Temperature control system and its control method

Abstract

The present invention discloses a temperature control system and a control method thereof. The temperature control method includes the following steps: driving a temperature control device to generate an air circulation for a first server and a second server; continuously monitoring the operating states of the temperature control device, the first server and the second server to establish a first learning model; receiving temperature control status data of the temperature control device, first status data of the first server and second status data of the second server, wherein the first status data includes a first temperature of the first server, and the second status data includes a second temperature of the second server; inputting the temperature control status data, the first status data and the second status data into the first learning model to obtain a first temperature prediction value output by the first learning model; and adjusting the temperature control device according to the first temperature prediction value.

Description

Technical Field

[0001] The present invention relates to a temperature control system and a control method thereof, in particular to a technology for driving a temperature control device according to the temperature of a computer room or a cabinet. Background Art

[0002] With the development of the Internet of Things (IoT), artificial intelligence, and network technology, the construction of "computer rooms" (i.e., data centers) has received increasing attention. A computer room refers to a space where a large number of servers are systematically installed in cabinets for the management and maintenance of data. Currently, the industry has established standardized guidelines for the specific technical requirements of computer rooms.

[0003] The Rack Cooling Index (RCI) refers to the temperature value of the air intake of each cabinet (rack intake temperature). For example, the air intake temperature of the cabinet should not exceed the range of 18 - 27 °C. When the air intake temperature is lower than 18 °C, it means that the temperature control system of the computer room is in a state of excessive energy consumption; on the contrary, if the air intake temperature is higher than 27 °C, it means that the cooling efficiency of the temperature control equipment is poor. When the intake air temperature of the cabinet is too high, the temperature control equipment is actuated. Due to the uncertainty of time delay, there may be concerns that the temperature control equipment may not respond in time; furthermore, due to the height effect of the cabinet and the change of the server load, the intake air temperature and the cabinet height show a variable gradient change, which also increases the difficulty of controlling the intake air temperature of the cabinet.

[0004] The control of the computer room temperature often faces a dilemma. When designing a computer room, its temperature control equipment is usually over-designed to ensure that the computer room can be cooled immediately to maintain a low temperature and keep the system operating well. However, this will also keep the operating cost high and cause too high cooling costs. Therefore, how to balance the energy consumption and efficiency of the temperature control system while ensuring the normal operation of the server will be a major issue at present. Summary of the Invention

[0005] The content of the present application is about a temperature control method, including the following steps: driving a temperature control device to generate an air circulation for a first server and a second server; continuously monitoring the operating states of the temperature control device, the first server, and the second server to establish a first learning model, where the first learning model is used to predict the temperature change of the first server; receiving the temperature control status data of the temperature control device, the first status data of the first server, and the second status data of the second server, where the first status data includes the first temperature of the first server, and the second status data includes the second temperature of the second server; inputting the temperature control status data, the first status data, and the second status data into the first learning model to obtain a first temperature prediction value output by the first learning model; and adjusting the temperature control device according to the first temperature prediction value.

[0006] Preferably, the method for continuously monitoring the operating states of the temperature control device, the first server, and the second server includes: obtaining a cold air temperature or a cold air flow rate of the temperature control device in a detection period; and obtaining a plurality of operating temperatures when the first server and the second server are operating in the detection period.

[0007] Preferably, it further includes: continuously monitoring the operating states of the temperature control device, the first server, and the second server to establish a second learning model, where the second learning model is used to predict a temperature change of the second server; receiving the temperature control status data, the first status data, and the second status data; and inputting the temperature control status data, the first status data, and the second status data into the second learning model to obtain a second temperature prediction value output by the second learning model.

[0008] Preferably, the first server and the second server are disposed in a cabinet, and the method for adjusting the temperature control device according to the first temperature prediction value includes: calculating an overall temperature prediction value of the cabinet according to the first temperature prediction value and the second temperature prediction value; and adjusting a cold air temperature or a cold air flow rate of the temperature control device according to the overall temperature prediction value.

[0009] Preferably, it further includes: when the overall temperature prediction value is higher than a preset temperature upper limit, reducing the cold air temperature or increasing the cold air flow rate; and when the overall temperature prediction value is lower than a preset temperature lower limit, temporarily stopping the operation of the temperature control device.

[0010] The present application also relates to a temperature control system applicable to a data center, including a temperature control device, a first sensor, a second sensor, and an analysis device. The temperature control device is used to generate air circulation for the first server and the second server. The first sensor is used to detect the operating state of the first server and obtain first status data. The second sensor is used to detect the operating state of the second server and obtain second status data. The analysis device is electrically connected to the temperature control device, the first sensor, and the second sensor, and is used to continuously monitor the operating states of the temperature control device, the first server, and the second server to establish a first learning model. The analysis device is further used to input the temperature control status data of the temperature control device, the first temperature of the first status data, and the second temperature of the second status data into the first learning model to obtain a first temperature prediction value output by the first learning model, and the analysis device is used to adjust the temperature control device according to the first temperature prediction value.

[0011] Preferably, the analysis device is used to detect a cold air temperature or a cold air flow rate of the temperature control device in a detection period, and obtain a plurality of operating temperatures when the first server and the second server are operating, so as to establish the first learning model.

[0012] Preferably, the analysis device is further used to continuously monitor the operating states of the temperature control device, the first server and the second server, so as to establish a second learning model. The analysis device is further used to input the temperature control state data, the first state data and the second state data into the second learning model, so as to obtain a second temperature prediction value output by the second learning model.

[0013] Preferably, the analysis device is further used to calculate an overall temperature prediction value of a cabinet according to the first temperature prediction value and the second temperature prediction value, and adjust a cold air temperature or a cold air flow rate of the temperature control device according to the overall temperature prediction value.

[0014] Preferably, when the overall temperature prediction value is higher than a preset temperature upper limit, the analysis device is used to control the temperature control device to lower the cold air temperature, or control the temperature control device to increase the cold air flow rate; and when the overall temperature prediction value is lower than a preset temperature lower limit, the analysis device is used to temporarily stop operating the temperature control device.

[0015] Accordingly, the temperature control system respectively establishes learning models according to each server. In addition to reducing the computational complexity of training models and predicting temperatures, it can also predict the changes in the intake air temperature gradient distribution of the cabinet after several time units through the learning models established for each server respectively, so that the temperature control system can pre-adjust the temperature control device when it judges that the future temperature will exceed the preset range, ensuring the operation efficiency of the server and the energy consumption of the temperature control device, and having the function of optimizing energy efficiency. Description of the Drawings

[0016] Figure 1A It is a schematic diagram of a temperature control system according to some embodiments of the present application.

[0017] Figure 1B It is a schematic diagram of a temperature control system according to some embodiments of the present application.

[0018] Figure 2 It is a flowchart of a temperature control system control method according to some embodiments of the present application.

[0019] Figure 3 It is a schematic diagram of a distributed learning model according to some embodiments of the present application.

[0020] Symbol Explanation:

[0021] 100: Temperature control system

[0022] 110: Temperature control device

[0023] 120: Analysis device

[0024] 200: Cabinet

[0025] 211: Air inlet

[0026] 212: Air outlet

[0027] 220: Intake port

[0028] D1 - D10: Server

[0029] S1 - S11: Sensor

[0030] M1 - Mn: Learning model

[0031] Ma: Feature extraction module

[0032] Mb: Training module

[0033] Mc: Temperature prediction module

[0034] Xt: Input feature

[0035] Y1 - Yn: Output target

[0036] Yt: Overall temperature prediction value

[0037] T1: First distance

[0038] S201 - S207: Steps Detailed implementation manners

[0039] The following will disclose multiple embodiments of the present invention with diagrams. For the sake of clear illustration, many practical details will be described together in the following narrative. However, it should be understood that these practical details are not used to limit the present invention. That is to say, in some embodiments of the present invention, these practical details are unnecessary. In addition, for the purpose of simplifying the diagrams, some conventional structures and elements will be shown in a simple schematic manner in the diagrams.

[0040] In this document, when an element is referred to as "connected" or "coupled", it may mean "electrically connected" or "electrically coupled". "Connected" or "coupled" can also be used to indicate the cooperative operation or interaction between two or more elements. In addition, although terms such as "first", "second",... are used in this document to describe different elements, these terms are only used to distinguish elements or operations described with the same technical terms. Unless the context clearly indicates otherwise, these terms do not specifically refer to or imply an order or sequence, nor are they used to limit the present invention.

[0041] This application is related to a temperature control system 100 and its control method. Figure 1A and Figure 1B Shown is a schematic diagram of a temperature control system 100 according to some embodiments of this application. In this embodiment, the temperature control system 100 is applied to the computer room of a data center (such as Figure 1A shown), and the computer room contains a plurality of cabinets 200. Each cabinet 200 has a plurality of placement spaces for setting a plurality of servers D1 to D10. Each of the servers D1 to D10 may include a heater, a fan, and its heat dissipation module, as well as a wireless module, and can control the fan speed, control the power of the heater, and acquire the internal temperature of the server. Since those skilled in the art can understand the structure of the cabinet 200 and the way the servers D1 to D10 are installed in the cabinet, it will not be elaborated here.

[0042] The temperature control system 100 includes a temperature control device 110, a plurality of sensors S1 to S10, and an analysis device 120. In some embodiments, the temperature control device 110 includes a blower, and the blower drives a fan through a motor to generate cold air in the direction of the cabinet 200. The cold air will form an air circulation among the plurality of cabinets 200 to control the temperature of the cabinet 200 and the servers D1 to D10. In one embodiment, the temperature control system 100 further includes an exhaust device (not shown in the figure), and the exhaust device is arranged above the cabinet 200 to assist in forming the air circulation. In other embodiments, the temperature control device 110 can also generate hot air, and the air circulation is not limited to cold air.

[0043] In some embodiments, the cabinet 200 has a cold chamber to guide cold air and hot air and prevent the cold and hot air from mixing with each other. As in Figure 1A the shown computer room, there is a cold air channel (such as: Figure 1A the middle arrow) among the plurality of cabinets 200, and the temperature control device 110 is used to generate cold air for the cold air channel. After the cold air flows through the servers D1 to D10, it will flow out from the hot air channel (such as: Figure 1AFlow out through the arrow signs on both sides). In addition, the blower of the temperature control device 110 can be arranged under the raised floor of the computer room (i.e., corresponding to the position of the cold air channel) to ensure the air flow rate.

[0044] In one embodiment, sensors S1 to S10 are respectively arranged at positions adjacent to each of the servers D1 to D10 in the cabinet 200 to detect the operating states of the servers D1 to D10. Please refer to Figure 1B , an air inlet 211 and an air outlet 212 are arranged at positions corresponding to each placement space on the cabinet 200. In some embodiments, the sensors S1 to S10 (such as: thermocouples) are arranged at positions corresponding to the air inlet 211 to detect the temperature of the cold air flowing to the servers D1 to D10. In other embodiments, the sensors S1 to S10 can be arranged at positions corresponding to the air outlet 212, or sensors can be respectively arranged for the air inlet 211 and the air outlet 212 of each of the servers D1 to D10 in the cabinet 200 to detect the temperature of the cold air flowing through the servers D1 to D10. In other partial embodiments, the sensors S1 to S10 can also be electrically connected to the servers D1 to D10 to detect the load power of the servers D1 to D10 and transmit the detected data to the analysis device 120.

[0045] The analysis device 120 is electrically connected to the temperature control device 110 and the sensors S1 to S10 to continuously monitor the operating states of the temperature control device 110 and the servers D1 to D10. The analysis device 120 is used to respectively establish a corresponding learning model for each server D1 according to the monitored operating states of the temperature control device 110 and the servers D1 to D10 to predict the temperature change of each server D1 in a future period of time.

[0046] Through the pre-established learning model, the temperature control system 100 can predict the temperature change after a future period of time by monitoring the current states of the servers D1 to D10. For example: after establishing the "first learning model" for predicting the future temperature of the first server D1, the first learning model can estimate the first temperature prediction value according to the state data of the temperature control device 110 and the first server D1 and the second server D2, and adjust the temperature control device 110 accordingly.

[0047] First, taking the first sensor S1 and the second sensor S2 as examples, the way for the temperature control system 100 to establish a learning model is described below. Since the operation modes of the remaining sensors S3 to S10 are similar to those of the sensors S1 and S2, they will not be repeated separately: The first sensor S1 and the second sensor S2 are used to detect the operation states of the first server D1 and the second server D2. In some embodiments, the "operation state" includes the current temperatures of the first server D1 and the second server D2. In addition, when establishing the "first learning model" for predicting the future temperature of the first server D1, the "operation state" may further include the current load power of the first server D1, the fan speed of the cooling fan in the first server D1, or the change trend of the fan speed.

[0048] The operation states used to establish the learning model include (but are not limited to) the following data: the cold air temperature and cold air flow rate of the temperature control device 110, the temperatures at the air inlet 211 or the air outlet 212 of the cabinet 200, the load powers, fan speeds, and temperatures of key components (such as the processing CPU) of the servers D1 to D11, the installation positions of the servers D1 to D11 (such as the cabinet height), etc.

[0049] In one embodiment, the analysis device 120 uses a deep learning model to train with a large amount of historical data to establish a learning model. Taking the first learning model for "predicting the temperature of the first server D1 (such as the temperature at the air inlet 211)" as an example, the analysis device 120 sets the operation states (such as the current temperatures) of the temperature control device 110, the first server D1, and the second server D2 as input features, and sets the temperature of the first server D1 after a period of time (such as three minutes later) as the output target, and conducts training in a deep learning manner. After a large amount of training, the mapping function relationship between the input features and the output target can be established, that is, the first learning model.

[0050] In some embodiments, the deep learning model used by the analysis device 120 is GRU (Gated Recurrent Unit). GRU is a type of recurrent neural network (RNN - Recurrent Neural Network), especially used to handle data problems with time sequences. Through an appropriate deep learning model, the mapping function relationship of the model can be updated periodically with new data.

[0051] The change in the overall intake air temperature of the cabinet 200 is affected by many physical variables, such as the fan speeds of each of the servers D1 to D10, the inlet air temperature, the outlet air temperature, and the power loads of the servers D1 to D20 themselves. If the analysis device 120 is to directly establish a learning model with the "overall temperature of the cabinet 200" as the output target, then all the variable factors of the servers D1 to D10 must be set as input features. In this way, the computational load of the analysis device 120 will be too large, making it difficult to accurately and quickly complete the prediction. Therefore, the content of this application sets the target of the learning model as the "temperature of each server", predicts the future temperature of each server separately, and then evaluates the overall temperature of the cabinet 200 based on multiple prediction results. This "distributed" learning model can balance the computational efficiency and prediction accuracy of the analysis device 120, enabling the temperature control system 100 to immediately adjust the temperature control device 110 when it is inferred that the future temperature of the cabinet 200 will exceed the expected range.

[0052] When establishing a learning model, the temperature control system 100 of the content of this application, in addition to monitoring the operating status of the target server, will also obtain the operating status adjacent to the target server based on the thermal interaction of adjacent servers to improve the analysis accuracy. For example, when establishing the first learning model for the first server D1, in addition to obtaining the temperature and load power of the first server D1, the temperature of the second server D2 (adjacent to the first server D1) will also be obtained simultaneously. In some embodiments, since the "temperature" of the second server D2 affects the temperature of the first server D1, but the load power of the second server D2 does not affect the temperature of the first server D1, the analysis device 120 only applies the temperature of the second server D2 to train the first learning model.

[0053] After establishing the learning model, the temperature control system 100 can continuously monitor the status data of the temperature control device 110 and each of the servers D1 to D10, and input the status data into the learning model to predict the future temperature of each server. It should be particularly mentioned here that the "operating status" used by the temperature control system 100 to establish / train the learning model and the "status data" that the temperature control system subsequently continuously monitors for the temperature control device 110 and each of the servers D1 to D10 can be of the same type of data. The difference between the two is that the "operating status" is used to train the learning model, and the "status data" is input into the learning model as input features to predict the temperature for a subsequent period of time.

[0054] The present application can predict the temperature of servers D1-D10 (e.g., the temperature inside the server or at the air inlet 211) through a learning model. When it is predicted that the temperature of servers D1-D10 is too high, the fan of the temperature control device 100 can be controlled in advance to improve air circulation, thereby realizing a state predictor of the temperature control system 100. Based on the future temperature changes predicted by the learning model, the temperature control system 100 can be dynamically controlled according to the "trend" of future temperature changes. In addition, the temperature control system 100 of the present application can also be applied to the problem of variable heat dissipation modes in the computer room configurations of different data centers.

[0055] Figure 2 The schematic diagram of the temperature control system control method of the present application is shown. Figure 1A to Figure 2 The operation of the present application is described as follows: In step S201 , the temperature control device 110 is driven to generate cold air for the servers D1 - D10 in the cabinet 200 to form an air circulation.

[0056] In step S202, the analysis device 120 continuously monitors the operating status of the temperature control device 110 and the servers D1-D10, and establishes a plurality of learning models according to the operating status and temperature of the temperature control device 110 and the servers D1-D10. The learning model is used to predict the temperature change of the corresponding server within a period of time in the future (e.g., within the next 1-3 minutes) or after a period of time in the future (after 1 minute).

[0057] Specifically, taking the first learning model as an example, the analysis device 120 will receive data from the temperature control device 110, the first sensor S1, and the second sensor S2 in a detection cycle (e.g., every three minutes). This data may include the cold air temperature or cold air flow rate set by the temperature control device 110 to be generated, the current operating temperature of the first server D1 (e.g., the air inlet temperature or the air outlet temperature) and the load power, and the current operating temperature of the second server D2 (e.g., the air inlet temperature). The analysis device 120 will continue to receive operating data to train the first learning model. For example, the "current operating data of the temperature control device 110, the first sensor S1, and the second sensor S2" is used as the input feature for training, and the "actual temperature after one minute" of the first sensor S1 is used as the confirmed output target, so that the first learning model can establish parameters or weight values ​​corresponding to each input feature after multiple trainings.

[0058] Similarly, the analysis device 120 will continuously detect the operating states of the temperature control device, the second server D2, and the first server D1 in a detection period to establish a second learning model. In other partial embodiments, since both the first server D1 and the third server D3 are adjacent to the second server D2 (i.e., above and below the same cabinet), the analysis device 120 is further configured to establish / train the second learning model according to the current temperature of the third server D3.

[0059] In step S203, after the learning model is established, the temperature control system 100 can accordingly predict whether the future temperature of the cabinet 200 or the servers D1-D10 will exceed the preset range. Specifically, the analysis device 120 continuously or periodically receives the temperature control status data of the temperature control device 110, the status data of the server to be predicted (e.g., the first status data of the first server D1), and the status data of the adjacent server (e.g., the second status data of the second server D2).

[0060] The temperature control status data may include the cold air temperature or the cold air flow rate of the temperature control device 110. In some embodiments, the cold air temperature or the cold air flow rate may be the set value of the temperature control device 110. The analysis device 120 is electrically connected to the temperature control device 110 to obtain or adjust the cold air temperature or the cold air flow rate of the temperature control device 110. In other partial embodiments, sensors are installed on the air inlet 220 corresponding to the temperature control device 110 in the computer room to detect the cold air temperature or the cold air flow rate of the temperature control device 110.

[0061] Continuing from the above, the status data of the server may be the current temperature of the server. The analysis device 120 can respectively obtain the current temperature of each server D1-D10 through the sensors S1-S10. In addition, the status data of the server may further include the load power, the current temperature, and / or the fan speed. In Figure 1B In the illustrated embodiment, the sensors are arranged at positions adjacent to the servers within the cabinet 200. However, in other embodiments, the sensors may also be arranged inside the corresponding servers.

[0062] In addition, the status data of the server may further include the distance between the server and the temperature control device. For example, the first status data of the first server D1 may include the first distance T1 between the first server D1 and the temperature control device 110. Similarly, if the temperature of the second server D2 is predicted through the second learning model, the second status data obtained by the analysis device 120 may include the second distance between the second server D2 and the temperature control device 110.

[0063] In step S204, the analysis device 120 inputs the temperature control state data, the state data of the server to be predicted and the adjacent servers into the corresponding learning model as input features. The learning model performs operations on these input features according to the established parameters or weight values to output a temperature prediction value. For example, the analysis device 120 inputs the temperature control state data, the first state data of the first server D1, and the second state data of the second server D2 into the first learning model to obtain a first temperature prediction value (e.g., the predicted temperature for 1 minute).

[0064] Similarly, the analysis device 120 inputs the temperature control state data, the first state data of the first server D1, and the second state data of the second server D2 into the second learning model to obtain a second temperature prediction value. In other embodiments, since both the first server D1 and the third server D3 are adjacent to the second server D2, the analysis device 120 also inputs the current temperature of the third server D3 as the third state data into the second learning model to obtain a second temperature prediction value.

[0065] After the analysis device 120 obtains the corresponding temperature prediction values from each learning model, the analysis device 120 calculates the overall temperature prediction value of the computer cabinet 200 based on these temperature prediction values. For example, the analysis device 120 calculates or predicts the overall temperature prediction value of the computer cabinet 200 based on the first temperature prediction value and the second temperature prediction value. The analysis device 120 can calculate the temperature distribution of the entire computer cabinet 200 according to the interpolation method, or can also use the obtained highest / lowest temperature prediction value as the overall temperature prediction value.

[0066] In step S205, after obtaining the overall temperature prediction value, the analysis device 120 determines whether the overall temperature prediction value exceeds the preset temperature range, and adjusts the temperature control device 110 accordingly. If the overall temperature prediction value is higher than the preset temperature upper limit, in step S206, the analysis device 120 controls the temperature control device 110 to lower the cold air temperature or increase the cold air flow rate. Conversely, if the overall temperature prediction value is lower than the preset temperature lower limit, in step S207, the analysis device 120 controls the temperature control device 110 to temporarily stop operating. Accordingly, the temperature of the computer cabinet 200 can be prevented from exceeding the preset temperature range.

[0067] Figure 3The figure shows a schematic diagram of a distributed learning model according to some embodiments of the present application. As shown in the figure, a plurality of learning models M1 to Mn are built into the temperature control system 100. Each of the learning models M1 to Mn corresponds to a server and includes a feature extraction module Ma, a training module Mb, and a temperature prediction module Mc. The feature extraction module Ma is used to obtain input features Xt from the temperature control device 110 and sensors S1 to S10, and extract the required data. For example, the first learning module M1 extracts the temperatures of the first server D1 and the second server D2 as input features.

[0068] If the input feature Xt corresponds to a known output target (i.e., the temperature after a period of time), the learning models M1 to Mn adjust the training module Mb according to the input feature (i.e., the operating state). If the extracted input feature Xt does not have a corresponding output target, the learning models M1 to Mn input the input feature (i.e., the status data) into the temperature prediction module Mc to predict the temperature. As shown in the figure, each of the learning models M1 to Mn outputs a corresponding output target Y1 to Yn. The output targets Y1 to Yn are the server temperatures predicted by each of the learning models M1 to Mn. According to the output targets Y1 to Yn, the analysis device 120 can calculate the overall temperature prediction value Yt of the cabinet 200.

[0069] In one embodiment of the present application, the temperature control method can be applied to a server, which can be used for artificial intelligence (AI) operations, edge computing, and can also be used as a 5G server, a cloud server, or a vehicle networking server.

[0070] The various elements, method steps, or technical features in the foregoing embodiments can be combined with each other, and are not limited by the order of text description or the order of diagram presentation in the present application.

[0071] Although the present application has been disclosed as above in embodiments, it is not intended to limit the present application. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present application. Therefore, the protection scope of the present application shall be determined by the scope defined in the appended claims.

Claims

1. A temperature control method, characterized in that, Including: Generate an air circulation for a first server and a second server through a temperature control device, where the first server and the second server are arranged in a cabinet; Through an analysis device, continuously monitor the operating states of the temperature control device, the first server and the second server as a plurality of first input features, and use the temperature change of the first server as an output target for deep learning to establish a first learning model; Through the analysis device, continuously monitor the operating states of the temperature control device, the first server and the second server as a plurality of second input features, and use the temperature change of the second server as an output target for deep learning to establish a second learning model; Through the analysis device, receive a temperature control status data of the temperature control device, a first status data of the first server and a second status data of the second server, where the first status data includes a first temperature of the first server, and the second status data includes a second temperature of the second server; Through the analysis device, input the temperature control status data, the first status data and the second status data into the first learning model and the second learning model to obtain a first temperature prediction value output by the first learning model and a second temperature prediction value output by the second learning model; And Through the analysis device, calculate an overall temperature prediction value of the cabinet according to the first temperature prediction value and the second temperature prediction value, and adjust the temperature control device according to the overall temperature prediction value.

2. The temperature control method according to claim 1, characterized in that, Where the method for continuously monitoring the operating states of the temperature control device, the first server and the second server includes: Through the analysis device, obtain a cold air temperature or a cold air flow rate of the temperature control device in a detection period; and Through the analysis device, obtain a plurality of operating temperatures when the first server and the second server are operating in the detection period.

3. The temperature control method according to claim 1, characterized in that Where The method for adjusting the temperature control device according to the first temperature prediction value includes: Adjust a cold air temperature or a cold air flow rate of the temperature control device according to the overall temperature prediction value.

4. The temperature control method according to claim 3, characterized in that, It further includes: Through the analysis device, when the overall temperature prediction value is higher than a preset temperature upper limit, reduce the cold air temperature or increase the cold air flow rate; And Through the analysis device, when the overall temperature prediction value is lower than a preset temperature lower limit, temporarily stop operating the temperature control device.

5. A temperature control system, applicable to a data center, characterized in that, Including: A temperature control device for generating an air circulation for a first server and a second server; A first sensor for detecting the operating state of the first server and obtaining a first status data; A second sensor for detecting the operating state of the second server and obtaining a second status data; and An analysis device is electrically connected to the temperature control device, the first sensor, and the second sensor, and is configured to continuously monitor the operating states of the temperature control device, the first server, and the second server as a plurality of first input features, and perform deep learning with a temperature change of the first server as an output target to establish a first learning model, and continuously monitor the operating states of the temperature control device, the first server, and the second server as a plurality of second input features, and perform deep learning with a temperature change of the second server as an output target to establish a second learning model; wherein the analysis device is further configured to input a temperature control state data of the temperature control device, a first temperature of the first state data, and a second temperature of the second state data into the first learning model and the second learning model to obtain a first temperature prediction value output by the first learning model and a second temperature prediction value output by the second learning model, and the analysis device is configured to calculate an overall temperature prediction value of the cabinet according to the first temperature prediction value and the second temperature prediction value, and adjust the temperature control device according to the overall temperature prediction value.

6. The temperature control system according to claim 5, characterized in that, Wherein the analysis device is configured to detect a cold air temperature or a cold air flow rate of the temperature control device in a detection period, and obtain a plurality of operating temperatures when the first server and the second server are operating, so as to establish the first learning model.

7. The temperature control system according to claim 5, wherein Wherein according to the overall temperature prediction value, a cold air temperature or a cold air flow rate of the temperature control device is adjusted; wherein the cabinet has a plurality of placement spaces for placing the first server and the second server.

8. The temperature control system according to claim 7, characterized in that, Wherein when the overall temperature prediction value is higher than a preset temperature upper limit, the analysis device is configured to control the temperature control device to lower the cold air temperature, or control the temperature control device to increase the cold air flow rate; and Wherein when the overall temperature prediction value is lower than a preset temperature lower limit, the analysis device is configured to temporarily stop operating the temperature control device.

Images