Server thermal power consumption real-time detection method and system

Abstract

One or more embodiments of the invention provide a server thermal power consumption real-time detection method and system. The method comprises the following steps: collecting operation data of a data center, wherein the data comprises computing power resource use data, refrigeration system operation data and environment influence data; preprocessing the data to remove abnormal values, and storing the preprocessed data in a computing power resource use data set, a refrigeration system operation data set and an environment monitoring data set; according to data in the computing power resource use data set, the refrigerating system operation data set and the environment monitoring data set, computing a computing power load characteristic index, a refrigerating system operation index and an environment monitoring index respectively; calculating an overall energy consumption influence value according to the calculation force load characteristic index, the refrigeration system operation index and the environment influence index; and according to the overall energy consumption influence value, abnormal early warning and distributed management are implemented.

Description

TECHNICAL FIELD

[0001] One or more embodiments of the present disclosure relate to the technical field of server management, and in particular to a server thermal power real-time detection method and system. BACKGROUND

[0002] It should be noted that the above introduction to the technical background is only for the convenience of clearly and completely describing the technical solutions of the present disclosure, and for the convenience of understanding by those skilled in the art. The above technical solutions cannot be considered as known to those skilled in the art merely because they are described in the background section of the present disclosure.

[0003] With the expansion of the scale of data centers, the improvement of computing power and power density, the energy consumption supporting the operation of IT equipment also increases rapidly. The high energy consumption of data computing, storage and exchange is accompanied by a large amount of heat generated by the equipment. More than 99% of the electric energy of IT equipment is converted into heat energy, and 70% of the heat energy must be removed through the data center cooling system, which further increases the power consumption of the data center. In the past five years, the power consumption of data centers has maintained an increase of more than 15%. The power consumption of the refrigeration system of a large data center as an auxiliary device is basically the same as that of the IT equipment, so the energy-saving potential of the refrigeration system is huge.

[0004] The existing server thermal power management has defects of high energy consumption and low energy efficiency, imbalance between heat dissipation and computing power, and insufficient fine management capability. There is a lack of multi-dimensional correlation analysis, which leads to the inability to accurately identify the composite influencing factors of thermal power, the inability to avoid risks in advance, high cost of post-fault processing, and limited energy efficiency optimization space. SUMMARY

[0005] Therefore, the purpose of one or more embodiments of the present disclosure is to propose a server thermal power real-time detection method and system to solve the problems in the background.

[0006] In order to achieve the above purpose, one or more embodiments of the present disclosure provide a server thermal power real-time detection method, which comprises:

[0007] Collecting data of the operation of a data center, the data comprising computing power resource usage data, refrigeration system operation data and environmental influence data;

[0008] Preprocessing the data to eliminate outliers, and saving the preprocessed data to a computing power resource usage dataset, a refrigeration system operation dataset and an environmental monitoring dataset;

[0009] According to the data in the computing power resource usage dataset, the refrigeration system operation dataset and the environmental monitoring dataset, computing power load characteristic indexes, refrigeration system operation indexes and environmental monitoring indexes are respectively calculated;

[0010] According to the computing power load characteristic index, the refrigeration system operation index and the environmental impact index, an overall energy consumption impact value is calculated;

[0011] According to the overall energy consumption impact value, an abnormality early warning and distributed management are implemented.

[0012] Optionally, the pre-processing of the data to eliminate abnormal values comprises:

[0013] A unique standard key is set for the computing power resource usage data, the refrigeration system operation data and the environmental impact data, and data in the computing power resource usage data, the refrigeration system operation data and the environmental impact data that does not conform to the unique standard key is eliminated;

[0014] A general data processing tool is used to eliminate missing values in the computing power resource usage data, the refrigeration system operation data and the environmental impact data;

[0015] A highest standard value and a lowest standard value are set for the computing power resource usage data, the refrigeration system operation data and the environmental impact data, and data higher than the highest standard value and data lower than the lowest standard value are eliminated;

[0016] The eliminated numerical gaps are replaced by historical average values.

[0017] Optionally, it is characterized in that,

[0018] The saving number expression of the computing power resource usage data set is: [S1, S2, S3, ···, Sn], S1 represents the first computing power resource usage data in the computing power resource usage data set, S2 represents the second computing power resource usage data in the computing power resource usage data set, and Sn represents the nth computing power resource usage data in the computing power resource usage data set. n n

[0019] The saving number expression of the refrigeration system operation data set is: [Z1, Z2, Z3, ···, Zn], Z1 represents the first refrigeration system operation data in the refrigeration system operation data set, Z2 represents the second refrigeration system operation data in the refrigeration system operation data set, and Zn represents the nth refrigeration system operation data in the refrigeration system operation data set. n n

[0020] The saving number expression of the environmental monitoring data set is: [H1, H2, H3, ···, Hn], H1 represents the first environmental monitoring data in the environmental monitoring data set, H2 represents the second environmental monitoring data in the environmental monitoring data set, and Hn represents the nth environmental monitoring data in the environmental monitoring data set. n n

[0021] ​​​​​​Optionally, the computing power resource usage data includes: CPU utilization, GPU utilization, memory occupancy, network bandwidth occupancy, read-write throughput, storage device response time, the number of currently running processes, and the number of tasks to be executed.

[0022] The refrigeration system operation data includes: compressor operation data, condenser operation data, evaporator operation data, and fan operation data, and the operation data includes: temperature, humidity, pressure, flow, vibration, and noise.

[0023] The environmental monitoring data includes: ambient temperature, ambient humidity, and dust concentration.

[0024] Optionally, the computing power load characteristic index is calculated according to the data in the computing power resource usage data set, including:

[0025] According to the data in the computing power resource usage data set, the computing power load characteristic index is obtained through the following computing power load characteristic index model:

[0026]

[0027] wherein, SLzs represents the computing power load characteristic index, s i represents the i-th data in the computing power resource usage data, S imax represents the maximum value in the i-th data in the computing power resource usage data, a i represents the weight of the i-th data in the computing power resource usage data.

[0028] Optionally, the refrigeration system operation index is calculated according to the data in the refrigeration system operation data set, including:

[0029] According to the data in the refrigeration system operation data set, the refrigeration system operation index is obtained through the following refrigeration system operation index model:

[0030]

[0031] wherein, ZLzs represents the refrigeration system operation index, ys represents the total value of the data in the compressor operation data that exceeds the standard operation data, ln represents the total value of the data in the condenser operation data that exceeds the standard operation data, zf represents the total value of the data in the evaporator operation data that exceeds the standard operation data, fj represents the total value of the data in the fan operation data that exceeds the standard operation data, and Z o represents the total value of the allowed exceeding value.

[0032] Optionally, the environmental impact index is calculated according to the data in the environmental impact data set, including:

[0033] According to the data in the environmental impact dataset, the environmental impact index is obtained through an environmental impact index model as follows:

[0034]

[0035] wherein HJzs represents the environmental impact index, H w represents the environmental temperature, H wo represents the standard value of the environmental temperature, H s represents the environmental humidity, H so represents the standard value of the environmental humidity, H h represents the dust concentration, H ho represents the standard value of the dust concentration.

[0036] Optionally, the calculation of the overall energy consumption impact value according to the computing power load characteristic index, the refrigeration system operation index and the environmental impact index comprises:

[0037] NHyx = (δ1*ZLzs) + (δ2*ZLzs) + (δ3*HJzs);

[0038] wherein NHyx represents the overall energy consumption impact value, SLzs represents the computing power load characteristic index, ZLzs represents the refrigeration system operation index, HJzs represents the environmental impact index, and δ1, δ2 and δ3 represent the weights corresponding to the computing power load characteristic index, the refrigeration system operation index and the environmental impact index.

[0039] Optionally, the implementation of the abnormal early warning and the distributed management according to the overall energy consumption impact value comprises:

[0040] When the overall energy consumption impact value is greater than an overall energy consumption impact value threshold, the abnormal early warning is implemented.

[0041] When the overall energy consumption impact value is abnormal, the computing power load characteristic index and the refrigeration system operation index are compared with a computing power load characteristic index threshold and a refrigeration system operation index threshold respectively, and the corresponding index greater than the corresponding threshold is extracted to send an early warning signal.

[0042] Based on the same inventive concept, one or more embodiments of the present disclosure also provide a server heat power consumption real-time detection system, comprising:

[0043] a data collection module configured to collect data center operation data, the data comprising computing power resource usage data, refrigeration system operation data and environmental impact data;

[0044] a data processing module configured to pre-process the data to eliminate outliers and save the pre-processed data to a computing power resource usage dataset, a refrigeration system operation dataset and an environmental monitoring dataset.

[0045] The first data calculation module is configured to calculate the computing load characteristic index, the cooling system operation index, and the environmental monitoring index based on the data in the computing power resource usage dataset, the cooling system operation dataset, and the environmental monitoring dataset, respectively.

[0046] The second data calculation module is configured to calculate the overall energy consumption impact value based on the computing power load characteristic index, the refrigeration system operation index and the environmental impact index.

[0047] The data analysis and early warning module is configured to implement anomaly warnings and distributed management based on the overall energy consumption impact value.

[0048] As can be seen from the above, the server thermal power consumption real-time detection method provided by one or more embodiments of this disclosure has the advantages of processing the collected data to remove outliers, ensuring the integrity and continuity of the dataset, reducing the impact of data deviation, effectively improving the accuracy, reliability and availability of data, integrating computing load characteristic index, cooling system operation index and environmental impact index, realizing real-time accurate detection and dynamic distributed management of server thermal power consumption, accurately locating the source of anomalies, thereby triggering targeted early warnings, optimizing resource scheduling, balancing computing power and cooling capacity, and ultimately achieving advantages such as energy efficiency optimization, reduced fault risk, saved operation and maintenance costs, and ensuring the safe and stable operation of data centers.

[0049] The real-time server thermal power consumption detection method disclosed herein has the following beneficial effects:

[0050] By processing the collected data to remove outliers and using historical averages as substitutes, the integrity and continuity of the dataset can be ensured, the impact of data bias can be reduced, and the accuracy, reliability and usability of the data can be effectively improved. This provides high-quality input for subsequent modeling and analysis, enhances the stability and prediction accuracy of the model, and reduces the risk of outliers misleading the results.

[0051] By integrating computing load characteristic index, cooling system operation index, and environmental impact index, the system achieves real-time and accurate detection and dynamic distributed management of server thermal power consumption. When the overall energy consumption impact value exceeds the threshold, the system can quickly identify abnormal thermal power consumption and accurately locate the source of the anomaly by comparing each index with the corresponding threshold, thereby triggering targeted warnings. This mechanism not only improves fault response speed and avoids equipment damage caused by local overheating, but also optimizes resource scheduling through distributed management, balances computing power and cooling capacity, and ultimately achieves energy efficiency optimization, reduced fault risk, and cost savings in operation and maintenance, ensuring the safe and stable operation of the data center.

[0052] The server thermal power real-time detection system provided by the present disclosure can implement the steps of the server thermal power real-time detection method, and therefore has the beneficial effects of the server thermal power real-time detection method. BRIEF DESCRIPTION OF DRAWINGS

[0053] In order to more clearly illustrate the technical solutions in the one or more embodiments of the present disclosure or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are only one or more embodiments of the present disclosure, and other drawings can be obtained by those skilled in the art without creative effort.

[0054] Figure 1 The flowchart of the server thermal power real-time detection method of the one or more embodiments of the present disclosure;

[0055] Figure 2 The structural diagram of the server thermal power real-time detection system of the one or more embodiments of the present disclosure. DETAILED DESCRIPTION

[0056] In order to make the purposes, technical solutions and advantages of the present disclosure clearer, the present disclosure will be further described in detail below with reference to specific embodiments and drawings.

[0057] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the one or more embodiments of the present disclosure should be understood as the general meaning understood by those skilled in the art to which the present disclosure belongs. The terms "first", "second" and the like used in the one or more embodiments of the present disclosure do not represent any order, quantity or importance, but are only used to distinguish different components. The terms "include", "contain" and the like mean that the elements or objects before the terms cover the elements or objects listed after the terms and their equivalents, and do not exclude other elements or objects. The terms "connect" or "connected" and the like do not mean physical or mechanical connection, but can include electrical connection, whether direct or indirect. The terms "up", "down", "left", "right" and the like only represent relative positional relationships, which can change accordingly when the absolute positions of the described objects change.

[0058] Reference Figure 1 The server thermal power real-time detection method of the one or more embodiments of the present disclosure includes the following steps:

[0059] Step S101: Collecting data of the operation of the data center, the data including computing resource usage data, refrigeration system operation data and environmental impact data;

[0060] Step S102: pre-processing the data to eliminate outliers, and saving the pre-processed data to a computing resource usage dataset, a refrigeration system operation dataset, and an environmental monitoring dataset;

[0061] Step S103: calculating a computing resource load characteristic index, a refrigeration system operation index, and an environmental monitoring index according to the data in the computing resource usage dataset, the refrigeration system operation dataset, and the environmental monitoring dataset, respectively;

[0062] Step S104: calculating an overall energy consumption impact value according to the computing resource load characteristic index, the refrigeration system operation index, and the environmental impact index;

[0063] Step S105: implementing an abnormality early warning and distributed management according to the overall energy consumption impact value.

[0064] The data of step S101 includes computing resource usage data, refrigeration system operation data, and environmental impact data.

[0065] In examples of the present disclosure, the computing resource usage data can include CPU utilization, GPU utilization, memory occupancy, network bandwidth occupancy, read-write throughput, storage device response time, number of currently running processes, and number of tasks to be executed.

[0066] The refrigeration system operation data can include compressor operation data, condenser operation data, evaporator operation data, and fan operation data, and the operation data includes temperature, humidity, pressure, flow rate, vibration, and noise.

[0067] The environmental monitoring data can include environmental temperature, environmental humidity, and dust concentration.

[0068] The computing resource usage data reflects the server computing load state, the refrigeration system operation data reveals the device health degree and energy efficiency, and the environmental monitoring data represents the influence of the external environment on the device. By fusing these multi-dimensional data, the server heat power real-time detection and distributed management can accurately identify the dynamic relationship between the computing load and the heat dissipation demand, optimize the refrigeration strategy, and reduce the invalid energy consumption. At the same time, based on the environmental data, potential overheating risks can be early warned, and combined with the distributed management, load balancing and resource scheduling can be realized to avoid local overload or idling, thereby improving the overall energy efficiency, prolonging the device life, reducing the downtime, and ultimately realizing the green and low-carbon operation and sustainable development of the data center.

[0069] In the example disclosed herein, preprocessing the data in step S102 to remove outliers may include: using a general data processing tool to remove missing values ​​for computing resource usage data, cooling system operation data, and environmental impact data; setting a maximum and minimum standard value for computing resource usage data, cooling system operation data, and environmental impact data, and removing data that is higher than the maximum standard value or lower than the minimum standard value; and replacing the missing values ​​with historical average values.

[0070] By processing the collected data to remove outliers and using historical averages as substitutes, the integrity and continuity of the dataset can be ensured, reducing the impact of data bias. This effectively improves the accuracy, reliability, and usability of the data, providing high-quality input for subsequent modeling and analysis, enhancing the stability and prediction accuracy of the model, and reducing the risk of outliers misleading the results.

[0071] In the examples disclosed herein, the processed data can also be saved to a computing resource usage dataset, a cooling system operation dataset, and an environmental monitoring dataset.

[0072] In the example disclosed herein, the storage number expression for the dataset used by the aforementioned computing resources is: [S1, S2, S3, ..., S...]. n S1 represents the first piece of computing resource usage data in the computing resource usage dataset. n This represents the nth data point in the dataset representing the usage of computing resources.

[0073] The storage number expression for the above refrigeration system operation dataset is: [Z1, Z2, Z3, ..., Z... n Z1 represents the first refrigeration system operation data in the refrigeration system operation dataset. n This represents the nth refrigeration system operation data in the refrigeration system operation dataset;

[0074] The storage number expression for the above environmental monitoring dataset is: [H1, H2, H3, ..., H... n H1 represents the first environmental monitoring data in the environmental monitoring dataset. n This represents the nth environmental monitoring data point in the environmental monitoring dataset.

[0075] In the examples disclosed herein, the computing power load characteristic index model can be:

[0076]

[0077] Where SLzs represents the computing power load characteristic index, S i S represents the i-th type of data in the computing resource usage data. imaxrepresents the maximum value in the i-th type of data in the computing power resource usage data, a i represents the weight of the i-th type of data in the computing power resource usage data.

[0078] The refrigeration system operation index model can be:

[0079]

[0080] wherein, SLzs represents the refrigeration system operation index, ys represents the total value of the standard operation data value exceeded in the compressor operation data, ln represents the total value of the standard operation data value exceeded in the condenser operation data, zf represents the total value of the standard operation data value exceeded in the evaporator operation data, and fj represents the total value of the standard operation data value exceeded in the fan operation data, Z o represents the total value of the allowed excess value.

[0081] The environmental impact index model can be:

[0082]

[0083] wherein, HJzs represents the environmental impact index, H w represents the ambient temperature, H wo represents the standard value of the ambient temperature, H s represents the ambient humidity, H so represents the standard value of the ambient humidity, H h represents the dust concentration, H ho represents the standard value of the dust concentration.

[0084] According to the above content, in examples of the present disclosure, the calculation formula of the overall energy consumption impact value can be:

[0085] NHyx = (δ1*SLzs) + (δ2*SLzs) + (δ3*HJzs);

[0086] wherein, NHyx represents the overall energy consumption impact value, SLzs represents the computing power load characteristic index, ZLzs represents the refrigeration system operation index, HJzs represents the environmental impact index, and δ1, δ2 and δ3 represent the weights corresponding to the computing power load characteristic index, the refrigeration system operation index and the environmental impact index.

[0087] Based on the overall energy consumption impact value, abnormal early warning and distributed management can be implemented, which can include:

[0088] When the overall energy consumption impact value NHyx is greater than the overall energy consumption impact value threshold, it represents that the current server thermal power consumption is abnormal, and abnormal early warning is implemented;

[0089] When the current server thermal power consumption is abnormal, the computing power load characteristic index and the refrigeration system operation index are compared with the computing power load characteristic index threshold and the refrigeration system operation index threshold respectively, the corresponding index greater than the corresponding threshold is extracted, and a warning signal is sent.

[0090] By fusing the computing power load characteristic index, the refrigeration system operation index and the environmental impact index, real-time and accurate detection and dynamic distributed management of server thermal power consumption are realized. When the overall energy consumption impact value exceeds the threshold, the system can quickly identify abnormal thermal power consumption, accurately locate the source of the abnormality by comparing each index with the corresponding threshold, and trigger a targeted warning. This mechanism not only improves the fault response speed and avoids local overheating causing equipment damage, but also optimizes resource scheduling through distributed management to balance computing power and refrigeration capacity, ultimately achieving energy efficiency optimization, fault risk reduction, operation and maintenance cost saving, and ensuring safe and stable operation of the data center.

[0091] It can be understood that the method can be executed by any device, equipment, platform, or device cluster with computing and processing capabilities.

[0092] It should be noted that the method of one or more embodiments of the present disclosure can be executed by a single device, such as a computer or a server. The method of the present embodiment can also be applied in a distributed scenario, and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only execute one or more steps in the method of one or more embodiments of the present disclosure, and the multiple devices will interact with each other to complete the method.

[0093] It should be noted that the above describes specific embodiments of the present disclosure. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be executed in an order different than the order in the embodiments and still achieve the desired result. In addition, the processes depicted in the figures do not necessarily require the particular order shown, or sequential order, to achieve the desired results. In certain implementations, multitasking and parallel processing can be advantageous or necessary.

[0094] Based on the same inventive concept, the present disclosure also provides a server thermal power consumption real-time detection system corresponding to the method of any of the above embodiments. As shown in Figure 2 The server thermal power consumption real-time detection system includes:

[0095] The data collection module 11 is configured to collect data center operation data, including computing power resource usage data, refrigeration system operation data and environmental impact data;

[0096] The data processing module 12 is configured to pre-process the data to eliminate outliers, and save the pre-processed data to a computing resource usage dataset, a refrigeration system operation dataset and an environment monitoring dataset;

[0097] The first data calculation module 13 is configured to calculate a computing resource load characteristic index, a refrigeration system operation index and an environment monitoring index according to the data in the computing resource usage dataset, the refrigeration system operation dataset and the environment monitoring dataset, respectively.

[0098] The second data calculation module 14 is configured to calculate an overall energy consumption impact value according to the computing resource load characteristic index, the refrigeration system operation index and the environment monitoring index.

[0099] The data analysis and early warning module 15 is configured to implement abnormal early warning and distributed management according to the overall energy consumption impact value.

[0100] For the convenience of description, the above device is described as various modules respectively according to functions. Of course, the functions of each module can be implemented in one or more software and / or hardware when implementing one or more embodiments of the present disclosure.

[0101] The device of the above embodiment is used to implement the corresponding method in the above embodiment, and has the beneficial effects of the corresponding method embodiment, which will not be described here.

[0102] Those skilled in the art should understand that the discussion of any of the above embodiments is only exemplary and is not intended to imply that the scope of the present disclosure (including the claims) is limited to these examples; under the idea of the present disclosure, the technical features in the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other changes of different aspects of the one or more embodiments of the present disclosure as described above, which are not provided in details for the sake of brevity.

[0103] Additionally, to simplify the description and discussion, and so as not to obscure one or more embodiments of the disclosure with details that are well known to those skilled in the art, the well-known power / ground connections of the integrated circuits (ICs) and other components can or can not be shown in the provided drawings. Furthermore, devices can be shown in block diagram form in order to avoid obscuring one or more embodiments of the disclosure, and this also acknowledges the fact that the details in regard to how the devices are implemented can vary depending on the platform in which one or more embodiments of the disclosure are being implemented, i.e., these details should be completely understood in view of the teachings provided herein. While the description sets forth various embodiments of the present disclosure, changes can be made, and equivalents can be substituted, without departing from the spirit and scope of one or more embodiments of the disclosure. For example, coherency can be implemented in other memory architectures (e.g., dynamic RAM (DRAM)) using the embodiments discussed.

[0104] Although the present disclosure has been described in connection with certain specific embodiments thereof, many modifications, changes, substitutions and variations will now occur to those skilled in the art. For example, other memory architectures (e.g., dynamic RAM (DRAM)) can use the embodiments discussed.

[0105] It should be understood that any of the embodiments of the present disclosure can include any of the features, alone or in combination, with the rest of the features. It should be understood that every maximum numerical limitation given throughout this disclosure will have a lower limit, whether or not that lower limit is already specifically disclosed. The upper and lower limits of these concentrations and dosage amounts can be expressly combined with one another. The disclosure of the foregoing detailed description together with the accompanying drawings makes clear that no single feature or combination of features is solely responsible and required for the practice of one or more embodiments of the disclosure. Thus, the disclosure is not to be understood as being confined to the illustrative embodiments of one or more embodiments of the disclosure. Rather, the scope of the disclosure is to be understood as eluding all such embodiments that can reasonably be called equivalents irrespective of whether such equivalents could fall within the literal scope of the claims.

Claims

1. A method for real-time detection of server thermal power dissipation, characterized in that, The method comprises the following steps: collecting data of data center operation, the data comprising computing resource usage data, refrigeration system operation data and environmental impact data; preprocessing the data to eliminate outliers, and saving the preprocessed data to computing resource usage dataset, refrigeration system operation dataset and environmental monitoring dataset; calculating computing load characteristic index, refrigeration system operation index and environmental monitoring index according to the data in the computing resource usage dataset, the refrigeration system operation dataset and the environmental monitoring dataset, respectively; calculating overall energy consumption impact value according to the computing load characteristic index, the refrigeration system operation index and the environmental impact index; implementing abnormal warning and distributed management according to the overall energy consumption impact value.

2. The method of claim 1, wherein, The preprocessing of the data to eliminate outliers comprises: setting unique standard keywords for the computing resource usage data, the refrigeration system operation data and the environmental impact data, and eliminating data in the computing resource usage data, the refrigeration system operation data and the environmental impact data that does not conform to the unique standard keywords; eliminating missing values in the computing resource usage data, the refrigeration system operation data and the environmental impact data by using general data processing tools; setting the highest standard value and the lowest standard value for the computing resource usage data, the refrigeration system operation data and the environmental impact data, and eliminating data higher than the highest standard value and data lower than the lowest standard value; replacing the eliminated numerical vacancies with historical average values.

3. The method of claim 2, wherein The saving number expression of the computing resource usage data set is: [S1, S2, S3, …, Sn] n ], S1 represents the first computing resource usage data in the computing resource usage data set, S n n represents the nth computing resource usage data in the computing resource usage data set; The saving number expression of the refrigeration system operation data set is: [Z1, Z2, Z3, ···, Zn] n ], Z1 represents the first refrigeration system operation data in the refrigeration system operation data set, Z n represents the n-th refrigeration system operation data in the refrigeration system operation data set; The saving number expression of the environment monitoring data set is: [H1, H2, H3, ···, H n ], H1 represents the first environment monitoring data in the environment monitoring data set, H n represents the nth environment monitoring data in the environment monitoring data set.

4. The method of claim 3, wherein the computing resource usage data comprises CPU utilization, GPU utilization, memory occupancy, network bandwidth occupancy, read-write throughput, storage device response time, number of currently running processes and number of tasks to be executed; the refrigeration system operation data comprises compressor operation data, condenser operation data, evaporator operation data and fan operation data, and the operation data comprises temperature, humidity, pressure, flow, vibration and noise; the environmental monitoring data comprises environmental temperature, environmental humidity and dust concentration.

5. The method of claim 4, wherein, The computing of the computing load characteristic index according to the data in the computing resource usage dataset comprises: obtaining the computing load characteristic index according to the data in the computing resource usage dataset by using the following computing load characteristic index model; wherein SLzs represents the computing power load characteristic index, S i represents the i-th type of data in the computing power resource usage data, S imax represents the maximum value in the i-th type of data in the computing power resource usage data, a i represents the weight of the i-th type of data in the computing power resource usage data.

6. The method of claim 5, wherein, The computing of the refrigeration system operation index according to the data in the refrigeration system operation dataset comprises: obtaining the refrigeration system operation index according to the data in the refrigeration system operation dataset by using the following refrigeration system operation index model; wherein ZLzs represents a refrigeration system operation index, ys represents a total value of the standard operation data values exceeded in the compressor operation data, ln represents a total value of the standard operation data values exceeded in the condenser operation data, zf represents a total value of the standard operation data values exceeded in the evaporator operation data, fj represents a total value of the standard operation data values exceeded in the fan operation data, and Z o represents a total value of the allowed excess values.

7. The method of claim 6, wherein, The computing of the environmental impact index according to the data in the environmental impact dataset comprises: obtaining the environmental impact index according to the data in the environmental impact dataset by using the following environmental impact index model; where HJzs represents an environmental impact index, H w represents an ambient temperature, H wo represents a standard value of an ambient temperature, H s represents an ambient humidity, H so represents a standard value of an ambient humidity, H h represents a dust concentration, H ho represents a standard value of a dust concentration.

8. The method of claim 7, wherein, The computing of the overall energy consumption impact value according to the computing load characteristic index, the refrigeration system operation index and the environmental impact index comprises: NHyx=(δ1*SLzs)+(δ2*ZLzs)+(δ3*HJzs); wherein NHyx represents an overall energy consumption impact value, SLzs represents a computing power load characteristic index, ZLzs represents a refrigeration system operation index, HJzs represents an environmental impact index, and δ1, δ2, and δ3 represent weights corresponding to the computing power load characteristic index, the refrigeration system operation index, and the environmental impact index.

9. The method of claim 8, wherein, The implementation of abnormal early warning and distributed management according to the overall energy consumption impact value comprises: When the overall energy consumption impact value is greater than an overall energy consumption impact value threshold, an abnormal early warning is implemented. When the overall energy consumption impact value is abnormal, the computing power load characteristic index and the refrigeration system operation index are compared with a computing power load characteristic index threshold and a refrigeration system operation index threshold, respectively, and corresponding indexes greater than the corresponding thresholds are extracted, and a warning signal is issued.

10. A server thermal power real-time detection system, characterized in that, It comprises: a data collection module configured to collect data center operation data, the data comprising computing power resource usage data, refrigeration system operation data, and environmental impact data; a data processing module configured to pre-process the data to eliminate outliers, and save the pre-processed data to a computing power resource usage dataset, a refrigeration system operation dataset, and an environmental monitoring dataset; a first data calculation module configured to calculate a computing power load characteristic index, a refrigeration system operation index, and an environmental monitoring index according to data in the computing power resource usage dataset, the refrigeration system operation dataset, and the environmental monitoring dataset, respectively; a second data calculation module configured to calculate an overall energy consumption impact value according to the computing power load characteristic index, the refrigeration system operation index, and the environmental impact index; a data analysis and early warning module configured to implement abnormal early warning and distributed management according to the overall energy consumption impact value.

Images