtain temperature characteristics, environmental characteristics, and load characteristics for multiple fluctuation types.

[0047] In an embodiment of the present application, the environmental data can characterize the external environmental state of the power supply unit, including temperature data and humidity data. The relevant load data can characterize the load data of different components in the server, including processor load, memory load, disk load, network load, etc. The fluctuation types can be determined according to the change characteristics and change trends of the time series data. The temperature characteristics, environmental characteristics, and load characteristics can be the characteristic representations of different fluctuation types obtained after feature extraction of the target temperature data, environmental data, and relevant load data. It can be understood that the target temperature data, environmental data, and relevant load data are time series data.

[0048] For example, sliding windows of different lengths are used to perform feature extraction on the temperature data, environmental data, and relevant load data respectively to obtain temperature characteristics, environmental characteristics, and load characteristics of different fluctuation types.

[0049] In operation S230, based on the temperature characteristics, environmental characteristics, and load characteristics, a temperature prediction result of the power supply unit at a later time is obtained.

[0050] In an embodiment of the present application, the temperature prediction result can be the temperature result of the power supply unit at a future moment predicted according to the characteristics of different fluctuation types obtained at the current moment.

[0051] For example, the same features of different fluctuation types can be concatenated first to obtain the concatenated features; then, the concatenated features are weighted with their respective corresponding weights to obtain the temperature prediction result. Alternatively, the features of different fluctuation types are weighted with their respective corresponding weights first to obtain the weighted features; then, the weighted features are fused to obtain the temperature prediction result.

[0052] In operation S240, the flow rate of the liquid cooling device in the power supply unit is controlled based on the temperature prediction result.

[0053] In an embodiment of the present application, the liquid cooling device can be used to cool the power supply unit so that the temperature of the power supply unit is maintained within a reasonable temperature range. The flow rate can be the flow rate of the heat dissipation refrigerant in the liquid cooling device, which can be controlled by the opening degree of the flow valve at the water inlet of the liquid cooling device.

[0054] In a feasible embodiment, the above control method can be implemented using a temperature prediction model. The temperature prediction model can be a hybrid model constructed based on a one-dimensional convolutional sub-model and a long short-term memory network sub-model. The one-dimensional convolutional sub-model can be used to perform feature extraction processing on target temperature data, environmental data, and related load data, and the long short-term memory network sub-model can be used to process the extracted features in each dimension to predict the temperature result. The obtained temperature features, environmental features, and load features can be input into the pre-trained temperature prediction model to output the temperature prediction result.

[0055] According to the embodiment of the present application, by processing the temperature data, environmental data, and load data for different fluctuation types, temperature features, environmental features, and load features at different time scales are obtained, improving the flexibility of processing different types of data. Since the temperature prediction result of the power supply unit at a later time is predicted in advance based on the multi-dimensional feature data at the current time, the flow control process of the liquid cooling device is made more scientific and reasonable, providing more comprehensive decision support for the operation and maintenance of the power supply unit, optimizing resource utilization, and reducing the operating cost of the system.

[0056] It can be understood that how to control the flow rate of the liquid cooling device has been described above, and below will describe how to obtain the features of different fluctuation types.

[0057] According to an embodiment of the present application, the fluctuation types include short-term fluctuation types and long-term fluctuation types; the target temperature data, the environmental data of the power supply unit, and the relevant load data are processed respectively to obtain temperature characteristics, environmental characteristics, and load characteristics for multiple fluctuation types, including: performing first feature extraction on the target temperature data, environmental data, and relevant load data respectively based on a first sliding window for short-term fluctuation types to obtain a first temperature characteristic, a first environmental characteristic, and a first load characteristic; performing second feature extraction on the target temperature data, environmental data, and relevant load data respectively based on a second sliding window for long-term fluctuation types to obtain a second temperature characteristic, a second environmental characteristic, and a second load characteristic.

[0058] In an embodiment of the present application, the first sliding window can capture short-term fluctuations in the data, and the short-term fluctuations can reflect the rapid change characteristics of the data in a relatively short period of time. The second sliding window can capture long-term fluctuations in the data, and the long-term fluctuations can reflect the change characteristics of the data over a long time span. It can be understood that the first temperature characteristic, the first environmental characteristic, and the first load characteristic are short-term fluctuation characteristics, and the second temperature characteristic, the second environmental characteristic, and the second load characteristic are long-term fluctuation characteristics.

[0059] Taking the extraction of local features of data by using a one-dimensional convolutional sub-model in a temperature prediction model as an example. For example, the target temperature data, the environmental data of the power supply unit, and the relevant load data are input into the one-dimensional convolutional sub-model, and a small convolutional kernel (for example, the convolutional kernel size is 3) is slid on the data, covering a small part of the data each time to extract short-term fluctuations in the data, obtaining a first temperature characteristic, a first environmental characteristic, and a first load characteristic.

[0060] For example, a large convolutional kernel (for example, the convolutional kernel size is 10) in the one-dimensional convolutional sub-model is slid on the data, covering a larger part of the data each time to extract long-term fluctuations in the data, and by spanning multiple time steps, extracting the long-term trend in the data, obtaining a second temperature characteristic, a second environmental characteristic, and a second load characteristic.

[0061] It can be understood that the small convolutional kernel can detect the rapid changes in the target temperature data within a few minutes, and then judge whether it is caused by a sudden increase in load or a rapid change in environmental conditions. The large convolutional kernel can identify the periodic changes in the target temperature data within a day, and then judge whether the change is caused by the day-night temperature difference or the daily change in server load.

[0062] According to an embodiment of the present application, by combining different sliding windows, short-term fluctuations and long-term fluctuations in the data can be captured simultaneously, thereby providing a more comprehensive feature representation for subsequent temperature prediction. Especially for complex time series data, features on different time scales can be extracted to achieve a better understanding and prediction of future temperature changes.

[0063] According to an embodiment of the present application, based on temperature characteristics, environmental characteristics, and load characteristics, a temperature prediction result of the power supply unit at a later time is obtained, including: splicing the first temperature characteristic with the second temperature characteristic, the first environmental characteristic with the second environmental characteristic, and the first load characteristic with the second load characteristic respectively to obtain a spliced temperature characteristic, a spliced environmental characteristic, and a spliced load characteristic; and obtaining a temperature prediction result based on the spliced temperature characteristic, the spliced environmental characteristic, the spliced load characteristic, and their respective weights.

[0064] In an embodiment of the present application, after obtaining the characteristics of different fluctuation types, the same characteristics of different fluctuation types can be fused. By utilizing the change characteristics of different time scales, a more comprehensive feature representation is provided to improve the accuracy of subsequent temperature prediction.

[0065] In an embodiment of the present application, splicing can be performed between different characteristics in a simple connection manner to obtain a spliced temperature characteristic, a spliced environmental characteristic, and a spliced load characteristic. The weights corresponding to different spliced characteristics can be determined according to the importance of the characteristics or variance analysis.

[0066] For example, the volatility of each characteristic is determined through variance analysis, and the characteristic with greater volatility has a greater weight.

[0067] In an embodiment of the present application, after obtaining the weights of different spliced characteristics, the spliced characteristics can be weighted averaged with the corresponding weights to obtain a temperature prediction result.

[0068] For example, the attention scores of short-term fluctuation characteristics and long-term fluctuation characteristics can be calculated respectively to determine the weights corresponding to each characteristic, so as to perform weighted summation on the short-term fluctuation characteristics and long-term fluctuation characteristics to obtain a temperature prediction result.

[0069] According to an embodiment of the present application, based on the spliced temperature characteristic, the spliced environmental characteristic, the spliced load characteristic, and their respective corresponding weights, a temperature prediction result is obtained, including: respectively weighting the spliced temperature characteristic, the spliced environmental characteristic, and the spliced load characteristic by using a preset temperature weight, a preset environmental weight, and a preset load weight to obtain an intermediate weighted result; and obtaining a temperature prediction result according to the intermediate weighted result and the state characteristic at the previous time.

[0070] In an embodiment of the present application, the intermediate weighted result can be obtained by weighting each spliced characteristic with its respective weight. The intermediate weighted result can include an intermediate temperature weighted result, an intermediate environmental weighted result, and an intermediate load weighted result based on different data types. The state characteristic can be the state data of the intermediate weighted result at the previous moment.

[0071] Taking the prediction of the temperature of the power supply unit using the long short-term memory network sub-model as an example, for instance, the spliced temperature feature, spliced environmental feature, and spliced load feature are input into the long short-term memory network sub-model. Based on the gating mechanism, the spliced temperature feature, spliced environmental feature, spliced load feature, and corresponding weights are used to obtain the intermediate temperature weighted result, intermediate environmental weighted result, and intermediate load weighted result. Furthermore, according to the intermediate weighted result and the state feature at the previous moment, the temperature prediction result of the power supply unit at the next moment is output.

[0072] For example, in the long short-term memory network sub-model, each time step can receive an input and the cell state (state feature) of the previous time step. Thus, through the control of the forget gate, input gate, and output gate, the long short-term memory network sub-model can determine which information is forgotten, which new information can be added, and which information is output to the hidden state to output the temperature prediction result of the power supply unit at the next moment.

[0073] According to the embodiments of the present application, through the state feature at the previous moment and the gating mechanism, the temperature prediction model can better capture the long-term dependencies in the sequence data. For different types of fluctuation features (short-term fluctuation feature and long-term fluctuation feature), by introducing different time steps, the requirement of processing data with multiple time scales in the actual prediction process is satisfied, and the flexibility and applicability of the control method are improved.

[0074] According to the embodiments of the present application, based on the temperature feature, environmental feature, and load feature, the temperature prediction result of the power supply unit at the next moment is obtained, including: performing a first operation on the first temperature feature, first environmental feature, and first load feature based on the first gating algorithm and the first time step to obtain a first operation result, and performing a second operation on the second temperature feature, second environmental feature, and second load feature based on the second gating algorithm and the second time step to obtain a second operation result, where the first time step is less than the second time step; splicing the first operation result and the second operation result to obtain the temperature prediction result.

[0075] In the embodiments of the present application, the first operation result can be the result obtained by using the first gating algorithm for the short-term fluctuation feature; the second operation result can be the result obtained by using the second gating algorithm for the long-term fluctuation feature. The time length corresponding to the first time step is less than the length corresponding to the second time step.

[0076] In the embodiments of the present application, the first operation result and the second operation result can be spliced using at least one of direct splicing, weighted splicing, attention mechanism splicing, and feature fusion to obtain the temperature prediction result.

[0077] For example, after obtaining the first operation result and the second operation result, the first operation result and the second operation result can be weighted and spliced to obtain an intermediate operation result, and then the intermediate operation result can be subjected to multiple non-linear transformations to obtain a transformed feature representation, and a temperature prediction result can be obtained.

[0078] According to an embodiment of the present application, by splicing and fusing the first operation result and the second operation result in a higher dimension during the gating algorithm process, the dynamic characteristics in the data can be captured more comprehensively, the expression ability of the features can be improved, and the accuracy of the temperature prediction result can be further enhanced.

[0079] According to an embodiment of the present application, processing multiple initial temperature data of multiple detection points in at least one power supply unit obtained, and determining target temperature data among the multiple initial temperature data includes: determining a weight value of a target detection point among the multiple detection points; and determining the target temperature data from the multiple initial temperature data based on the weight value and the initial temperature data.

[0080] In an embodiment of the present application, the target detection point can be any detection point determined from the multiple detection points according to actual needs. The weight value can be determined according to the initial temperature data of different detection points. The importance degree of the detection point and the credibility of the initial temperature data are directly proportional to the weight value. The target temperature data can be temperature data that meets the requirements of data quality and data quantity determined from the initial temperature data.

[0081] For example, a power supply device includes 5 detection points (denoted as T1, T2, T3, T4, and T5), and corresponding weight values are set for the importance degrees of the 5 detection points. For example, higher weight values are set for the detection points closest to the water inlet and the water outlet. Or, if the matching degree between the initial temperature data of T3 and the initial temperature data of T1, T2, T4, and T5 is the highest, the weight value of T3 can be set to the maximum.

[0082] In the related art, the number of set detection points is small, and in the case where the temperature sensor is abnormal, basic data that meets the requirements cannot be obtained in time. To solve this technical problem, the present application sets multiple temperature monitoring points to increase the data volume of the initial temperature data, so that the initial temperature data of multiple detection points can be used to verify abnormal detection data and meet subsequent prediction requirements.

[0083] It can be understood that how to determine the target temperature data has been described above, and below, how to determine the weight value of the target point will be described.

[0084] According to an embodiment of the present application, determining the weight value of a target detection point among a plurality of detection points includes: determining the support degree between the initial temperature data of the target detection point and the initial temperature data of other detection points; determining the weight value of the target detection point based on the support degree, and the weight value is directly proportional to the support degree.

[0085] In an embodiment of the present application, the support degree between different initial temperature data can be determined by using at least one of correlation analysis and distance metric. Assume that the initial temperature data of the target detection point and other detection points follow a certain probability distribution (such as Gaussian distribution), and then calculate the confidence distance between the target detection point and other detection points, and the confidence distance is inversely proportional to the support degree.

[0086] For example, a support degree matrix can be constructed, and each element in the matrix represents the support degree of one detection point to another detection point. The support degree matrix can be constructed by calculating the similarity or correlation between the initial temperature data of different detection points. Or, by calculating the eigenvalues and eigenvectors of the support degree matrix, the main support features between detection points can be extracted, and the magnitude of the eigenvalue represents the support strength in the direction of the corresponding eigenvector, so as to determine the support relationship between sensors. It can be understood that the above methods for determining the support degree can be used alone or in combination to more accurately determine the support degree between different detection points.

[0087] According to an embodiment of the present application, using data fusion technology, the initial temperature data of a plurality of detection points are comprehensively processed to improve the accuracy and reliability of measurement. During the data fusion process, the support degree and weight between them can also be determined according to factors such as the accuracy and stability of the temperature sensors. During the data fusion process, reasonable weights are assigned to the data of different sensors according to the support degree, which can more accurately reflect the actual temperature situation. In a multi-sensor system, even if some sensors fail or the data is abnormal, by considering the support degree, the system can still rely on the data of other reliable sensors to maintain normal operation, improving the fault tolerance and reliability of the system, ensuring that the temperature monitoring system can work stably in complex and harsh environments, and providing continuous and accurate data support for subsequent temperature prediction.

[0088] According to an embodiment of the present application, based on the weight value and the initial temperature data, determining the target temperature data from a plurality of initial temperature data includes: determining a target value corresponding to the target detection point based on the initial temperature data and the weight value; in the case where the target value is greater than or equal to a preset threshold, determining the initial temperature data of the target detection point as the target temperature data.

[0089] In an embodiment of the present application, the target value may be the average or standard deviation of the initial temperature data of the target detection point. The preset threshold can be used to determine whether the initial temperature data meets the data quality requirements, and can be specifically determined according to the actual situation, which is not limited here.

[0090] For example, under the same temperature conditions, multiple measurements are taken using temperature sensors of the same model at multiple detection points to obtain a set of initial temperature data. Calculate statistics such as the average and standard deviation of the measurement data of each sensor, and compare them with the true value of the standard temperature source to evaluate the measurement error and accuracy of the sensor. The sensors at the target detection points with smaller standard deviations have relatively higher measurement accuracies, and thus the initial temperature data obtained by them is determined as the target temperature data.

[0091] According to the embodiment of the present application, the analysis of support can help detect the faults of the detection point sensors in a timely manner. For example, in a temperature monitoring system, if the support of the data of a temperature sensor is continuously low compared with that of other sensors, the system can automatically issue an alarm to remind the maintenance personnel to check the sensor. When the support of the data of a certain sensor suddenly decreases compared with that of other sensors, it may indicate that the sensor has failed or been interfered with, and measures should be taken in a timely manner for repair or replacement to avoid the measurement errors caused by sensor failures from affecting the normal operation of the entire system.

[0092] Figure 3 Schematically shows a flowchart of another control method according to an embodiment of the present application.

[0093] As Figure 3 shown, the power management module can be used to achieve the water and electricity control of the liquid-cooled power supply rack. The water and electricity control method of the liquid-cooled power supply rack may include operations S310 to S370.

[0094] In operation S310, initialize the control system. The initialization method may include: detecting and recording the in-place status of the power supply unit, performing a leak detection of the liquid-cooling device, and reading the data of each leak sensor. Reading the temperatures of the liquid-cooling inlet and outlet, and determining whether the water temperature is normal.

[0095] In operation S320, after the control system initialization is completed, control the power supply unit to power on and operate the liquid-cooling device.

[0096] In operation S330, obtain multiple initial temperature data of multiple detection points in at least one power supply unit in the liquid-cooling device.

[0097] In operation S340, based on the weight value and the initial temperature data of the target detection point among the multiple detection points, determine the target temperature data from the multiple initial temperature data.

[0098] In operation S350, based on the target temperature data, the environmental data of the power supply unit, and the load data of the server, temperature characteristics, environmental characteristics, and load characteristics of multiple fluctuation types are obtained.

[0099] In operation S360, based on the temperature characteristics, environmental characteristics, and load characteristics, a temperature prediction result of the power supply unit at a later time is obtained.

[0100] In operation S370, the flow rate of the liquid cooling device in the power supply unit is controlled according to the temperature prediction result.

[0101] Among them, controlling the flow rate of the liquid cooling device in the power supply unit according to the temperature prediction result may include operations S371 to S373.

[0102] Five temperature detection points, denoted as T1, T2, T3, T4, and T5, can be set inside the power supply unit. By detecting the temperatures of these five detection points inside the power supply unit, the solenoid valve of the liquid cooling temperature adjustment water distributor is judged to control the flow rate of the liquid cooling device.

[0103] In operation S371, keep the solenoid valve of the water distributor closed. When it is predicted that the highest temperature among the temperatures of T1, T2, T3, T4, and T5 is lower than the inlet water temperature, it can indicate that the load of the power supply unit is low, the internal heat generation is small, and liquid cooling is not required at the current moment, and this path of the water distributor remains closed.

[0104] In operation S372, request overheat protection and adjust the opening degree of the solenoid valve of the water distributor. When it is predicted that the highest temperature among the temperatures of T1, T2, T3, T4, and T5 is greater than or equal to the highest temperature warning value, it can indicate that abnormal overheating occurs inside the power supply unit and effective heat dissipation cannot be achieved currently. The control system reports an over-temperature warning and requests overheat protection at the same time.

[0105] The methods of overheat protection may include: querying whether the power supply unit is fully configured. If there are vacant positions in the power supply unit, request to configure more power supply units to balance the load; in the case where the power supply unit is fully configured, query whether the load power reaches the maximum value of the power supply rack. If it reaches the maximum power, request the server host to reduce the service load; in the case where the high-temperature warning persists for a long time without feedback, report to the host to save the service, forcibly shut down some device units, and when the temperature drops to the normal range, restart the above nodes and continue to run from the saved service nodes.

[0106] In operation S373, adjust the opening degree of the solenoid valve according to the real-time predicted temperature. When it is predicted that the temperature of at least one of the detection points T1, T2, T3, T4, and T5 is higher than the inlet water temperature and lower than the highest temperature warning value, open the inlet of the solenoid valve and adjust the opening degree of the solenoid valve according to the real-time predicted temperature.

[0107] According to an embodiment of the present application, the environmental features include temperature features and humidity features at a previous moment; the method further includes: updating the weight value of the target detection point when the fluctuation value between the temperature feature and humidity feature at the current moment and the temperature feature and humidity feature at the previous moment is greater than or equal to a preset fluctuation threshold.

[0108] In an embodiment of the present application, considering that the temperature data and humidity data in the external environment have a greater impact on the temperature prediction result, therefore, the weight values of different detection points can be updated by comparing the humidity features and temperature features at different moments in real time, and then the subsequent temperature result prediction can be performed according to the updated weight values.

[0109] For example, calculate the fluctuation value between the temperature feature and humidity feature at the current moment and the temperature feature and humidity feature at the previous moment; then compare the calculated fluctuation value with the preset fluctuation threshold. If the fluctuation value is greater than or equal to the preset fluctuation threshold, the weight update of the target detection point is triggered. The larger the fluctuation value, the lower the weight, and vice versa. It can be understood that after updating the weights, ensure that the sum of the weights of all detection point sensors is 1 to maintain the balance of the system.

[0110] According to an embodiment of the present application, in the process of multi-sensor data fusion, dynamically adjusting the weights according to the fluctuation value can more accurately reflect the reliability of each sensor in the current environment, thereby optimizing the data fusion result and improving the measurement accuracy of the overall system. By dynamically updating the weight values of the detection points, the system can automatically optimize the sensor combination and data processing method according to the requirements of the specific application scenario, thereby improving the adaptability and performance of the system in different application scenarios.

[0111] Based on the above control method, the present application also provides a control device. The following will be combined with Figure 4 to describe this device in detail.

[0112] Figure 4 Schematically shows a structural block diagram of a control device according to an embodiment of the present application.

[0113] As Figure 4 shown, the control device of this embodiment includes a first data processing module 410, a second data processing module 420, a result obtaining module 430, and a control module 440.

[0114] The first data processing module 410 is configured to process multiple initial temperature data of multiple detection points in at least one power supply unit obtained, and determine the target temperature data among the multiple initial temperature data. In one embodiment, the first data processing module 410 can be configured to perform the operation S210 described above, which will not be elaborated here.

[0115] The second data processing module 420 is configured to process the target temperature data, the environmental data of the power supply unit, and the relevant load data respectively, so as to obtain temperature characteristics, environmental characteristics, and load characteristics for multiple fluctuation types. In one embodiment, the second data processing module 420 may be configured to perform the operation S220 described above, which will not be elaborated here.

[0116] The result obtaining module 430 is configured to obtain the temperature prediction result of the power supply unit at a later time based on the temperature characteristics, environmental characteristics, and load characteristics. In one embodiment, the result obtaining module 430 may be configured to perform the operation S230 described above, which will not be elaborated here.

[0117] The control module 440 is configured to control the flow rate of the liquid cooling device in the power supply unit based on the temperature prediction result. In one embodiment, the control module 440 may be configured to perform the operation S240 described above, which will not be elaborated here.

[0118] According to the embodiments of the present application, based on the first data processing module 410, the second data processing module 420, the result obtaining module 430, and the control module 440 in the control device, by processing the temperature data, environmental data, and load data for different fluctuation types, temperature characteristics, environmental characteristics, and load characteristics at different time scales are obtained, improving the flexibility of data processing for different types of data. Since the temperature prediction result of the power supply unit at a later time is predicted in advance according to the multi-dimensional feature data at the current time, the flow rate control process of the liquid cooling device is made more scientific and reasonable, providing more comprehensive decision-making support for the operation and maintenance of the power supply unit, optimizing the resource usage, and reducing the operation cost of the system.

[0119] According to the embodiments of the present application, the fluctuation types include short-term fluctuation types and long-term fluctuation types; the second data processing module 420 includes: a first extraction sub-module and a second extraction sub-module. The first extraction sub-module is configured to perform first feature extraction on the target temperature data, environmental data, and relevant load data respectively based on a first sliding window for short-term fluctuation types, so as to obtain first temperature characteristics, first environmental characteristics, and first load characteristics. The second extraction sub-module is configured to perform second feature extraction on the target temperature data, environmental data, and relevant load data respectively based on a second sliding window for long-term fluctuation types, so as to obtain second temperature characteristics, second environmental characteristics, and second load characteristics.

[0120] According to an embodiment of the present application, the result obtaining module 430 includes: a splicing sub-module and a result obtaining sub-module. The splicing sub-module is configured to splice the first temperature feature with the second temperature feature, the first environmental feature with the second environmental feature, and the first load feature with the second load feature respectively to obtain a spliced temperature feature, a spliced environmental feature, and a spliced load feature; the result obtaining sub-module is configured to obtain a temperature prediction result based on the spliced temperature feature, the spliced environmental feature, the spliced load feature, and their respective weights.

[0121] According to an embodiment of the present application, the result obtaining sub-module includes: a weighting unit and a prediction result obtaining unit. The weighting unit is configured to weight the spliced temperature feature, the spliced environmental feature, and the spliced load feature respectively by using a preset temperature weight, a preset environmental weight, and a preset load weight to obtain an intermediate weighted result; the prediction result obtaining unit is configured to obtain a temperature prediction result according to the intermediate weighted result and the state feature at the previous moment.

[0122] According to an embodiment of the present application, the result obtaining module 430 includes: an operation sub-module and a splicing sub-module. The operation sub-module is configured to perform a first operation on the first temperature feature, the first environmental feature, and the first load feature based on a first gating algorithm and a first time step to obtain a first operation result, and perform a second operation on the second temperature feature, the second environmental feature, and the second load feature based on a second gating algorithm and a second time step to obtain a second operation result, where the first time step is less than the second time step; the splicing sub-module is configured to splice the first operation result and the second operation result to obtain a temperature prediction result.

[0123] According to an embodiment of the present application, the first data processing module 410 includes: a weight value determination sub-module and a temperature data determination sub-module. The weight value determination sub-module is configured to determine the weight value of a target detection point among multiple detection points; the temperature data determination sub-module is configured to determine the target temperature data from multiple initial temperature data based on the weight value and the initial temperature data.

[0124] According to an embodiment of the present application, the weight value determination sub-module includes: a support degree determination unit and a weight value determination unit. The support degree determination unit is configured to determine the support degree between the initial temperature data of the target detection point and the initial temperature data of other detection points; the weight value determination unit is configured to determine the weight value of the target detection point based on the support degree, and the weight value is proportional to the support degree.

[0125] According to an embodiment of the present application, the temperature data determination sub-module includes: a target value determination unit and a temperature data determination unit. The target value determination unit is configured to determine a target value corresponding to the target detection point based on the initial temperature data and the weight value; the temperature data determination unit is configured to determine the initial temperature data of the target detection point as the target temperature data when the target value is greater than or equal to a preset threshold.

[0126] According to an embodiment of the present application, the environmental features include temperature features and humidity features at a previous moment; the apparatus further includes: a weight value updating sub-module, configured to update the weight value of a target detection point when the fluctuation value between the temperature feature and humidity feature at the current moment and the temperature feature and humidity feature at the previous moment is greater than or equal to a preset fluctuation threshold.

[0127] According to an embodiment of the present application, any plurality of modules among the first data processing module 410, the second data processing module 420, the result obtaining module 430, and the control module 440 may be combined and implemented in one module, or any one of them may be split into multiple modules. Alternatively, at least part of the functions of one or more of these modules may be combined with at least part of the functions of other modules and implemented in one module. According to an embodiment of the present application, at least one of the first data processing module 410, the second data processing module 420, the result obtaining module 430, and the control module 440 may be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or may be implemented by any other reasonable manner of integrating or packaging circuits, etc., in hardware or firmware, or implemented in any one of the three implementation manners of software, hardware, and firmware, or in an appropriate combination of any several of them. Alternatively, at least one of the first data processing module 410, the second data processing module 420, the result obtaining module 430, and the control module 440 may be at least partially implemented as a computer program module, and when the computer program module is run, corresponding functions may be executed.

[0128] Figure 5 A block diagram of an electronic device suitable for implementing a control method according to an embodiment of the present application is schematically shown.

[0129] As Figure 5 shown, the electronic device according to an embodiment of the present application includes a processor 501, which may perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage section 508 into a random access memory (RAM) 503. The processor 501 may include, for example, a general microprocessor (such as a CPU), an instruction set processor, and / or a related chipset, and / or a dedicated microprocessor (such as an application specific integrated circuit (ASIC)), etc. The processor 501 may also include on-board memory for caching purposes. The processor 501 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present application.

[0130] In the RAM 503, various programs and data required for the operation of the electronic device are stored. The processor 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. The processor 501 performs various operations of the method flow according to the embodiments of the present application by executing the programs in the ROM 502 and / or the RAM 503. It should be noted that the programs may also be stored in one or more memories other than the ROM 502 and the RAM 503. The processor 501 may also perform various operations of the method flow according to the embodiments of the present application by executing the programs stored in the one or more memories.

[0131] According to an embodiment of the present application, the electronic device may further include an input / output (I / O) interface 505, and the input / output (I / O) interface 505 is also connected to the bus 504. The electronic device may further include one or more of the following components connected to the input / output (I / O) interface 505: an input portion 506 including a keyboard, a mouse, etc.; an output portion 507 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc. and a speaker, etc.; a storage portion 508 including a hard disk, etc.; and a communication portion 509 including a network interface card such as a LAN card, a modem, etc. The communication portion 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the input / output (I / O) interface 505 as needed. A removable medium 511, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 510 as needed so that a computer program read from it can be installed into the storage portion 508 as needed.

[0132] The present application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or may exist separately without being assembled into the device / apparatus / system. The above computer-readable storage medium carries one or more programs, and when the one or more programs are executed, the method according to the embodiments of the present application is implemented.

[0133] According to an embodiment of the present application, the computer-readable storage medium may be a non-volatile computer-readable storage medium, for example, it may include but is not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In the present application, the computer-readable storage medium may be any tangible medium that contains or stores a program, and this program can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to an embodiment of the present application, the computer-readable storage medium may include the above-described ROM 502 and / or RAM 503 and / or one or more memories other than ROM 502 and RAM 503.

[0134] An embodiment of the present application also includes a computer program product, which includes a computer program that contains program code for executing the method shown in the flowchart. When the computer program product runs in a computer system, the program code is used to enable the computer system to implement the control method provided by the embodiment of the present application.

[0135] When the computer program is executed by the processor 501, it executes the above functions defined in the system / apparatus of the embodiment of the present application. According to an embodiment of the present application, the above-described systems, apparatuses, modules, units, etc. can be implemented by computer program modules.

[0136] In one embodiment, the computer program can rely on tangible storage media such as optical storage devices and magnetic storage devices. In another embodiment, the computer program can also be transmitted and distributed in the form of a signal on a network medium, and is downloaded and installed through the communication part 509, and / or installed from the removable medium 511. The program code contained in the computer program can be transmitted by any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0137] In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 509, and / or installed from the removable medium 511. When the computer program is executed by the processor 501, it executes the above functions defined in the system of the embodiment of the present application. According to an embodiment of the present application, the above-described systems, devices, apparatuses, modules, units, etc. can be implemented by computer program modules.

[0138] According to embodiments of the present application, program code for executing the computer programs provided by the embodiments of the present application can be written in any combination of one or more programming languages. Specifically, these computing programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages include, but are not limited to, such as Java, C++, Python, the "C" language, or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user's computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (for example, by using an Internet service provider to connect through the Internet).

[0139] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the above-mentioned module, program segment, or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram or flowchart, and the combination of blocks in the block diagram or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0140] Those skilled in the art can understand that the features described in the various embodiments of the present application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in the present application. In particular, without departing from the spirit and teachings of the present application, the features described in the various embodiments of the present application can be combined and / or combined in various ways. All such combinations and / or combinations fall within the scope of the present application.

[0141] The above describes the embodiments of the present application. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of the present application. Although the embodiments are described separately above, this does not mean that the measures in each embodiment cannot be used advantageously in combination. Without departing from the scope of the present application, those skilled in the art can make various substitutions and modifications, and all such substitutions and modifications should fall within the scope of the present application.

Claims

1. A control method, characterized in that: The method comprises: Processing the acquired multiple initial temperature data of multiple detection points in at least one power supply unit to determine target temperature data in the multiple initial temperature data; The target temperature data, the environmental data of the power supply unit and the related load data are processed respectively to obtain temperature characteristics, environmental characteristics and load characteristics for multiple fluctuation types; Based on the temperature characteristics, the environmental characteristics and the load characteristics, obtaining a temperature prediction result of the power supply unit at a later time; The flow rate of the liquid cooling device in the power supply unit is controlled based on the temperature prediction result.

2. The method according to claim 1, characterized in that The fluctuation types include short-term fluctuation types and long-term fluctuation types; The target temperature data, the environmental data of the power supply unit and the related load data are processed respectively to obtain temperature characteristics, environmental characteristics and load characteristics for multiple fluctuation types, including: Based on a first sliding window for the short-term fluctuation type, first feature extraction is performed on the target temperature data, the environmental data, and the relevant load data to obtain a first temperature feature, a first environmental feature, and a first load feature; Based on the second sliding window for the long-term fluctuation type, second feature extraction is performed on the target temperature data, the environmental data and the related load data respectively to obtain a second temperature feature, a second environmental feature and a second load feature.

3. The method according to claim 2, characterized in that Obtaining a temperature prediction result of the power supply unit at a later time based on the temperature feature, the environment feature, and the load feature, including: Respectively splicing the first temperature characteristic with the second temperature characteristic, the first environment characteristic with the second environment characteristic, and the first load characteristic with the second load characteristic to obtain a spliced ​​temperature characteristic, a spliced ​​environment characteristic, and a spliced ​​load characteristic; The temperature prediction result is obtained based on the splicing temperature characteristics, the splicing environment characteristics, the splicing load characteristics and their respective weights.

4. The method according to claim 3, characterized in that Based on the splicing temperature characteristics, the splicing environment characteristics, the splicing load characteristics and their respective weights, the temperature prediction result is obtained, including: Respectively weighting the splicing temperature feature, the splicing environment feature and the splicing load feature using a preset temperature weight, a preset environment weight and a preset load weight to obtain an intermediate weighted result; The temperature prediction result is obtained according to the intermediate weighted result and the state characteristics at the previous moment.

5. The method according to claim 2, characterized in that: Obtaining a temperature prediction result of the power supply unit at a later time based on the temperature feature, the environment feature, and the load feature, including: performing a first operation on the first temperature characteristic, the first environmental characteristic, and the first load characteristic based on a first gating algorithm and a first time step to obtain a first operation result, and performing a second operation on the second temperature characteristic, the second environmental characteristic, and the second load characteristic based on a second gating algorithm and a second time step to obtain a second operation result, wherein the first time step is smaller than the second time step; The first operation result and the second operation result are concatenated to obtain the temperature prediction result.

6. The method according to claim 1, characterized in that Processing the acquired multiple initial temperature data of multiple detection points in at least one power supply unit to determine target temperature data in the multiple initial temperature data includes: Determining a weight value of a target detection point among the multiple detection points; The target temperature data is determined from the plurality of initial temperature data based on the weight value and the initial temperature data.

7. The method according to claim 6, characterized in that Determining a weight value of a target detection point among the multiple detection points includes: Determine the support between the initial temperature data of the target detection point and the initial temperature data of other detection points; A weight value of the target detection point is determined based on the support, and the weight value is proportional to the support.

8. The method according to claim 6, characterized in that Determining the target temperature data from the plurality of initial temperature data based on the weight value and the initial temperature data includes: Determine a target value corresponding to the target detection point based on the initial temperature data and the weight value; When the target value is greater than or equal to a preset threshold, the initial temperature data of the target detection point is determined as the target temperature data.

9. The method according to any one of claims 6 to 8, characterized in that The environmental characteristics include temperature characteristics and humidity characteristics at the previous moment; the method also includes: When the fluctuation value between the temperature characteristic and the humidity characteristic at the current moment and the temperature characteristic and the humidity characteristic at the previous moment is greater than or equal to a preset fluctuation threshold, the weight value of the target detection point is updated.

10. A control device, characterized in that: The device comprises: A first data processing module, configured to process a plurality of initial temperature data of a plurality of detection points in at least one power supply unit, and determine target temperature data from the plurality of initial temperature data; A second data processing module is used to process the target temperature data, the environmental data of the power supply unit and the related load data respectively to obtain temperature characteristics, environmental characteristics and load characteristics for multiple fluctuation types; A result obtaining module, used for obtaining a temperature prediction result of the power supply unit at a later time based on the temperature characteristics, the environmental characteristics and the load characteristics; A control module is used to control the flow of the liquid cooling device in the power supply unit based on the temperature prediction result.

Images