Temperature and humidity control method and device, computer device, and storage medium

Abstract

The embodiment of the application provides a kind of temperature and humidity control method, device, computer equipment and storage medium, method includes: in the case where it is detected that at least one cabinet load in target area exists change, obtain changed cabinet load;According to the current temperature and humidity control parameter of target area and changed cabinet load, predict the temperature and humidity of temperature and humidity monitoring point in target area after the change of cabinet load;According to the temperature and humidity of temperature and humidity monitoring point in target area after the change of cabinet load and changed cabinet load, determine the target temperature and humidity control parameter of target area.The change of target area temperature and humidity is predicted based on the first change of load, and the corresponding temperature and humidity control parameter is predicted based on the change of target area temperature and humidity, to realize the accurate control of the temperature and humidity of target area after the change of load, to avoid the case that temperature and humidity local anomaly is caused by load up and down rack, to ensure the safe operation of target area.

Description

Technical Field

[0001] This application relates to the field of computer room equipment control, and in particular to a temperature and humidity control method, device, computer equipment, and storage medium. Background Technology

[0002] Data center server rooms play a crucial role in modern information technology architecture, primarily consisting of servers and storage devices, network equipment, power systems, cooling systems, security systems, management software, and network connectivity. Servers and storage devices generate significant heat during operation. Excessive heat can degrade equipment performance and even cause hardware failures. Prolonged high-temperature operation accelerates the aging of electronic components and shortens equipment lifespan. Furthermore, excessive humidity within the server room can cause condensation on equipment surfaces, leading to short circuits and corrosion, damaging sensitive electronic components. Conversely, low humidity can generate static electricity, which can damage electronic equipment and degrade the performance of insulation materials, resulting in equipment malfunctions.

[0003] During the initial construction of a data center, computational fluid dynamics (CFD) simulation tools are typically used to model the room under different loads and temperature settings to ensure safe operation. However, as the data center operates over time, and with the addition and removal of racks, the temperature directly impacts its condition. Parameters such as the rack placement of servers or storage devices and the load rate of different racks can lead to localized hotspots or high humidity levels, which can pose significant operational safety risks over time. Therefore, precise control of temperature and humidity in the data center, coinciding with the addition and removal of racks, is crucial for its safe operation. Summary of the Invention

[0004] This application provides a temperature and humidity control method, apparatus, computer equipment, and storage medium to solve the problem of local temperature and humidity changes caused by the mounting and dismounting of servers or storage devices in a data center.

[0005] In a first aspect, embodiments of this application provide a temperature and humidity control method, the method comprising:

[0006] Real-time monitoring of rack load in the target area; if at least one rack load changes, the changed rack load is obtained.

[0007] Based on the changed rack load and the current temperature and humidity control parameters of the target area, predict the temperature and humidity of the temperature and humidity monitoring points in the target area after the change in rack load.

[0008] Based on the changed rack load and the temperature and humidity at the monitoring points in the target area after the change in rack load, the target temperature and humidity control parameters for the target area are determined.

[0009] In the method provided in the above embodiments, through two prediction processes, the first prediction is based on the change in load to predict the change in temperature and humidity of the target area, and the second prediction is based on the change in temperature and humidity of the target area to predict the corresponding temperature and humidity control parameters, so as to achieve precise control of the temperature and humidity of the target area after the load change, avoid the situation of local abnormal temperature and humidity caused by the load being put on or taken off, and ensure the safe operation of the target area.

[0010] In one possible implementation, based on the changed rack load and the current temperature and humidity control parameters of the target area, the temperature and humidity at the monitoring points within the target area are predicted after the change in rack load, including:

[0011] The changed rack load and the current temperature and humidity control parameters of the target area are input into the first prediction model to obtain the temperature and humidity of the temperature and humidity monitoring points in the target area after the change of rack load.

[0012] The first prediction model is trained based on historical and simulated operational data of the target area. The historical and simulated operational data include cabinet load, temperature and humidity of temperature and humidity monitoring points, and temperature and humidity control parameters.

[0013] The method provided in the above embodiments uses a multi-dimensional input-output model for prediction, which can accurately predict the temperature and humidity changes at each temperature and humidity monitoring point.

[0014] In one possible implementation, the simulation data is obtained by simulating the operating state of the target area under extreme conditions.

[0015] The method provided in the above embodiments combines historical operating data and simulation operating data to make the training data of the model more comprehensive, provide boundary data for model training, and enable the first prediction model after training to be applicable to a variety of working conditions.

[0016] In one possible implementation, the extreme operating conditions include at least one of the following four operating conditions: the rack load is less than a first load threshold and the temperature and humidity control parameter is less than a first parameter threshold; the rack load is less than the first load threshold and the temperature and humidity control parameter is greater than a second parameter threshold; the rack load is greater than the second load threshold and the temperature and humidity control parameter is less than the first parameter threshold; and the rack load is greater than the second load threshold and the temperature and humidity control parameter is greater than the second parameter threshold.

[0017] The method provided in the above embodiments directly determines extreme operating conditions through cabinet load and temperature and humidity control parameters, intuitively reflecting the operating status and temperature and humidity control status of the target area.

[0018] In one possible implementation, the training process of the first prediction model includes:

[0019] For each sample in the historical and simulation data, data preprocessing is performed. Data preprocessing includes at least one of the following: data deduplication, missing value removal, outlier removal, and data normalization.

[0020] The preprocessed historical running data is divided into a training set and a test set, and the preprocessed simulation running data is used as the training set.

[0021] The first prediction model is trained based on each sample in the training set to obtain the trained first prediction model.

[0022] The first prediction model after training is validated based on the test set. If the validation is successful, the first prediction model is determined.

[0023] In the method provided in the above embodiments, the training effect of the first prediction model is improved by preprocessing the data and dividing the dataset.

[0024] In one possible implementation, both historical running data and simulation running data further include floor opening; correspondingly, the first prediction model is trained based on each sample in the training set, including:

[0025] The rack load, temperature and humidity control parameters, and floor opening of each sample in the training set are used as inputs to the first prediction model, and the temperature and humidity of the temperature and humidity monitoring points in each sample are used as sample labels to train the first prediction model.

[0026] In the method provided in the above embodiments, the first prediction model is predicted in combination with the floor opening degree. In practical application scenarios, the floor opening degree parameter can be incorporated to improve the accuracy of temperature and humidity control in the target area.

[0027] In one possible implementation, based on the changed rack load and the temperature and humidity at monitoring points within the target area after the change in rack load, the target temperature and humidity control parameters for the target area are predicted, including:

[0028] The changed rack load and the temperature and humidity monitoring points in the target area after the rack load change are input into the second prediction model to obtain the target temperature and humidity control parameters for the target area.

[0029] The second prediction model is trained based on historical and simulated operational data of the target area. The historical and simulated operational data include cabinet load, temperature and humidity at temperature and humidity monitoring points, and temperature and humidity control parameters.

[0030] In the method provided in the above embodiments, the target temperature and humidity control parameters of the target area can be accurately predicted by performing secondary prediction through a multi-dimensional input-output model.

[0031] In one possible implementation, the method further includes:

[0032] Based on the location of temperature and humidity monitoring points, the temperature and humidity of the monitoring points in the target area after the rack load changes are presented in the target area map as a cloud map.

[0033] The method provided in the above embodiments enables the visualization of prediction results.

[0034] Secondly, embodiments of this application provide a temperature and humidity control device, the device comprising:

[0035] The detection module is used to monitor the rack load of the target area in real time. When it detects that the load of at least one rack has changed, it obtains the changed rack load.

[0036] The first determining module is used to predict the temperature and humidity of the temperature and humidity monitoring points in the target area after the change in the rack load, based on the changed rack load and the current temperature and humidity control parameters of the target area.

[0037] The second determining module is used to determine the target temperature and humidity control parameters of the target area based on the changed cabinet load and the temperature and humidity of the temperature and humidity monitoring points in the target area after the change in cabinet load.

[0038] Thirdly, embodiments of this application provide a computer device, including: a memory and a processor; the memory stores computer execution instructions; the processor executes the computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0039] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0040] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0041] The temperature and humidity control method, apparatus, computer equipment, and storage medium provided in this application include: when a change in the load of at least one rack in a target area is detected, acquiring the changed rack load; predicting the temperature and humidity of the monitoring points in the target area after the change in rack load based on the changed rack load and the current temperature and humidity control parameters of the target area; and determining the target temperature and humidity control parameters of the target area based on the changed rack load and the temperature and humidity of the monitoring points in the target area after the change in rack load. Through two prediction processes—the first predicting the change in temperature and humidity of the target area based on the changed load, and the second predicting the corresponding temperature and humidity control parameters based on the change in temperature and humidity of the target area—precise control of the temperature and humidity of the target area after the load change is achieved, avoiding localized temperature and humidity anomalies caused by load changes and ensuring the safe operation of the target area. Attached Figure Description

[0042] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0043] Figure 1 A schematic diagram illustrating a scenario for the temperature and humidity control method provided in this application;

[0044] Figure 2 Flowchart of the temperature and humidity control method provided in this application Figure 1 ;

[0045] Figure 3 Flowchart 2 of the temperature and humidity control method provided in this application;

[0046] Figure 4 Flowchart of the temperature and humidity control method provided in this application Figure 3 ;

[0047] Figure 5 A schematic diagram of the temperature and humidity control device provided in this application;

[0048] Figure 6 A schematic diagram of the structure of the computer device provided in this application.

[0049] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0050] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0051] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. It is understood that the terms “first,” “second,” etc., as used herein may be used to describe various information or data, but these elements are not limited by these terms. These terms are only used to distinguish first information from another type of information. For example, without departing from the scope of this application, first action information may be referred to as second action information, and similarly, second action information may be referred to as first action information. Both first action information and second action information are action information, but they are not the same action information.

[0052] The temperature and humidity control method provided in this application can be applied to scenarios where the data center is operating in real time. According to the method provided in this application, the required temperature and humidity control parameters can be determined after a change in the load on the server racks in a target area of ​​the data center. It can also be applied to simulation control scenarios of the target area; that is, before the load on the target area needs to be adjusted, the required temperature and humidity control parameters can be predicted according to the method provided in this application, thereby directly adjusting the temperature and humidity control parameters when the server rack load changes. The following will describe the embodiments of this application using an example of a real-time data center operation scenario.

[0053] The temperature and humidity control method provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, terminal 102 acquires operational data of the target area, including data on load, temperature and humidity, floor opening, and temperature and humidity control parameters. Terminal 102 communicates with server 104 via a network, sending the acquired operational data to server 104. Server 104 processes the data to obtain the target temperature and humidity control parameters, thereby regulating and controlling the temperature and humidity of the target area. A data storage system can store the data processed by server 104. The data storage system can be integrated onto server 104 or located in the cloud or on other network servers.

[0054] The terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, and smart in-vehicle systems. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted devices. The server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0055] In one embodiment, such as Figure 2 As shown, a temperature and humidity control method is provided, which can be applied to... Figure 1 Taking server 104 as an example, the method includes:

[0056] Step 202: Monitor the rack load of the target area in real time. If a change in the rack load of at least one rack is detected, obtain the changed rack load.

[0057] Step 204: Based on the changed rack load and the current temperature and humidity control parameters of the target area, predict the temperature and humidity of the monitoring points in the target area after the change in rack load.

[0058] Step 206: Determine the target temperature and humidity control parameters for the target area based on the changed rack load and the temperature and humidity of the monitoring points in the target area after the change in rack load.

[0059] In a data center, server racks are the infrastructure used to install and support various IT devices (such as servers, storage devices, network devices, etc.). Typically, a data center includes multiple server racks. Rack load refers to the power consumption or load generated by all IT devices within a rack during operation. The target area is a portion or all of the data center, and the monitored rack load is the rack load within that target area. Correspondingly, the monitored rack load represents some or all of the rack loads within the data center. Within the target area, the number of racks whose load changes at any given time can be one or more. Furthermore, the rack load change in this embodiment includes both increases and decreases. The changed rack load refers to the load of all racks within the target area after a change in the load of at least one rack is detected.

[0060] In this embodiment, the target area refers to a target area where the server racks are already in operation. This target area is equipped with temperature and humidity acquisition devices, temperature and humidity control devices, and IT loads. For example, temperature and humidity are collected in the target area using temperature and humidity sensors. Since the target area is large and there are usually multiple server racks being monitored, in order to achieve more accurate monitoring of the temperature and humidity in the target area, multiple temperature and humidity monitoring points are usually set up in the target area, distributed in various locations within the target area. Each temperature and humidity monitoring point is equipped with a temperature and humidity sensor.

[0061] Accordingly, the temperature and humidity monitoring points in the target area mentioned in step 204 can be some of the temperature and humidity monitoring points in the computer room, or all the temperature and humidity monitoring points in the entire computer room, depending on the scope of the target area.

[0062] Temperature and humidity control devices are devices capable of regulating the temperature and humidity within a target area, such as air conditioners, dehumidifiers, humidifiers, variable air volume systems, and fresh air handling units. In one embodiment, the floor in the target area also functions as a temperature and humidity control device; by changing the floor's opening degree, the ventilation within the target area can be altered, thereby adjusting the temperature and humidity. Correspondingly, temperature and humidity control parameters refer to the control parameters of the temperature and humidity control device, such as the supply air temperature, return air temperature, and return air humidity of an air conditioner. Multiple temperature and humidity control devices may be present in the target area.

[0063] Similarly, the current temperature and humidity control parameters of the target area mentioned in step 204 can refer to the control parameters of all temperature and humidity control devices in the computer room, or to the control parameters of some temperature and humidity control devices, depending on the range of the target area.

[0064] It should be noted that all subsequent explanations are based on the target area. When the target area is a part of the data center, the monitored rack load, the corresponding temperature and humidity monitoring points, and the temperature and humidity control devices that need to be adjusted are also partial. Accordingly, the data involved later will be data from a partial rack, a partial temperature and humidity monitoring point, or a partial temperature and humidity control device; otherwise, all data will be used.

[0065] For target areas that are already in operation, the load of the cabinets within the target area is monitored in real time. When a change in cabinet load is detected, the temperature and humidity of the target area environment after the load change are predicted based on the changed cabinet load and the current temperature and humidity control parameters that have not yet been adjusted. The temperature and humidity control parameters are then adjusted based on the predicted temperature and humidity to reduce the temperature and humidity changes caused by load changes in actual application scenarios.

[0066] In one embodiment, the temperature and humidity control parameters of all temperature and humidity control devices in the computer room can be adjusted more comprehensively based on the load of all cabinets, thereby achieving precise control of the temperature and humidity of the entire computer room. Therefore, the target area is the entire computer room. In this case, for a computer room with m cabinets, the cabinet load in the target area is the load of the m cabinets [b1,b2,...,bm]. The load of the m cabinets [b1,b2,...,bm] is collected. Taking the detection that b1 changes from 1kW to 10kW as an example, the load of all cabinets before detection is [1,b2,...,bm], and the load of all cabinets after the change is [10,b2,...,bm] (load unit is kW).

[0067] When determining temperature and humidity based on the changed rack load and current temperature and humidity control parameters, methods such as functions, prediction models, and lookup tables can be used. These methods are based on various data from the rack over a historical period. For example, in one embodiment, historical operating data is used to determine the mapping relationship between the rack load, current temperature and humidity control parameters, and temperature and humidity after each rack load change, thus establishing a lookup table. In practical applications, the corresponding temperature and humidity values ​​are retrieved from the lookup table. Alternatively, historical operating data can be used to determine the function relating temperature and humidity changes to the rack load and temperature and humidity control parameters. In practical applications, the amount of temperature and humidity change is determined using the changed rack load and the current temperature and humidity control parameters of the target area, thereby determining the temperature and humidity.

[0068] After determining the temperature and humidity changes in the target area due to the load change based on the changed rack load and the current temperature and humidity control parameters, the corresponding target temperature and humidity control parameters can be determined to adjust the temperature and humidity within the target area. In one embodiment, the optimal adjustment strategy is selected through a trained model to determine the target temperature and humidity control parameters.

[0069] The method provided in the above embodiments monitors the rack load of a target area in real time. When a change in the rack load is detected, the changed rack load is obtained. Based on the changed rack load and the current temperature and humidity control parameters of the target area, the temperature and humidity of the target area after the rack load change are predicted. Based on the changed rack load and the temperature and humidity of the target area after the rack load change, the target temperature and humidity control parameters of the target area are determined. Through two prediction processes—the first predicting the change in temperature and humidity of the target area based on the changed load, and the second predicting the corresponding temperature and humidity control parameters based on the change in temperature and humidity of the target area—precise control of the temperature and humidity of the target area after the load change is achieved, avoiding localized temperature and humidity anomalies caused by load changes and ensuring the safe operation of the target area.

[0070] In one embodiment, based on the changed rack load and the current temperature and humidity control parameters of the target area, the temperature and humidity at the monitoring points within the target area are predicted after the change in rack load, including:

[0071] The changed rack load and the current temperature and humidity control parameters of the target area are input into the first prediction model to obtain the temperature and humidity of the temperature and humidity monitoring points in the target area after the change of rack load.

[0072] The first prediction model is trained based on historical and simulated operational data of the target area. The historical and simulated operational data include rack load, temperature and humidity, and temperature and humidity control parameters.

[0073] Because the target area has multiple temperature and humidity sensors, multiple server racks, and multiple temperature and humidity control devices, using functions or lookup tables is not suitable for the complex scenario. Therefore, a predictive model can be built to predict the temperature and humidity of the target area after changes in server rack load. For example, a first predictive model can be built based on multi-dimensional input-output Support Vector Regression (SVR), taking the changed server rack load and the current temperature and humidity control parameters of the target area as inputs to predict the temperature and humidity of the target area after changes in server rack load.

[0074] In this embodiment, the first prediction model is obtained by training with historical operating data and simulated operating data. The historical operating data is determined by collecting and determining the actual operating data of the target area within a historical time period, while the simulated operating data is obtained by simulating the operating state of the target area under extreme conditions.

[0075] Compared to historical operating data, simulation operating data can obtain data under operating conditions not involved in actual application scenarios. For example, operating data under extreme operating conditions, which include at least one of the following four conditions: the rack load is less than a first load threshold and the temperature and humidity control parameters are less than a first parameter threshold; the rack load is less than the first load threshold and the temperature and humidity control parameters are greater than a second parameter threshold; the rack load is greater than the second load threshold and the temperature and humidity control parameters are less than the first parameter threshold; and the rack load is greater than the second load threshold and the temperature and humidity control parameters are greater than the second parameter threshold.

[0076] Operating data under extreme conditions is unavailable in real-world applications. Examples include operating data under conditions such as zero rack load and minimum temperature and humidity control parameters, zero rack load and maximum temperature and humidity control parameters, maximum rack load and minimum temperature and humidity control parameters, and maximum rack load and maximum temperature and humidity control parameters. Furthermore, to ensure the normal operation of racks in the target area, these conditions are typically not encountered in real-world applications.

[0077] Specifically, simulation software, such as Energy Plus, can be used to simulate various extreme operating conditions in the target area. These extreme conditions include, but are not limited to, extreme rack loads and extreme control parameters of temperature and humidity control devices. The simulation data under these extreme conditions is used to train the first prediction model, providing edge data for model training. Therefore, the more extreme conditions involved in the simulation, the more accurate the results will be. In one embodiment, the extreme conditions include at least the four types mentioned above.

[0078] It should be noted that historical operational data can be collected at a first fixed time interval during operation in the target area, while simulation operational data is collected at a second fixed time interval during the simulated operating conditions. The first and second fixed time intervals can be the same or different, and the data dimensions of the collected historical and simulation operational data are consistent. For example, both historical and simulation operational data include the temperature and humidity control parameters of n air conditioners, the load of m server racks, k floor openings, and the temperature and humidity data collected by j temperature and humidity sensors.

[0079] The method provided in this application combines historical operating data and simulation operating data to make the training data of the model more comprehensive, provide boundary data for model training, and enable the first prediction model after training to be applicable to various working conditions.

[0080] In one embodiment, such as Figure 3 As shown, the training process of the first prediction model includes:

[0081] Step 302: Perform data preprocessing on each sample in the historical running data and simulation running data. Data preprocessing includes at least one of data deduplication, missing value removal, outlier data removal, and data normalization.

[0082] Step 304: Divide the preprocessed historical running data into a training set and a test set, and use the preprocessed simulation running data as the training set.

[0083] Step 306: Train the first prediction model for each sample in the training set to obtain the trained first prediction model.

[0084] Step 308: Validate the trained first prediction model based on the test set. If the validation is successful, determine the first prediction model.

[0085] For the collected historical and simulation data, each collected data point is treated as a sample. All samples undergo preprocessing to improve data validity. This preprocessing includes:

[0086] (1) Data deduplication: Keep one set of data for all data with the same values ​​in all dimensions to eliminate the impact of duplicate data on the overall training results;

[0087] (2) Remove missing values: Remove missing data. Since the collected data may contain some missing data, the data with some missing data is directly removed.

[0088] (3) Outlier removal: Outlier detection is performed on the data, and outliers are removed. For example, the Z-score method can be used for outlier detection. For each point... Calculate its Z value. If the absolute value of Z is greater than a preset threshold, such as 3, then the value is considered an outlier. The specific calculation formula is as follows:

[0089] ,

[0090] in To represent a certain data, Indicates the number of data points. Indicates the first Data.

[0091] (4) Data normalization: To eliminate the impact of excessive data differences on the training model, the data is normalized. The normalization formula is as follows:

[0092] ;

[0093] in: This represents the normalized data. This represents the data before normalization. This represents the maximum value in the data before normalization. This represents the minimum value in the data before normalization.

[0094] In training the first prediction model, all collected data needs to be divided into training and test sets. However, since simulation data has a significant impact on training, it is not used as the test set. Specifically, a portion of the preprocessed historical data is randomly selected as the test set, and the remaining data, along with the simulation data, is used as the training set.

[0095] In one embodiment, k-fold cross-validation can be used to divide the test and training sets. Specifically, the preprocessed historical running data and simulation running data sets (without order) are divided into K subsets of equal size. One subset is used as the test set in turn, and the rest are used as the training set. This process is repeated K times, and the final results are averaged. The subset containing simulation running data samples is not used as the test set. Specifically, each time, a subset can be randomly selected from the historical running data as the test set, and then the remaining historical running data and simulation running data can be divided into K-1 subsets as the training set.

[0096] In the method provided in the above embodiments, the training effect of the first prediction model is improved by preprocessing the data and dividing the dataset.

[0097] In one embodiment, both historical running data and simulation running data further include floor opening; correspondingly, the first prediction model is trained based on each sample in the training set, including:

[0098] The first prediction model is trained by using all cabinet loads, temperature and humidity control parameters, and floor openings of each sample in the training set as inputs, and the temperature and humidity of all temperature and humidity monitoring points in each sample as sample labels.

[0099] For a data center with n air conditioners, k floor openings, and m rack loads, and the target area is the entire data center, each sample in the training set includes the supply air temperature [as1,as2,as3,......,asn], return air temperature [ar1,ar2,ar3,......,arn], return air humidity [am1,a2,am3,......,amn], power on / off status [ao1,a2,ao3,......,aon], m rack loads [b1,b2,b3,......,bm], k floor openings [c1,c2,c3,......,ck], and the temperature [d1,d2,d3,......,dj] and humidity [e1,e2,e3,......,ej] of j temperature and humidity monitoring points.

[0100] In the first prediction model, the inputs are the supply air temperature [as1,as2,as3,......,asn], return air temperature [ar1,ar2,ar3,......,arn], return air humidity [am1,a2,am3,......,amn], on / off status [ao1,a2,ao3,......,aon] of n air conditioners, the load of m server racks [b1,b2,b3,......,bm], and k floor openings [c1,c2,c3,......,ck]. The outputs are j predicted values ​​for temperature and humidity. During training, the temperature [d1,d2,d3,......,dj] and humidity [e1,e2,e3,......,ej] of j temperature and humidity monitoring points in each sample are used as sample labels and compared with the model output to complete model training.

[0101] In one embodiment, taking the SVR algorithm as an example of the first prediction model, the function kernel is a radial basis function kernel (RBF kernel) because the radial basis function kernel has good performance in processing highly nonlinear data; the regularization parameter C and the penalty coefficient ε are performed using a grid search method, where the grid search method systematically tries all possible combinations by defining the range of values ​​for C and ε. The range of C is 0.1-10 with a step size of 0.1, and the range of ε is 0.001-1 with a step size of 0.001. Finally, the set of parameters with the smallest error is obtained as the final model parameters.

[0102] Alternatively, mean squared error (MSE) can be used as an evaluation parameter for the model. The smaller the MSE, the better the model's predictive performance. The calculation formula is shown below:

[0103] ,

[0104] in: Indicates the number of samples. Represents the actual value. The predicted values ​​are represented by ε. Finally, the model is trained using the set of C and ε that minimizes the error, and this trained model is used as the first prediction model to predict temperature and humidity.

[0105] In one embodiment, such as Figure 4 As shown, based on the changed rack load and the temperature and humidity of the target area after the change in rack load, the target temperature and humidity control parameters for the target area are predicted, including:

[0106] Step 402: Input the changed rack load and the temperature and humidity monitoring points in the target area after the rack load change into the second prediction model to obtain the target temperature and humidity control parameters of the target area.

[0107] Step 404: The second prediction model is obtained by training based on historical operating data and simulated operating data of the target area. The historical operating data and simulated operating data include cabinet load, temperature and humidity of temperature and humidity monitoring points, and temperature and humidity control parameters.

[0108] The second prediction model is used to determine the target temperature and humidity control parameters based on the changed rack load and the temperature and humidity of the target area after the rack load change. The second prediction model can use the same algorithm as the first prediction model, and is trained using the same historical and simulation data, which will not be elaborated here.

[0109] However, it should be noted that when using the loads of m racks [b1,b2,b3,......,bm] and the temperatures [d1,d2,d3,......,dj] and humidity [e1,e2,e3,......,ej] of j temperature and humidity monitoring points as inputs to the second prediction model, if there are closed cold and hot aisle temperatures, then all closed cold aisle temperatures will be set to 25, closed hot aisle temperatures will be set to 30, and humidity will be uniformly set to 50. If there are no closed cold and hot aisles, then the temperature can be uniformly set to 25 and the humidity can be uniformly set to 50.

[0110] In addition, when using the first and second prediction models, since the input data of the models has been normalized, the output needs to be denormalized to obtain the final output result.

[0111] In one embodiment, the method further includes:

[0112] Based on the location of the temperature and humidity monitoring points, the temperature and humidity of the monitoring points after the rack load changes are presented in the target area map as a cloud map.

[0113] In the process of temperature and humidity control, to make the prediction results visible, the predicted temperature and humidity data from all temperature and humidity monitoring points can be presented in the form of a temperature and humidity cloud map, combined with the location of the monitoring points. That is, in a two-dimensional or three-dimensional model of the target area, the predicted data is displayed as a temperature and humidity cloud map based on the installation location of each temperature and humidity sensor. Different colors in the display map correspond to different temperatures or humidity levels, making it easy to understand at a glance. The target temperature and humidity control parameters can be presented in a tabular format.

[0114] In one specific implementation, the entire data center is considered the target area. During the operation of the data center, historical operational data is collected at fixed time intervals. Data collection is performed on the operating data center, which contains 9 air conditioners and 14 micro-module enclosed cold aisles. Therefore, each set of data collected includes 9 supply air temperature settings, 9 return air temperature settings, and 9 humidity settings. In this actual case, the supply and return air temperatures and humidity settings of the shut-down air conditioners are both 0. The data also includes 318 rack loads, 14 floor openings, 60 temperature data points, and 60 humidity data points.

[0115] The Energy Plus software was used to simulate various conditions in the target area, including at least the following conditions: (1) when the load is all 0, the air supply and return air of the air conditioner are set to the minimum; (2) when the load is all 0, the air supply and return air of the air conditioner are set to the maximum; (3) when the load is all at the maximum, the air supply and return air of the air conditioner are set to the minimum; (4) when the load is all at the maximum, the air supply and return air of the air conditioner are set to the maximum. The dimensions of the collected simulation operation data are the same as the dimensions of the historical operation data.

[0116] Historical and simulated data were merged to obtain a dataset (the order is not shown), containing 383 data points. The data in the dataset was preprocessed, and then divided into training and test sets. This was done using k-fold cross-validation, dividing the dataset into K equal subsets. One subset was used as the test set, and the rest as the training set. This process was repeated K times, and the final results were averaged. Because the simulation data significantly influenced the training of the first and second prediction models, subsets containing simulation data were not used as the test set.

[0117] The training steps for the first prediction model are as follows:

[0118] (1) First, the input and output are determined. The dimension of the input is 359 and the dimension of the output is 120.

[0119] (2) Algorithm parameters are determined. Multi-dimensional input-output SVR is adopted, and radial basis function kernel is used. According to the optimization results, the optimal C value with the minimum error is 0.1 and the optimal ε is 0.01.

[0120] (3) The corresponding minimum model evaluation parameter MSE is 0.089;

[0121] (4) Finally, the model is trained using the set of C and ε with the smallest error, and the first prediction model obtained from the training is saved.

[0122] (5) Call the saved first prediction model and verify it through the test set. The MSE of the verification result is 0.326.

[0123] The training steps for the second prediction model are as follows:

[0124] (1) Determine the input and output dimensions. The input dimension is 438 and the output dimension is 41.

[0125] (2) Algorithm parameters are determined. Multi-dimensional input-output SVR is adopted, and radial basis function kernel is used. According to the optimization results, the optimal C value with the minimum error is 10, and the optimal ε is 0.001.

[0126] (3) The corresponding minimum model evaluation parameter MSE is 0.935;

[0127] (4) Finally, the model is trained using the set of C and ε with the smallest error, and the second prediction model obtained from the training is saved;

[0128] (5) Call the saved second prediction model and verify it through the test set. The MSE of the verification result is 3.927.

[0129] In practical application scenarios, the system first predicts temperature and humidity based on the saved first prediction model, according to the rack load, air supply settings, air return settings, air return humidity, and floor opening settings. Then, based on the saved second prediction model, the system predicts the air supply settings, air return settings, air return humidity, and floor opening settings based on the rack load and temperature and humidity.

[0130] In the above embodiments, the multidimensional input-output SVR model is used twice for prediction. That is, the environmental temperature change of the target area can be accurately predicted based on the load and air conditioning settings, and the air conditioning settings and floor opening of the target area can also be accurately predicted. This embodiment combines the target area operation data with the simulation operation data, which can provide effective boundary data for the SVR model, provide better data support for model training, and make predictions on the environmental temperature and humidity changes of the target area more quickly.

[0131] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0132] Based on the same inventive concept, this application also provides a temperature and humidity control device. The solution provided by this temperature and humidity control device is similar to the solution described in the above method. Therefore, the specific limitations of one or more temperature and humidity control device embodiments provided below can be found in the limitations of the temperature and humidity control method above, and will not be repeated here.

[0133] In one embodiment, such as Figure 5 As shown, the temperature and humidity control device includes a detection module 501, a first determination module 502, and a second determination module 503, wherein:

[0134] The detection module 501 is used to monitor the rack load of the target area in real time, and to obtain the changed rack load when it is detected that at least one rack load has changed.

[0135] The first determining module 502 is used to predict the temperature and humidity of the temperature and humidity monitoring points in the target area after the change in the rack load, based on the changed rack load and the current temperature and humidity control parameters of the target area.

[0136] The second determining module 503 is used to determine the target temperature and humidity control parameters of the target area based on the changed cabinet load and the temperature and humidity of the temperature and humidity monitoring points in the target area after the change of cabinet load.

[0137] In one possible implementation, the first determining module 502 is further configured to:

[0138] The changed rack load and the current temperature and humidity control parameters of the target area are input into the first prediction model to obtain the temperature and humidity of the temperature and humidity monitoring points in the target area after the change of rack load.

[0139] The first prediction model is trained based on historical and simulated operational data of the target area. The historical and simulated operational data include cabinet load, temperature and humidity of temperature and humidity monitoring points, and temperature and humidity control parameters.

[0140] In one possible implementation, the first determining module 502 is further used to determine that the simulation running data is obtained by collecting running data of the target area under extreme operating conditions.

[0141] In one possible implementation, the first determining module 502 is further configured to determine that the extreme operating conditions include at least one of the following four operating conditions: the rack load is less than the first load threshold and the temperature and humidity control parameter is less than the first parameter threshold; the rack load is less than the first load threshold and the temperature and humidity control parameter is greater than the second parameter threshold; the rack load is greater than the second load threshold and the temperature and humidity control parameter is less than the first parameter threshold; and the rack load is greater than the second load threshold and the temperature and humidity control parameter is greater than the second parameter threshold.

[0142] In one possible implementation, the temperature and humidity control device further includes a first training module for:

[0143] For each sample in the historical and simulation data, data preprocessing is performed. Data preprocessing includes at least one of the following: data deduplication, missing value removal, outlier removal, and data normalization.

[0144] The preprocessed historical running data is divided into a training set and a test set, and the preprocessed simulation running data is used as the training set.

[0145] The first prediction model is trained based on each sample in the training set to obtain the trained first prediction model.

[0146] The first prediction model after training is validated based on the test set. If the validation is successful, the first prediction model is determined.

[0147] In one possible implementation, the first training module is also used for:

[0148] The rack load, temperature and humidity control parameters, and floor opening of each sample in the training set are used as inputs to the first prediction model, and the temperature and humidity of the temperature and humidity monitoring points in each sample are used as sample labels to train the first prediction model.

[0149] In one possible implementation, the second determining module 503 is further configured to:

[0150] The changed rack load and the temperature and humidity monitoring points in the target area after the rack load change are input into the second prediction model to obtain the target temperature and humidity control parameters for the target area.

[0151] The second prediction model is trained based on historical and simulated operational data of the target area. The historical and simulated operational data include cabinet load, temperature and humidity at temperature and humidity monitoring points, and temperature and humidity control parameters.

[0152] In one possible implementation, the temperature and humidity control device further includes a display module for:

[0153] Based on the location of temperature and humidity monitoring points, the temperature and humidity of the monitoring points in the target area after the rack load changes are presented in the target area map as a cloud map.

[0154] Each module in the aforementioned temperature and humidity control device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0155] Figure 6 A schematic diagram of the structure of the computer device provided in this application. Figure 6 As shown, the electronic device 60 provided in this embodiment includes at least one processor 601 and a memory 602. Optionally, the device 60 further includes a communication component 603. The processor 601, memory 602, and communication component 603 are connected via a bus.

[0156] In a specific implementation, at least one processor 601 executes computer execution instructions stored in memory 602, causing at least one processor 601 to perform the above-described method.

[0157] The specific implementation process of processor 601 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0158] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0159] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0160] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0161] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0162] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0163] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0164] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0165] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0166] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0167] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0168] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0169] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0170] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for controlling temperature and humidity, characterized in that, The method includes: Real-time monitoring of rack load in the target area; if a change in the load of at least one rack is detected, the changed rack load is obtained. Based on the changed rack load and the current temperature and humidity control parameters of the target area, predict the temperature and humidity of the temperature and humidity monitoring points in the target area after the change in rack load. Based on the changed rack load and the temperature and humidity of the temperature and humidity monitoring points in the target area after the change in rack load, the target temperature and humidity control parameters for the target area are determined.

2. The method according to claim 1, characterized in that, The step of predicting the temperature and humidity at the monitoring points in the target area after the change in rack load, based on the changed rack load and the current temperature and humidity control parameters of the target area, includes: The changed rack load and the current temperature and humidity control parameters of the target area are input into the first prediction model to obtain the temperature and humidity of the temperature and humidity monitoring points in the target area after the change of rack load; The first prediction model is obtained by training based on historical operating data and simulated operating data of the target area. The historical operating data and the simulated operating data include cabinet load, temperature and humidity of temperature and humidity monitoring points, and temperature and humidity control parameters.

3. The method according to claim 2, characterized in that, The simulation data is obtained by simulating the operating state of the target area under extreme conditions.

4. The method according to claim 3, characterized in that, The extreme operating conditions include at least one of the following four operating conditions: the rack load is less than the first load threshold and the temperature and humidity control parameter is less than the first parameter threshold; the rack load is less than the first load threshold and the temperature and humidity control parameter is greater than the second parameter threshold; the rack load is greater than the second load threshold and the temperature and humidity control parameter is less than the first parameter threshold; and the rack load is greater than the second load threshold and the temperature and humidity control parameter is greater than the second parameter threshold.

5. The method according to claim 2, characterized in that, The training process of the first prediction model includes: Each sample in the historical running data and the simulation running data is preprocessed. The data preprocessing includes at least one of data deduplication, missing value removal, outlier data removal, and data normalization. The preprocessed historical running data is divided into a training set and a test set, and the preprocessed simulation running data is used as the training set. The first prediction model is trained based on each sample in the training set to obtain the trained first prediction model. The trained first prediction model is validated based on the test set, and if the validation is successful, the first prediction model is determined.

6. The method according to claim 5, characterized in that, Both the historical operating data and the simulation operating data include floor opening; correspondingly, training the first prediction model based on each sample in the training set includes: The cabinet load, temperature and humidity control parameters, and floor opening of each sample in the training set are used as inputs to the first prediction model, and the temperature and humidity of the temperature and humidity monitoring points in each sample are used as sample labels to train the first prediction model.

7. The method according to claim 1, characterized in that, The step of predicting the target temperature and humidity control parameters for the target area based on the changed rack load and the temperature and humidity at the monitoring points within the target area after the change in rack load includes: The changed rack load and the temperature and humidity of the monitoring points in the target area after the change of rack load are input into the second prediction model to obtain the target temperature and humidity control parameters of the target area. The second prediction model is obtained by training based on historical operating data and simulated operating data of the target area. The historical operating data and the simulated operating data include cabinet load, temperature and humidity of temperature and humidity monitoring points, and temperature and humidity control parameters.

8. The method according to any one of claims 1-7, characterized in that, The method further includes: Based on the location of the temperature and humidity monitoring points, the temperature and humidity of the monitoring points in the target area after the cabinet load changes are presented in the target area map as a cloud map.

9. A temperature and humidity control device, characterized in that, The device includes: The detection module is used to monitor the rack load of the target area in real time. When it detects that the load of at least one rack has changed, it obtains the changed rack load. The first determining module is used to predict the temperature and humidity of the temperature and humidity monitoring points in the target area after the change in the rack load, based on the changed rack load and the current temperature and humidity control parameters of the target area. The second determining module is used to determine the target temperature and humidity control parameters of the target area based on the changed cabinet load and the temperature and humidity of the temperature and humidity monitoring points in the target area after the change in cabinet load.

10. A computer device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1 to 8.

12. A computer program product, characterized in that, Includes computer execution instructions, which, when executed by a processor, implement the method as described in any one of claims 1 to 8.

Images