An adaptive energy-saving control system and method for data centers

Abstract

This invention discloses an adaptive energy-saving control system and method for data centers. The invention relates to the field of data center technology. It rapidly determines whether data center energy consumption meets standards through periodic monitoring: when the operating parameters of all secondary and primary parameters are within the same ranking correlation range, it indicates good system coordination and reasonable energy consumption; otherwise, it is immediately identified as abnormal energy consumption. This assessment method does not require complex energy consumption model calculations; it can be completed simply by matching interval rankings, significantly improving the efficiency of identifying abnormal energy consumption, allowing time for timely intervention and adjustment, and avoiding the continuous consumption of ineffective energy. When energy consumption meets standards, the data throughput is adjusted by preset fluctuation values, selecting the optimal data throughput with the goal of "minimizing moment-specific characteristics," achieving further energy-saving optimization while ensuring stable system operation.

Description

Technical Field

[0001] This invention relates to the field of data center technology, specifically to an adaptive energy-saving control system and method for data centers. Background Technology

[0002] With the rapid development of the digital economy, data centers, as the core of computing infrastructure, are experiencing continuous expansion in scale and computing power demand, leading to increasingly prominent energy consumption issues. According to industry statistics, the proportion of data center electricity consumption in global total electricity consumption is rising year by year, with cooling systems, server clusters, and supporting equipment accounting for the majority of the energy consumption. How to achieve energy conservation and consumption reduction while ensuring the stability of computing power supply and service quality has become a core requirement for data center operation and management.

[0003] Currently, energy-saving control methods for data centers mostly focus on optimizing single devices or using fixed strategies. However, these methods have significant limitations: Data center operation is a complex system involving the coordinated action of multiple devices and parameters. There are strong correlations between operational items such as server data throughput, cooling power, and power supply load. Local optimization of a single device is difficult to take into account the overall system's energy consumption coordination, which can easily lead to the problem of "local energy saving but overall energy efficiency decline." Fixed strategy control relies on manually preset parameters, which cannot adapt to dynamic scenarios such as fluctuations in data throughput and changes in ambient temperature. When the data volume suddenly increases or the external environment changes abruptly, the preset strategy may lead to mismatch in device operating parameters. Either the parameters are too high, resulting in energy waste, or the parameters are insufficient, affecting service stability. It is difficult to achieve the energy-saving goal of "dynamic adaptation and global optimization."

[0004] In addition, although some existing energy-saving control systems have attempted to introduce correlation parameter analysis, they are still crude in terms of quantifying the correlation between operating items and dividing the intervals: most methods use fixed intervals to divide parameter intervals without considering the actual characteristics of parameter fluctuations in historical operating data, resulting in the correlation intervals being disconnected from the actual operating patterns, which in turn affects the accuracy of energy consumption status assessment and the effectiveness of control strategies.

[0005] Therefore, there is an urgent need for an energy-saving control method that can accurately capture the correlation patterns of multiple operating items, dynamically assess energy consumption status, and adaptively adjust parameters, so as to overcome the shortcomings of existing technologies in terms of coordination, dynamism, and accuracy, and achieve global optimization and stable improvement of data center energy utilization efficiency. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides an adaptive energy-saving control system and method for data centers. It solves the problem that preset strategies may lead to mismatch in equipment operating parameters, resulting in either excessively high parameters causing energy waste or insufficient parameters affecting service stability, making it difficult to achieve the energy-saving goal of "dynamic adaptation and global optimization".

[0007] To achieve the above objectives, the present invention provides the following technical solution: a data center adaptive energy-saving control method, comprising the following steps:

[0008] Step 1: Extract historical operational data associated with the data center. Identify the operational parameters associated with different operational items from this data. Based on the correlation between the operational parameters of different operational items, identify and record the correlation intervals between them. The specific method is as follows:

[0009] Select the primary running item from the confirmed running items and use it as the primary item. Use the other running items as secondary items. Confirm the historical running parameters associated with the primary item and select the minimum and maximum values ​​from the confirmed historical running parameters as the parameter running interval. Divide the parameter running interval associated with the primary item into N micro intervals according to the preset segmentation value N.

[0010] In the same way as the main item confirms the parameter operating range, the parameter operating range associated with other sub-items is confirmed, and the confirmed parameter operating range is simultaneously divided into N micro-intervals, denoted as sub-micro-intervals, and the micro-interval associated with the main item is denoted as the main micro-interval.

[0011] Perform several sub-micro interval partitioning processes, confirm the correlation value between the sub-micro interval and the main micro interval in the corresponding partitioning process, and average the correlation values ​​confirmed in this partitioning process to obtain the process characteristics belonging to this partitioning process.

[0012] The process characteristics associated with different partitioning processes are confirmed in turn. From the confirmed process characteristics, the maximum value is selected, and the partitioning process associated with the maximum value is recorded as the selected process. The sub-micro interval associated with the corresponding sub-item of the selected process is recorded as the associated interval of the corresponding sub-item.

[0013] Step 2: Based on the confirmed correlation intervals between different operating items, monitor the operating parameters associated with the operating items, and confirm whether the energy consumption of multiple operating items meets the standards according to the monitoring and evaluation process. If the standards are met, proceed to Step 3 for optimization; if the standards are not met, proceed to Step 4 for adjustment. The specific method is as follows:

[0014] The operating parameters associated with each operating item are monitored, and the operating parameters associated with different operating items are denoted as C. i , where i represents different running items;

[0015] Confirm the runtime parameter C associated with the main item. iConfirm the associated interval to which its running parameters belong, and record the sorting rank of its associated interval. Then identify the sorting rank of the associated interval to which the running parameters of other sub-items belong. Confirm whether the sorting ranks confirmed between all sub-items and the main item are consistent. If yes, proceed to step three; otherwise, proceed to step four.

[0016] Step 3: Confirm the data throughput associated with the data center, and adaptively adjust the data throughput according to the preset fluctuation value. Perform range verification on the operating parameters associated with different running items in the adjustment process, select the optimal data throughput, and execute it. The specific method is as follows:

[0017] The current data throughput is denoted as TT, and based on a preset fluctuation value Y1, the data throughput is adjusted from (TT-Y1) to (TT+Y1). The time period associated with this adaptive adjustment is denoted as the adjustment period, and the different operating parameters of different running items associated with different times within the adjustment period are denoted as Y. q-o , where q represents different running items and o represents different times;

[0018] For the operating parameter Y associated at the same moment q-o Perform feature verification: identify its operation Y q-o The interval position located within the associated interval, denoted by Y q-o As a dividing value, the corresponding correlation interval is divided into two numerical segments. The segment with the lower value is designated as the low-value segment, and the segment with the higher value is designated as the high-value segment. The numerical range F1 of the low-value segment is determined, and the numerical range F2 of the high-value segment is determined. The formula is: F2 ÷ F1 = Tz q-o Confirmation of the feature Tz segment q-o And simultaneously associated with several value segment features Tz q-o Variance processing is performed to obtain the time features associated with the corresponding time.

[0019] The time characteristics associated with different times are confirmed, the minimum value is selected, the data throughput associated with the minimum value is recorded as the optimal data throughput and executed;

[0020] Step 4: Identify the specific operating items with abnormal operating parameters, and restrict the operating parameters associated with those specific operating items to ensure that all operating items are within the same related interval in the same order, thus completing the energy-saving control process. The specific method is as follows:

[0021] The ranking of the associated interval to which the main item's operating parameters belong is recorded as the primary ranking, and the ranking of the associated interval to which the specific operating item's operating parameters belong is recorded as the secondary ranking. The differences between the secondary ranking and the primary ranking are then identified.

[0022] If the secondary ranking is lower than the primary ranking, adjust the operating parameters of the specific operating item upwards until the secondary ranking matches the primary ranking;

[0023] If the secondary ranking is higher than the primary ranking, the operating parameters of the specific operating item are adjusted down until the secondary ranking is consistent with the primary ranking, at which point the operation stops.

[0024] Preferably, in step one, the specific method for dividing the parameter operating range into micro-intervals is as follows:

[0025] Within the confirmed parameter operating range, (N-1) boundary values ​​are identified. These boundary values ​​change within the parameter operating range. Simultaneously, the minimum and maximum values ​​associated with the parameter operating range are recorded as boundary values. The range of numerical segments between adjacent boundary values ​​is confirmed, and the mean characteristics associated with each range of numerical segments are confirmed.

[0026] The operating parameters within the range of numerical values ​​are sorted in ascending order. The changes between adjacent operating parameters within the range of numerical values ​​are then confirmed. The variance of the confirmed changes is then processed and recorded as the value segment feature of the current range of numerical values. The value segment features of the N confirmed range of numerical values ​​are confirmed, and the mean of the confirmed value segment features is processed to confirm the mean of the current process.

[0027] Then execute other processing processes. The boundary values ​​of each processing process are not the same. The verification mean associated with each processing process is confirmed in turn. From the confirmed verification mean values, the minimum value is selected. The processing process associated with the minimum value is recorded as the standard process. According to the location of the boundary value in the standard process, the parameter running interval is divided into micro intervals and recorded as the associated interval of the corresponding main item.

[0028] Preferably, the specific method for confirming the correlation between the secondary and primary micro-intervals during the partitioning process is as follows:

[0029] The sub-intervals are sorted in ascending order, and the primary intervals are sorted simultaneously. Sub-intervals at the same sorting position are associated with and confirmed to be related to the primary intervals. The running parameters that appear at the same time in the sub-intervals and primary intervals are marked. The proportion of the corresponding running parameters located in the sub-intervals or primary intervals is then confirmed. The confirmed proportions are averaged to obtain the correlation value between the primary intervals and the sub-intervals.

[0030] Preferably, a data center adaptive energy-saving control system includes:

[0031] The associated interval recording end extracts the historical operation data associated with the data center, identifies the operation parameters associated with different operation items from the historical operation data, and identifies and records the associated intervals between different operation items based on the correlation between the corresponding operation parameters of different operation items.

[0032] The parameter assessment center monitors the operating parameters associated with different operating items based on the confirmed correlation intervals between them, and confirms whether the energy consumption of multiple operating items meets the standards based on the assessment process.

[0033] The optimization processing end confirms the data throughput associated with the data center and adaptively adjusts the data throughput according to the preset fluctuation value. It performs range verification processing on the running parameters associated with different running items in the adjustment process, selects the optimal data throughput and executes it.

[0034] Adjust the processing end to identify the specific operating items with abnormal operating parameters, and restrict the operating parameters associated with the specific operating items so that multiple operating items are all within the same sorting position and the energy-saving control process is completed.

[0035] This invention provides an adaptive energy-saving control system and method for data centers. Compared with existing technologies, it has the following advantages:

[0036] This invention selects the standard process with the "most stable value segment characteristics" through variance analysis and mean verification to ensure that the main item correlation interval can truly reflect the fluctuation pattern of its operating parameters; and determines the secondary item correlation interval with the goal of "maximum correlation between main and secondary items", so that the interval division of each operating item forms a strongly coupled correlation system, accurately capturing the collaborative operation pattern between different operating items, and providing a quantitative basis that fits the actual operating scenario for subsequent energy consumption monitoring and control.

[0037] By periodically monitoring, we can quickly determine whether the energy consumption of the data center meets the standards: when the operating parameters of all sub-items and the main item are in the same sorted correlation range, it means that the system operation is coordinated and the energy consumption is reasonable; otherwise, it is immediately identified as an energy consumption anomaly. This evaluation method does not require complex energy consumption model calculations. It can be judged simply by matching the interval rankings, which greatly improves the efficiency of identifying abnormal energy consumption, buys time for timely intervention and adjustment, and avoids the continuous consumption of ineffective energy.

[0038] When energy consumption meets the standards, data throughput is adjusted by preset fluctuation values. The optimal data throughput is selected with the goal of "minimizing moment-specific characteristics" (i.e., the most balanced distribution of parameters among various operating items within the associated interval), achieving further energy-saving optimization while ensuring stable system operation. When energy consumption does not meet the standards, parameters are precisely adjusted upwards or downwards based on the difference in ranking between the secondary and primary items within their associated intervals. This ensures that the secondary and primary items return to the same ranking interval, avoiding both the problem of "energy waste due to excessively high parameters" and the hidden danger of "inefficient system due to excessively low parameters." This adjustment method is entirely based on the matching relationship between the actual operating parameters of the system and the associated intervals, requiring no manual intervention. It achieves automated and intelligent energy-saving control while maximizing the operational stability of the data center. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0040] Figure 2 This is a schematic diagram of the principle framework of the present invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] First Embodiment

[0043] Please see Figure 1 This application provides an adaptive energy-saving control method for data centers, comprising the following steps:

[0044] Step 1: Extract the historical operational data associated with the data center, identify the operational parameters associated with different operational items from the historical operational data, and identify and record the correlation intervals between different operational items based on the correlation between the corresponding operational parameters of different operational items. Specifically, within the corresponding data center, each different operational item is related to the others. Generally, there is a master operational item, such as data throughput, and other operational items follow the changes in the operational parameters of the master operational item (data throughput).

[0045] The specific method for confirming the corresponding correlation intervals between different operation items is as follows:

[0046] Select the primary running item from the confirmed running items, and treat the other running items as secondary items. Confirm the historical running parameters associated with the primary running item, and select the minimum and maximum values ​​from the confirmed historical running parameters as the parameter running interval. Based on the preset segmentation value N, which is a preset value determined in advance by relevant operators, if the segmentation value is 5, then the corresponding parameter running interval is divided into five micro-intervals. Divide the parameter running interval associated with the primary running item into N micro-intervals. Within the confirmed parameter running interval, confirm (N-1) boundary values, which change within the parameter running interval (or can be understood as a numerical segment, with the boundary values ​​moving within the numerical segment). Simultaneously, record the minimum and maximum values ​​associated with the parameter running interval as boundary values. Confirm the range of numerical segments included between adjacent boundary values, and confirm the mean characteristics associated with each range of numerical segments.

[0047] The operating parameters within the range of numerical values ​​are sorted in ascending order. The changes between adjacent operating parameters within the range of numerical values ​​are then confirmed. The variance of the confirmed changes is then processed and recorded as the value segment feature of the current range of numerical values. The value segment features of the N confirmed range of numerical values ​​are confirmed, and the mean of the confirmed value segment features is processed to confirm the mean of the current process.

[0048] Then execute other processing processes. The boundary values ​​of each processing process are not the same. The verification mean associated with each processing process is confirmed in turn. From the confirmed verification mean values, the minimum value is selected. The processing process associated with the minimum value is recorded as the standard process. According to the location of the boundary value in the standard process, the parameter running interval is divided into micro intervals and recorded as the associated interval of the corresponding main item.

[0049] In the same way as the main item confirms the parameter operating range, the parameter operating range associated with other sub-items is confirmed, and the confirmed parameter operating range is simultaneously divided into N micro-intervals, denoted as sub-micro-intervals, and the micro-interval associated with the main item is denoted as the main micro-interval.

[0050] Perform several sub-micro interval division processes to confirm the correlation between the sub-micro intervals and the main micro intervals in the corresponding division processes: sort the sub-micro intervals in ascending order, and simultaneously sort the main micro intervals. Confirm the correlation between the sub-micro intervals and the main micro intervals at the same sorting position. Mark the running parameters that appear at the same time in the sub-micro intervals and the main micro intervals (confirmed from historical data). Then confirm the proportion of the corresponding running parameters located in the sub-micro intervals or the main micro intervals. Average the confirmed proportions to obtain the correlation values ​​between the main micro intervals and the sub-micro intervals. Average the several correlation values ​​confirmed in this division process to obtain the process characteristics belonging to this division process.

[0051] Then, the process characteristics associated with different partitioning processes are confirmed in turn. From the confirmed process characteristics, the maximum value is selected, and the partitioning process associated with the maximum value is recorded as the selected process. The sub-micro interval associated with the corresponding sub-item of the selected process is recorded as the associated interval of the corresponding sub-item.

[0052] Similarly, different sub-items have different correlation intervals. The correlation intervals of these sub-items are most closely related to the correlation intervals of the main item. After confirming the corresponding correlation intervals, the running process of subsequent running items can be effectively monitored, and it can be identified whether the running parameters associated with their running process exceed the standard or are abnormal. In this way, a comprehensive evaluation can be made to determine whether the running item needs to be processed to reduce energy consumption in order to achieve specific energy-saving control effects.

[0053] Step 2: Based on the confirmed correlation intervals between different operating items, monitor the operating parameters associated with the operating items, and based on the evaluation process of the monitoring, confirm whether the energy consumption of multiple operating items meets the standards. If it meets the standards, proceed to Step 3 for optimization. If it does not meet the standards, proceed to Step 4 for adjustment.

[0054] The specific method for assessing whether energy consumption meets standards across multiple operational items is as follows:

[0055] The operating parameters associated with each operating item are monitored, and the operating parameters associated with different operating items are denoted as C. i , where i represents different running items;

[0056] Confirm the runtime parameter C associated with the main item. i First, confirm the associated interval to which the operating parameters belong and record the ranking of the associated interval. Then, identify the ranking of the associated interval to which the operating parameters of other sub-items belong. Check whether the rankings of all sub-items and the main item are consistent. If so, it means that the energy consumption of the multiple operating items monitored at the current time meets the standard, and step three is executed for specific optimization. If not, it means that the energy consumption of the multiple operating items monitored at the current time does not meet the standard, and step four is executed for specific adjustment.

[0057] Specifically, in the actual monitoring and processing process, it is generally periodic monitoring, for example, once an hour. In each monitoring process, it is necessary to confirm whether the operating parameters associated with the corresponding operating item meet the associated interval of the confirmed corresponding characteristics, so as to conduct a comprehensive comparison and verification. If they all belong to the associated interval at the same sorting position, it means that the energy consumption between the corresponding operating items is normal. Otherwise, it means that the energy consumption between the corresponding operating items is abnormal, and relevant adjustments and processing are required.

[0058] Step 3: Confirm the data throughput associated with the data center, and adaptively adjust the data throughput according to the preset fluctuation value. Perform range verification on the operating parameters associated with different running items in the adjustment process, select the optimal data throughput and execute it.

[0059] The specific method for selecting the optimal data throughput is as follows:

[0060] The current data throughput is denoted as TT. Based on a preset fluctuation value Y1, where Y1 is a preset value determined by the operator based on experience, the data throughput is adjusted from (TT-Y1) to (TT+Y1). The time period associated with this adaptive adjustment is denoted as the adjustment period. The different operating parameters of different operating items associated with different times within the adjustment period are denoted as Y. q-o , where q represents different running items and o represents different times;

[0061] For the operating parameter Y associated at the same moment q-o Perform feature verification: identify its operation Y q-o The interval position located within the associated interval, denoted by Y q-o As a dividing value, the corresponding correlation interval is divided into two numerical segments. The segment with the lower value is designated as the low-value segment, and the segment with the higher value is designated as the high-value segment. The numerical range F1 of the low-value segment is determined, and the numerical range F2 of the high-value segment is determined. The formula is: F2 ÷ F1 = Tz q-o Confirmation of the feature Tz segment q-o And simultaneously associated with several value segment features Tz q-o Variance processing is performed to obtain the time features associated with the corresponding time.

[0062] The time characteristics associated with different times are confirmed, the minimum value is selected, the data throughput associated with the minimum value is recorded as the optimal data throughput and executed;

[0063] Specifically, during the execution of optimal data throughput, the operating parameters associated with different operating items are all in a relatively balanced position within their respective associated intervals. In other words, no operating item has excessively low or high parameters, thus achieving a relatively balanced state. This effectively achieves the parameter optimization process of the corresponding data center, reaching dynamic balance, and ensuring that all parameters associated with the data center reach a relatively balanced optimization state.

[0064] Step 4: Identify the specific operating items with abnormal operating parameters, and restrict the operating parameters associated with the specific operating items to ensure that multiple operating items are all within the same sorting position and associated interval, thus completing the energy-saving control process.

[0065] The specific methods for restricting specific running items are as follows:

[0066] The ranking of the associated interval to which the main item's operating parameters belong is recorded as the primary ranking, and the ranking of the associated interval to which the specific operating item's operating parameters belong is recorded as the secondary ranking. The differences between the secondary ranking and the primary ranking are then identified.

[0067] If the secondary ranking is lower than the primary ranking, adjust the operating parameters of the specific operating item upwards until the secondary ranking matches the primary ranking;

[0068] If the secondary ranking is higher than the primary ranking, then the operating parameters of the specific operating item are lowered until the secondary ranking is consistent with the primary ranking.

[0069] Specifically, when there are significant differences in the operating characteristics between corresponding operating items, it is necessary to adjust the operating parameters associated with the operating items upwards or downwards. When downward adjustment is required, it indicates that the energy consumption of the corresponding operating item is too high, so it needs to be adjusted downwards. This not only ensures the stable operation of the data center but also significantly reduces energy consumption. When the energy consumption of a certain operating item parameter is too low, it needs to be adjusted upwards to ensure that the data center does not experience low load and effectively improve the overall operating efficiency of the data center.

[0070] Second Embodiment

[0071] Combination Figure 2 An adaptive energy-saving control system for data centers, comprising:

[0072] The associated interval recording end extracts the historical operation data associated with the data center, identifies the operation parameters associated with different operation items from the historical operation data, and identifies and records the associated intervals between different operation items based on the correlation between the corresponding operation parameters of different operation items.

[0073] The parameter assessment center monitors the operating parameters associated with different operating items based on the confirmed correlation intervals between them, and confirms whether the energy consumption of multiple operating items meets the standards based on the assessment process.

[0074] The optimization processing end confirms the data throughput associated with the data center and adaptively adjusts the data throughput according to the preset fluctuation value. It performs range verification processing on the running parameters associated with different running items in the adjustment process, selects the optimal data throughput and executes it.

[0075] Adjust the processing end to identify the specific operating items with abnormal operating parameters, and restrict the operating parameters associated with the specific operating items so that multiple operating items are all within the same sorting position and the energy-saving control process is completed.

[0076] Some of the data in the above formulas are numerical calculations with dimensions removed, and the contents not described in detail in this specification are all prior art known to those skilled in the art.

[0077] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A data center adaptive energy-saving control method, characterized in that, Includes the following steps: Step 1: Extract historical operational data associated with the data center. Identify the operational parameters associated with different operational items from this data. Based on the correlation between the operational parameters of different operational items, identify and record the correlation intervals between them. The specific method is as follows: Select the primary running item from the confirmed running items and use it as the primary item. Use the other running items as secondary items. Confirm the historical running parameters associated with the primary item and select the minimum and maximum values ​​from the confirmed historical running parameters as the parameter running interval. Divide the parameter running interval associated with the primary item into N micro intervals according to the preset segmentation value N. In the same way as the main item confirms the parameter operating range, the parameter operating range associated with other sub-items is confirmed, and the confirmed parameter operating range is simultaneously divided into N micro-intervals, denoted as sub-micro-intervals, and the micro-interval associated with the main item is denoted as the main micro-interval. Perform several sub-micro interval partitioning processes, confirm the correlation value between the sub-micro interval and the main micro interval in the corresponding partitioning process, and average the correlation values ​​confirmed in this partitioning process to obtain the process characteristics belonging to this partitioning process. The process characteristics associated with different partitioning processes are confirmed in turn. From the confirmed process characteristics, the maximum value is selected, and the partitioning process associated with the maximum value is recorded as the selected process. The sub-micro interval associated with the corresponding sub-item of the selected process is recorded as the associated interval of the corresponding sub-item. The specific method for dividing the parameter operating range into micro-intervals is as follows: Within the confirmed parameter operating range, identify (N-1) boundary values. These boundary values ​​change within the parameter operating range. Simultaneously, record the minimum and maximum values ​​associated with the parameter operating range as boundary values. Confirm the range of numerical segments between adjacent boundary values ​​and confirm the mean characteristics associated with each range of numerical segments. The operating parameters within the range of numerical values ​​are sorted in ascending order. The changes between adjacent operating parameters within the range of numerical values ​​are then confirmed. The variance of the confirmed changes is then processed and recorded as the value segment feature of the current range of numerical values. The value segment features of the N confirmed range of numerical values ​​are confirmed, and the mean of the confirmed value segment features is processed to confirm the mean of the current process. Then execute other processing processes. The boundary values ​​of each processing process are not the same. The verification mean associated with each processing process is confirmed in turn. From the confirmed verification mean values, the minimum value is selected. The processing process associated with the minimum value is recorded as the standard process. According to the location of the boundary value in the standard process, the parameter running interval is divided into micro intervals and recorded as the associated interval of the corresponding main item. Step 2: Based on the confirmed correlation intervals between different operating items, monitor the operating parameters associated with the operating items, and based on the evaluation process of the monitoring, confirm whether the energy consumption of multiple operating items meets the standards. If it meets the standards, proceed to Step 3 for optimization. If it does not meet the standards, proceed to Step 4 for adjustment. Step 3: Confirm the data throughput associated with the data center, and adaptively adjust the data throughput according to the preset fluctuation value. Perform range verification on the operating parameters associated with different running items in the adjustment process, select the optimal data throughput and execute it. Step 4: Identify the specific operating items with abnormal operating parameters, and restrict the operating parameters associated with the specific operating items to ensure that multiple operating items are all within the same sorting position and related interval, thus completing the energy-saving control process.

2. The adaptive energy-saving control method for data centers according to claim 1, characterized in that, In step one, the specific method for confirming the correlation between the secondary micro-intervals and the primary micro-intervals during the partitioning process is as follows: The sub-intervals are sorted in ascending order, and the primary intervals are sorted simultaneously. Sub-intervals at the same sorting position are associated with and confirmed to be related to the primary intervals. The running parameters that appear at the same time in the sub-intervals and primary intervals are marked. The proportion of the corresponding running parameters located in the sub-intervals or primary intervals is then confirmed. The confirmed proportions are averaged to obtain the correlation value between the primary intervals and the sub-intervals.

3. The adaptive energy-saving control method for data centers according to claim 1, characterized in that, In step two, the method for assessing the energy consumption compliance among multiple operating items is as follows: The operating parameters associated with each operating item are monitored, and the operating parameters associated with different operating items are denoted as C. i , where i represents different running items; Confirm the running parameters associated with the main item, and confirm the associated interval to which the running parameters belong, and record the ranking of the associated interval. Then identify the ranking of the associated interval to which the running parameters of other sub-items belong, and confirm whether the rankings of all sub-items and the main item are consistent. If yes, proceed to step three; otherwise, proceed to step four.

4. The adaptive energy-saving control method for data centers according to claim 1, characterized in that, In step three, the specific method for selecting the optimal data throughput is as follows: The current data throughput is denoted as TT. Based on a preset fluctuation value Y1, the data throughput is adjusted from (TT-Y1) to (TT+Y1). The time period associated with this adaptive adjustment is denoted as the adjustment period. The different operating parameters of different running items associated with different times within the adjustment period are denoted as Y. q-o , where q represents different running items and o represents different times; For the operating parameter Y associated at the same moment q-o Perform feature verification: identify its operation Y q-o The interval position located within the associated interval, denoted by Y q-o As a dividing value, the corresponding correlation interval is divided into two numerical segments. The segment with the lower value is designated as the low-value segment, and the segment with the higher value is designated as the high-value segment. The numerical range F1 of the low-value segment is determined, and the numerical range F2 of the high-value segment is determined. The formula is: F2 ÷ F1 = Tz q-o Confirmation of the feature Tz segment q-o And simultaneously associated with several value segment features Tz q-o Variance processing is performed to obtain the time features associated with the corresponding time. The time characteristics associated with different times are confirmed, the minimum value is selected, the data throughput associated with the minimum value is recorded as the optimal data throughput and executed.

5. The adaptive energy-saving control method for data centers according to claim 1, characterized in that, In step four, the specific method for restricting specific running items is as follows: The ranking of the associated interval to which the main item's operating parameters belong is recorded as the primary ranking, and the ranking of the associated interval to which the specific operating item's operating parameters belong is recorded as the secondary ranking. The differences between the secondary ranking and the primary ranking are then identified. If the secondary ranking is lower than the primary ranking, adjust the operating parameters of the specific operating item upwards until the secondary ranking matches the primary ranking; If the secondary ranking is higher than the primary ranking, the operating parameters of the specific operating item are adjusted down until the secondary ranking is consistent with the primary ranking, at which point the operation stops.

6. A data center adaptive energy-saving control system, wherein the system operates according to any one of claims 1-5 of the data center adaptive energy-saving control method, characterized in that, include: The associated interval recording end extracts the historical operation data associated with the data center, identifies the operation parameters associated with different operation items from the historical operation data, and identifies and records the associated intervals between different operation items based on the correlation between the corresponding operation parameters of different operation items. The parameter assessment center monitors the operating parameters associated with different operating items based on the confirmed correlation intervals between them, and confirms whether the energy consumption of multiple operating items meets the standards based on the assessment process. The optimization processing end confirms the data throughput associated with the data center and adaptively adjusts the data throughput according to the preset fluctuation value. It performs range verification processing on the running parameters associated with different running items in the adjustment process, selects the optimal data throughput and executes it. Adjust the processing end to identify the specific operating items with abnormal operating parameters, and restrict the operating parameters associated with the specific operating items so that multiple operating items are all within the same sorting position and the energy-saving control process is completed.

Images