Temperature prediction method, model training method and related equipment

Abstract

The present disclosure provides a temperature prediction method and device, a model training method and device, an electronic device and a computer-readable storage medium, and relates to the field of artificial intelligence technology. The temperature prediction method includes: obtaining air conditioning outlet data, server load data and cold pool air inlet air volume data; inputting the air conditioning outlet data into a pre-trained first machine learning model to obtain cold pool air inlet temperature and humidity prediction data; inputting the cold pool air inlet temperature and humidity prediction data, server load data and cold pool air inlet air volume data into a pre-trained second machine learning model to obtain cabinet air outlet temperature prediction data. The present disclosure can accurately predict the cabinet air outlet temperature, thereby obtaining the cabinet air outlet temperature in real time, conveniently and efficiently.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology, and in particular to a temperature prediction method and device, a model training method and device, an electronic device, and a computer-readable storage medium. Background Art

[0002] There are usually cabinets of different heights in a computer room, and these cabinets may have excessive temperatures during operation.

[0003] In order to detect whether the cabinet temperature is too high during operation, it is necessary to obtain the temperature of the cabinet air outlet. The traditional method of obtaining the temperature of the cabinet air outlet usually requires deploying a temperature sensor or a patrol robot in the computer room. Summary of the invention

[0004] A technical problem solved by the present disclosure is how to obtain the cabinet air outlet temperature in real time, conveniently and efficiently.

[0005] According to the first aspect of an embodiment of the present disclosure, a temperature prediction method is provided, including: obtaining air conditioning outlet data, server load data, and cold pool air inlet air volume data; inputting the air conditioning outlet data into a pre-trained first machine learning model to obtain cold pool air inlet temperature and humidity prediction data; inputting the cold pool air inlet temperature and humidity prediction data, server load data, and cold pool air inlet air volume data into a pre-trained second machine learning model to obtain cabinet air outlet temperature prediction data.

[0006] In some embodiments, the temperature prediction method also includes: determining the target air volume data of the cold pool air inlet based on the cold pool air inlet temperature and humidity prediction data, server load data, cabinet air outlet temperature target data and a pre-trained second machine learning model; adjusting the cold pool air inlet air volume data to the cold pool air inlet target air volume data, so as to adjust the cabinet air outlet temperature prediction data to the cabinet air outlet temperature target data.

[0007] In some embodiments, based on the cold pool air inlet temperature and humidity prediction data, server load data, cabinet air outlet temperature target data and a pre-trained second machine learning model, determining the cold pool air inlet target air volume data includes: traversing the cold pool air inlet air volume test data with a preset step size; inputting the cold pool air inlet temperature and humidity prediction data, server load data and cold pool air inlet air volume test data into the pre-trained second machine learning model to obtain the cabinet air outlet temperature test data; when the cabinet air outlet temperature test data is equal to the cabinet air outlet temperature target data, using the cold pool air inlet air volume test data as the cold pool air inlet target air volume data.

[0008] According to a second aspect of an embodiment of the present disclosure, a model training method is provided, including: collecting air conditioning outlet sample data and cold pool air inlet temperature and humidity sample data; using the air conditioning outlet sample data and the cold pool air inlet temperature and humidity sample data to train a first machine learning model, so that the first machine learning model can obtain cold pool air inlet temperature and humidity prediction data based on the air conditioning outlet data; collecting cold pool air inlet temperature and humidity sample data, server load sample data, cold pool air inlet air volume sample data, and cabinet air outlet temperature sample data; using the cold pool air inlet temperature and humidity sample data, server load sample data, cold pool air inlet air volume sample data, and cabinet air outlet temperature sample data to train a second machine learning model, so that the second machine learning model can obtain cabinet air outlet temperature prediction data based on the cold pool air inlet temperature and humidity prediction data, server load data, and cold pool air inlet air volume data.

[0009] According to a third aspect of an embodiment of the present disclosure, a temperature prediction device is provided, including: a data acquisition module, configured to: acquire air conditioning outlet data, server load data and cold pool air inlet air volume data; a first data prediction module, configured to: input the air conditioning outlet data into a pre-trained first machine learning model to obtain cold pool air inlet temperature and humidity prediction data; a second data prediction module, configured to: input the cold pool air inlet temperature and humidity prediction data, server load data and cold pool air inlet air volume data into a pre-trained second machine learning model to obtain cabinet air outlet temperature prediction data.

[0010] In some embodiments, the temperature prediction device also includes: a data determination module, configured to determine the target air volume data of the cold pool air inlet based on the cold pool air inlet temperature and humidity prediction data, server load data, cabinet air outlet temperature target data and a pre-trained second machine learning model; a data adjustment module, configured to adjust the cold pool air inlet air volume data to the cold pool air inlet target air volume data, so as to adjust the cabinet air outlet temperature prediction data to the cabinet air outlet temperature target data.

[0011] In some embodiments, the data determination module is configured to: traverse the cold pool air inlet air volume test data with a preset step size; input the cold pool air inlet temperature and humidity prediction data, server load data and cold pool air inlet air volume test data into a pre-trained second machine learning model to obtain the cabinet air outlet temperature test data; when the cabinet air outlet temperature test data is equal to the cabinet air outlet temperature target data, use the cold pool air inlet air volume test data as the cold pool air inlet target air volume data.

[0012] According to a fourth aspect of an embodiment of the present disclosure, a model training device is provided, including: a first model training module, configured to: collect air conditioning outlet sample data and cold pool air inlet temperature and humidity sample data; use the air conditioning outlet sample data and the cold pool air inlet temperature and humidity sample data to train a first machine learning model, so that the first machine learning model can obtain cold pool air inlet temperature and humidity prediction data based on the air conditioning outlet data; a second model training module, configured to: collect cold pool air inlet temperature and humidity sample data, server load sample data, cold pool air inlet air volume sample data and cabinet air outlet temperature sample data; use the cold pool air inlet temperature and humidity sample data, server load sample data, cold pool air inlet air volume sample data and cabinet air outlet temperature sample data to train a second machine learning model, so that the second machine learning model can obtain cabinet air outlet temperature prediction data based on the cold pool air inlet temperature and humidity prediction data, server load data and cold pool air inlet air volume data.

[0013] According to a fifth aspect of an embodiment of the present disclosure, there is provided an electronic device, comprising: a memory; and a processor coupled to the memory, wherein the processor is configured to execute the aforementioned temperature prediction method or the aforementioned model training method based on instructions stored in the memory.

[0014] According to a sixth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and when the instructions are executed by a processor, the aforementioned temperature prediction method or the aforementioned model training method is implemented.

[0015] The present disclosure can accurately predict the cabinet air outlet temperature, thereby obtaining the cabinet air outlet temperature in real time, conveniently and efficiently.

[0016] Other features and advantages of the present disclosure will become apparent from the following detailed description of exemplary embodiments of the present disclosure with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in 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 described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0018] Figure 1 A schematic diagram showing cooling of a computer room is shown.

[0019] Figure 2 A flowchart of a model training method according to some embodiments of the present disclosure is shown.

[0020] Figure 3A schematic flow chart of a temperature prediction method according to some embodiments of the present disclosure is shown.

[0021] Figure 4 A schematic flow chart of a temperature prediction method according to some other embodiments of the present disclosure is shown.

[0022] Figure 5 A schematic diagram of the structure of a temperature prediction device according to some embodiments of the present disclosure is shown.

[0023] Figure 6 A schematic diagram of the structure of a model training device according to some embodiments of the present disclosure is shown.

[0024] Figure 7 A schematic structural diagram of an electronic device according to some embodiments of the present disclosure is shown. DETAILED DESCRIPTION

[0025] The following will be combined with the drawings in the embodiments of the present disclosure to clearly and completely describe the technical solutions in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all of the embodiments. The following description of at least one exemplary embodiment is actually only illustrative and is by no means intended to limit the present disclosure and its application or use. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.

[0026] After research, the inventor found that the traditional method of obtaining the temperature of the cabinet air outlet requires the modification of the computer room. Since the air inlet and outlet temperatures at different heights of the cabinet are different, it is necessary to install sensors at different heights of the cabinet, which is costly and difficult to operate and maintain. At the same time, the large number of temperature measurement points will cause the inspection robot to take too long to inspect the computer room for a week, making it difficult to monitor the cabinet air outlet temperature in real time, and unable to promptly detect whether the cabinet has too high a temperature during operation.

[0027] Figure 1 A schematic diagram of the computer room cooling is shown. After analysis, the inventor believes that the excessively high local temperature in the computer room is caused by the unreasonable organization of the airflow in the computer room and the high-load operation of the server in the cabinet. Since the computer room environment generally does not change dramatically, the temperature and humidity of the air outlet of the air conditioner and the inlet temperature and humidity of each cold pool inlet generally have a stable functional mapping relationship. At the same time, the air outlet temperature of each cabinet has a relatively stable functional mapping relationship with the temperature and humidity of the cold pool inlet, the air volume, and the server load. These functional mapping relationships can be learned and represented by artificial intelligence algorithms.

[0028] First combine Figure 2 Some embodiments of the disclosed model training method are described.

[0029] Figure 2 FIG. 1 is a flow chart showing a model training method according to some embodiments of the present disclosure. Figure 2 As shown, this embodiment includes steps S201 to S204.

[0030] In step S201, air conditioner outlet sample data and cold pool inlet temperature and humidity sample data are collected.

[0031] For example, the air conditioning outlet sample data can be collected from the dynamic environment system of the computer room, which can include temperature data, humidity data and air volume data; sensors can also be installed at the cold pool air inlet to collect the cold pool air inlet temperature data and humidity data. These data can be uploaded to the intelligent hotspot detection and elimination system.

[0032] In step S202, the first machine learning model is trained using the air conditioning outlet sample data and the cold pool inlet temperature and humidity sample data.

[0033] The first machine learning model is trained to learn the following mapping function .

[0034]

[0035] Among them, T pool_in_m is the cold pool inlet temperature at position m, H pool_in_m is the humidity of the cold pool inlet at position m, T acn is the outlet temperature of the nth air conditioner, H acn is the outlet humidity of the nth air conditioner, F acn is the air volume of the nth air conditioner. The trained first machine learning model can obtain the predicted data of the temperature and humidity at the cold pool inlet according to the air conditioner air output data.

[0036] In step S203, sample data of temperature and humidity at the cold pool air inlet, sample data of server load, sample data of air volume at the cold pool air inlet, and sample data of temperature at the cabinet air outlet are collected. These data can also be uploaded to the intelligent hotspot detection and elimination system.

[0037] For example, sensors can be installed at the cold pool air inlet to collect cold pool air inlet sample data and air volume sample data; server load sample data can be collected from the computer room server management system, which may include CPU utilization, disk input and output speed, etc.; sensors can also be installed at the cabinet air outlet to collect cabinet air outlet temperature sample data.

[0038] In step S204, the second machine learning model is trained using the cold pool air inlet temperature and humidity sample data, the server load sample data, the cold pool air inlet air volume sample data, and the cabinet air outlet temperature sample data.

[0039] The second machine learning model is trained to learn the following mapping function .

[0040]

[0041] Among them, T cain_m_h is the temperature of the air outlet of the cabinet at position m at height h, T pool_in_m is the air inlet temperature of the cold pool at position m, H pool_in_m is the air humidity at the cold pool inlet at position m, F pool_in_m is the air volume at the cold pool inlet at position m, R cpu_mn is the CPU utilization of the server at the height of n in the cabinet at position m, and Rio_n is the disk input and output read and write speed of the server at the height of n in the cabinet at position m. The trained second machine learning model can obtain the cabinet outlet temperature prediction data based on the cold pool inlet temperature and humidity prediction data, server load data, and cold pool inlet air volume data.

[0042] Those skilled in the art should understand that the first machine learning model and the second machine learning model may adopt machine learning models that can better learn nonlinear functions and perform logistic regression, specifically, they may be distributed gradient boosting library XGBoost, SVM (Support Vector Machine), neural network model, etc.

[0043] Combine the following Figure 3 Some embodiments of the temperature prediction method of the present disclosure are described.

[0044] Figure 3 FIG. 2 is a flow chart showing a temperature prediction method according to some embodiments of the present disclosure. Figure 3 As shown, this embodiment includes steps S301 to S303.

[0045] In step S301, air outlet data of the air conditioner, server load data and air volume data of the cold pool inlet are obtained.

[0046] For example, the air conditioner outlet data, server load data, and cold pool air inlet air volume data in real time can be updated to the intelligent hotspot detection and elimination system in real time. Then, the air conditioner outlet data, server load data, and cold pool air inlet air volume data in real time are obtained from the intelligent hotspot detection and elimination system.

[0047] In step S302, the air-conditioning outlet data is input into a pre-trained first machine learning model to obtain the temperature and humidity prediction data of the cold pool air inlet.

[0048] In step S303, the cold pool air inlet temperature and humidity prediction data, server load data, and cold pool air inlet air volume data are input into a pre-trained second machine learning model to obtain the cabinet air outlet temperature prediction data.

[0049] This embodiment provides an intelligent temperature prediction method for a computer room. In the early data collection stage, sensors are used to collect data and perform modeling. In the later temperature prediction stage, there is no need to use sensors. Instead, a machine learning model is used to achieve full-time and real-time prediction of the cabinet air outlet temperature, thereby obtaining the cabinet air outlet temperature in real time, conveniently and efficiently, with low cost and easy operation and maintenance.

[0050] It should be noted that: on the one hand, this embodiment does not directly predict the cabinet outlet temperature through the air conditioner outlet data and server load data, but adds the modeling of the cold pool inlet temperature and humidity data, which can reflect the heat loss in the intermediate process of the air flow direction; on the other hand, this embodiment also considers the impact of the fluctuation of the server load on the cabinet outlet temperature from the perspective of heat generation, and considers the impact of the cold pool inlet temperature and humidity on the cabinet outlet temperature from the perspective of cooling. Therefore, this embodiment can more accurately predict the cabinet outlet temperature.

[0051] The inventors further studied and found that the traditional method of obtaining the temperature of the cabinet outlet cannot provide accurate suggestions to alleviate the phenomenon of overheating of the cabinet during operation after execution. In fact, the computer room usually adopts a large-scale over-cooling method to avoid the phenomenon of overheating of the cabinet during operation, which will cause huge energy loss. To this end, the present disclosure also provides corresponding embodiments to further solve this problem.

[0052] Combine the following Figure 4 Other embodiments of the temperature prediction method disclosed herein are described.

[0053] Figure 4 FIG. 2 is a flow chart showing a temperature prediction method according to another embodiment of the present disclosure. Figure 4 As shown, this embodiment includes steps S404 to S405.

[0054] In step S404, the target air volume data of the cold pool air inlet is determined according to the temperature and humidity prediction data of the cold pool air inlet, the server load data, the cabinet air outlet temperature target data and the pre-trained second machine learning model.

[0055] First, the cold pool inlet air volume test data is traversed with a preset step size. , , , , , … , Under these parameters, Then, the cold pool air inlet temperature and humidity prediction data, server load data, and the cold pool air inlet air volume test data after traversal are input into the pre-trained second machine learning model in turn to obtain the cabinet air outlet temperature test data. When the cabinet air outlet temperature test data is equal to the cabinet air outlet temperature target data T target When , the cold pool air inlet air volume test data is used as the cold pool air inlet target air volume data. That is, find the second machine learning model output equal to T target The corresponding input parameter F target .

[0056] In step S405, the cold pool air inlet air volume data is adjusted to the cold pool air inlet target air volume data, so as to adjust the cabinet air outlet temperature prediction data to the cabinet air outlet temperature target data.

[0057] When adjusting the cold pool air inlet air volume data, assume that the current cold pool air inlet air volume data is F now , F target As the target air volume data for the cold pool inlet. Then, That is, the adjustment amount for the current cold pool inlet air volume data.

[0058] This embodiment can provide relatively accurate cold pool air inlet target air volume data and cold pool air inlet air volume data adjustment amount based on the detection of cabinet outlet temperature prediction data, without the need for computer room operation and maintenance personnel to adjust the cold pool air inlet air volume based on work experience. Therefore, it can conveniently and efficiently alleviate or even avoid the phenomenon of overtemperature of the cabinet during operation in the computer room production environment.

[0059] Combine the following Figure 5 Some embodiments of the temperature prediction device of the present disclosure are described.

[0060] Figure 5 FIG. 2 shows a schematic diagram of the structure of a temperature prediction device according to some embodiments of the present disclosure. Figure 5 As shown, the temperature prediction device 50 in this embodiment includes: a data acquisition module 501, configured to: acquire air conditioning outlet data, server load data and cold pool air inlet air volume data; a first data prediction module 502, configured to: input the air conditioning outlet data into a pre-trained first machine learning model to obtain cold pool air inlet temperature and humidity prediction data; a second data prediction module 503, configured to input the cold pool air inlet temperature and humidity prediction data, server load data and cold pool air inlet air volume data into a pre-trained second machine learning model to obtain cabinet air outlet temperature prediction data.

[0061] This embodiment provides an intelligent temperature prediction method for a computer room. In the early data collection stage, sensors are used to collect data and perform modeling. In the later temperature prediction stage, there is no need to use sensors. Instead, a machine learning model is used to achieve full-time and real-time prediction of the cabinet air outlet temperature, thereby obtaining the cabinet air outlet temperature in real time, conveniently and efficiently, with low cost and easy operation and maintenance.

[0062] It should be noted that: on the one hand, this embodiment does not directly predict the cabinet outlet temperature through the air conditioner outlet data and server load data, but adds the modeling of the cold pool inlet temperature and humidity data, which can reflect the heat loss in the intermediate process of the air flow direction; on the other hand, this embodiment also considers the impact of the fluctuation of the server load on the cabinet outlet temperature from the perspective of heat generation, and considers the impact of the cold pool inlet temperature and humidity on the cabinet outlet temperature from the perspective of cooling. Therefore, this embodiment can more accurately predict the cabinet outlet temperature.

[0063] In some embodiments, the temperature prediction device also includes: a data determination module, configured to determine the target air volume data of the cold pool air inlet based on the cold pool air inlet temperature and humidity prediction data, server load data, cabinet air outlet temperature target data and a pre-trained second machine learning model; a data adjustment module, configured to adjust the cold pool air inlet air volume data to the cold pool air inlet target air volume data, so as to adjust the cabinet air outlet temperature prediction data to the cabinet air outlet temperature target data.

[0064] In some embodiments, the data determination module is configured to: traverse the cold pool air inlet air volume test data with a preset step size; input the cold pool air inlet temperature and humidity prediction data, server load data and cold pool air inlet air volume test data into a pre-trained second machine learning model to obtain the cabinet air outlet temperature test data; when the cabinet air outlet temperature test data is equal to the cabinet air outlet temperature target data, use the cold pool air inlet air volume test data as the cold pool air inlet target air volume data.

[0065] This embodiment can provide relatively accurate cold pool air inlet target air volume data and cold pool air inlet air volume data adjustment amount based on the detection of cabinet outlet temperature prediction data, without the need for computer room operation and maintenance personnel to adjust the cold pool air inlet air volume based on work experience. Therefore, it can conveniently and efficiently alleviate or even avoid the phenomenon of overtemperature of the cabinet during operation in the computer room production environment.

[0066] Combine the following Figure 6 Some embodiments of the disclosed model training apparatus are described.

[0067] Figure 6 FIG. 1 is a schematic diagram showing the structure of a model training device in some embodiments of the present disclosure. Figure 6As shown, the model training device 60 in this embodiment includes:

[0068] The first model training module 601 is configured to: collect air conditioning outlet sample data and cold pool air inlet temperature and humidity sample data; use the air conditioning outlet sample data and the cold pool air inlet temperature and humidity sample data to train the first machine learning model, so that the first machine learning model can obtain the cold pool air inlet temperature and humidity prediction data based on the air conditioning outlet data; the second model training module 602 is configured to: collect cold pool air inlet temperature and humidity sample data, server load sample data, cold pool air inlet air volume sample data and cabinet air outlet temperature sample data; use the cold pool air inlet temperature and humidity sample data, server load sample data, cold pool air inlet air volume sample data and cabinet air outlet temperature sample data to train the second machine learning model, so that the second machine learning model can obtain the cabinet air outlet temperature prediction data based on the cold pool air inlet temperature and humidity prediction data, server load data and cold pool air inlet air volume data.

[0069] Combine the following Figure 7 Some embodiments of the disclosed electronic device are described.

[0070] Figure 7 FIG. 1 is a schematic diagram showing the structure of an electronic device according to some embodiments of the present disclosure. Figure 7 As shown, the electronic device 70 of this embodiment includes: a memory 710 and a processor 720 coupled to the memory 710, and the processor 720 is configured to execute the temperature prediction method in any of the aforementioned embodiments or the aforementioned model training method based on instructions stored in the memory 710.

[0071] The memory 710 may include, for example, a system memory, a fixed non-volatile storage medium, etc. The system memory may store, for example, an operating system, an application program, a boot loader, and other programs.

[0072] The temperature prediction device 70 may also include an input / output interface 730, a network interface 740, a storage interface 750, etc. These interfaces 730, 740, 750, the memory 710, and the processor 720 may be connected, for example, via a bus 760. The input / output interface 730 provides a connection interface for input / output devices such as a display, a mouse, a keyboard, and a touch screen. The network interface 740 provides a connection interface for various networked devices. The storage interface 750 provides a connection interface for external storage devices such as SD cards and USB flash drives.

[0073] The present disclosure also includes a computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implement the temperature prediction method in any of the aforementioned embodiments or the aforementioned model training method.

[0074] The present disclosure is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present disclosure. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0075] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0076] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0077] The above description is only a preferred embodiment of the present disclosure and is not intended to limit the present disclosure. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present disclosure should be included in the protection scope of the present disclosure.

Claims

1. A temperature prediction method, comprising: Obtain air conditioning outlet data, server load data, and cold pool inlet air volume data; Input the air-conditioning outlet data into a pre-trained first machine learning model to obtain the cold pool air inlet temperature and humidity prediction data, wherein the trained first machine learning model can obtain the cold pool air inlet temperature and humidity prediction data according to the air-conditioning outlet data, the air-conditioning outlet data includes the air-conditioning outlet temperature, the air-conditioning outlet humidity, and the air-conditioning outlet volume, and the cold pool air inlet temperature and humidity prediction data includes the cold pool air inlet temperature prediction data and the cold pool air inlet humidity prediction data; The cold pool air inlet temperature and humidity prediction data, server load data, and cold pool air inlet air volume data are input into a pre-trained second machine learning model to obtain the cabinet air outlet temperature prediction data, wherein the trained second machine learning model can obtain the cabinet air outlet temperature prediction data based on the cold pool air inlet temperature and humidity prediction data, server load data, and cold pool air inlet air volume data, and the server load data includes the server's CPU utilization and the server's disk input and output read and write speeds.

2. The temperature prediction method according to claim 1, further comprising: Determine the target air volume data of the cold pool air inlet according to the predicted data of temperature and humidity of the cold pool air inlet, the server load data, the target data of the cabinet air outlet temperature, and the pre-trained second machine learning model; The cold pool air inlet air volume data is adjusted to the cold pool air inlet target air volume data, so as to adjust the cabinet air outlet temperature prediction data to the cabinet air outlet temperature target data.

3. The temperature prediction method according to claim 2, wherein: Determining the target air volume data of the cold pool air inlet according to the cold pool air inlet temperature and humidity prediction data, server load data, cabinet air outlet temperature target data, and a pre-trained second machine learning model includes: Traverse the cold pool inlet air volume test data with a preset step length; Input the cold pool air inlet temperature and humidity prediction data, server load data, and cold pool air inlet air volume test data into a pre-trained second machine learning model to obtain the cabinet air outlet temperature test data; When the cabinet air outlet temperature test data is equal to the cabinet air outlet temperature target data, the cold pool air inlet air volume test data is used as the cold pool air inlet target air volume data.

4. A model training method, comprising: Collect air conditioning outlet sample data and cold pool inlet temperature and humidity sample data; The first machine learning model is trained using the air conditioning outlet sample data and the cold pool inlet temperature and humidity sample data, so that the first machine learning model can obtain the cold pool inlet temperature and humidity prediction data according to the air conditioning outlet data, wherein the air conditioning outlet data includes the air conditioning outlet temperature, the air conditioning outlet humidity, and the air conditioning outlet volume, and the cold pool inlet temperature and humidity prediction data includes the cold pool inlet temperature prediction data and the cold pool inlet humidity prediction data; Collect cold pool air inlet temperature and humidity sample data, server load sample data, cold pool air inlet air volume sample data, and cabinet air outlet temperature sample data; The second machine learning model is trained using the cold pool air inlet temperature and humidity sample data, server load sample data, cold pool air inlet air volume sample data and cabinet air outlet temperature sample data, so that the second machine learning model can obtain the cabinet air outlet temperature prediction data based on the cold pool air inlet temperature and humidity prediction data, server load data and cold pool air inlet air volume data, wherein the server load data includes the server's CPU utilization and the server's disk input and output read and write speeds.

5. A temperature prediction device, comprising: The data acquisition module is configured to: acquire air outlet data of the air conditioner, server load data, and air volume data of the cold pool inlet; The first data prediction module is configured to: input the air conditioning outlet data into a pre-trained first machine learning model to obtain the temperature and humidity prediction data of the cold pool air inlet; The second data prediction module is configured to input the cold pool air inlet temperature and humidity prediction data, server load data and cold pool air inlet air volume data into a pre-trained second machine learning model to obtain the cabinet air outlet temperature prediction data.

6. The temperature prediction device according to claim 5, further comprising: The data determination module is configured to: determine the target air volume data of the cold pool air inlet according to the predicted data of the temperature and humidity of the cold pool air inlet, the server load data, the target data of the cabinet air outlet temperature, and the pre-trained second machine learning model; The data adjustment module is configured to adjust the cold pool air inlet air volume data to the cold pool air inlet target air volume data, so as to adjust the cabinet air outlet temperature prediction data to the cabinet air outlet temperature target data.

7. The temperature prediction device according to claim 6, wherein: The data determination module is configured to: Traverse the cold pool inlet air volume test data with a preset step length; Input the cold pool air inlet temperature and humidity prediction data, server load data, and cold pool air inlet air volume test data into a pre-trained second machine learning model to obtain the cabinet air outlet temperature test data; When the cabinet air outlet temperature test data is equal to the cabinet air outlet temperature target data, the cold pool air inlet air volume test data is used as the cold pool air inlet target air volume data.

8. A model training device, comprising: The first model training module is configured to: collect air conditioner outlet sample data and cold pool inlet temperature and humidity sample data; use the air conditioner outlet sample data and cold pool inlet temperature and humidity sample data to train the first machine learning model, so that the first machine learning model can obtain cold pool inlet temperature and humidity prediction data based on the air conditioner outlet data; The second model training module is configured to: collect cold pool air inlet temperature and humidity sample data, server load sample data, cold pool air inlet air volume sample data and cabinet air outlet temperature sample data; use the cold pool air inlet temperature and humidity sample data, server load sample data, cold pool air inlet air volume sample data and cabinet air outlet temperature sample data to train the second machine learning model, so that the second machine learning model can obtain the cabinet outlet temperature prediction data based on the cold pool air inlet temperature and humidity prediction data, server load data and cold pool air inlet air volume data.

9. An electronic device, comprising: Memory; as well as A processor coupled to the memory, the processor being configured to execute the temperature prediction method according to any one of claims 1 to 3 or the model training method according to claim 4 based on instructions stored in the memory.

10. A computer-readable storage medium, wherein: The computer-readable storage medium stores computer instructions, which, when executed by a processor, implement the temperature prediction method according to any one of claims 1 to 3 or the model training method according to claim 4.

Images