CN117850494A - Data center energy-saving control method based on AI algorithm - Google Patents
Data center energy-saving control method based on AI algorithm Download PDFInfo
- Publication number
- CN117850494A CN117850494A CN202311720120.4A CN202311720120A CN117850494A CN 117850494 A CN117850494 A CN 117850494A CN 202311720120 A CN202311720120 A CN 202311720120A CN 117850494 A CN117850494 A CN 117850494A
- Authority
- CN
- China
- Prior art keywords
- data
- algorithm
- data center
- model
- energy
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 25
- 238000012544 monitoring process Methods 0.000 claims abstract description 32
- 238000005265 energy consumption Methods 0.000 claims abstract description 30
- 230000002787 reinforcement Effects 0.000 claims abstract description 26
- 238000007781 pre-processing Methods 0.000 claims abstract description 18
- 238000001816 cooling Methods 0.000 claims description 15
- 230000009471 action Effects 0.000 claims description 8
- 238000009499 grossing Methods 0.000 claims description 8
- 238000010801 machine learning Methods 0.000 claims description 8
- 238000005457 optimization Methods 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 6
- 230000001105 regulatory effect Effects 0.000 claims description 6
- 238000012706 support-vector machine Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 5
- 230000008569 process Effects 0.000 claims description 5
- 230000000694 effects Effects 0.000 claims description 4
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 238000011217 control strategy Methods 0.000 claims description 3
- 238000013135 deep learning Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 238000013508 migration Methods 0.000 claims description 3
- 230000005012 migration Effects 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 3
- 238000002360 preparation method Methods 0.000 abstract description 3
- 230000006978 adaptation Effects 0.000 abstract description 2
- 238000007726 management method Methods 0.000 description 8
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 238000010223 real-time analysis Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/092—Reinforcement learning
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D23/00—Control of temperature
- G05D23/19—Control of temperature characterised by the use of electric means
- G05D23/20—Control of temperature characterised by the use of electric means with sensing elements having variation of electric or magnetic properties with change of temperature
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
Abstract
The invention provides a data center energy-saving control method based on an AI algorithm, which comprises the following steps of data acquisition monitoring and data preprocessing, and performing an intelligent prediction model according to the AI algorithm; according to the intelligent prediction model structure dynamic resource scheduling, temperature regulation and control are carried out; real-time monitoring and feedback are carried out, and energy management is optimized; according to application scene adaptation, the system can acquire key parameter data in real time, can predict future load and energy consumption trend, helps to make corresponding energy consumption adjustment preparation in advance, and can intelligently adjust resource configuration of a data center through real-time monitoring and prediction, including the on-off state of a server so as to adapt to actual requirements, thereby reducing energy consumption, making real-time decisions by utilizing a deep reinforcement learning model, and dynamically adjusting equipment states so as to find the balance of optimal performance and energy cost.
Description
Technical Field
The invention relates to the technical field of data centers, in particular to a data center energy-saving control method based on an AI algorithm.
Background
Data centers are a facility specifically designed to store, process and manage large amounts of electronic data, which plays a key role of modern information technology infrastructure, providing reliable data storage and processing services for various organizations and enterprises, generally including infrastructure such as large servers, network devices, storage systems and cooling devices, aiming to ensure continuous efficient operation of the servers while maintaining safety and reliability of the data, which employ highly advanced technologies such as virtualization, cloud computing and automation management systems to improve efficiency and scalability of the data center, main functions of the data center including data storage, processing, backup and restoration, while providing high-speed network connection enabling users to quickly access information stored in the data center, and in the digital age, the data center is critical in supporting enterprise operations, cloud services, large data analysis, etc., while operating costs of the data center mainly include energy costs, and the operation of the servers and devices in the data center under high-load environments may cause hardware failures and damages in advance, so that an AI-based data center energy saving method is particularly required.
However, in the existing data center, most of the data centers generally use a large number of servers and devices in use, resulting in high power consumption, the devices are operated in all weather, a large amount of power is used to maintain normal operation and cooling, and there may be space waste of hardware resources due to rapid development of hardware and expansion of the data center. Some servers may not fully utilize their computing power, resulting in wasted resources.
Disclosure of Invention
The invention aims to provide an AI algorithm-based data center energy-saving control method, so as to solve the problems that in the prior data center, in the use process, the running cost of the data center mainly comprises energy cost, and the running of a server and equipment in the data center under a high-temperature and high-load environment can cause hardware faults and early damages.
In order to achieve the above object, the present invention provides a data center energy-saving control method based on AI algorithm, the control method comprising the steps of:
s1, data acquisition monitoring and data preprocessing, and performing an intelligent prediction model according to an AI algorithm;
s2, carrying out temperature regulation and control according to the intelligent prediction model structure dynamic resource scheduling;
s3, monitoring and feeding back in real time to optimize energy management;
s4, adapting according to the application scene.
Preferably, the step S1 includes:
s1.1, data acquisition and monitoring, namely deploying sensors such as a temperature sensor, a humidity sensor, power monitoring equipment and the like in a data center, and acquiring key parameter data in real time;
s1.2, preprocessing data, namely processing the acquired data by utilizing a preprocessing algorithm, wherein the preprocessing algorithm comprises outlier removal, data smoothing and normalization, so that the quality and consistency of the data are ensured;
s1.3, an intelligent prediction model is used for establishing a machine learning model, a long-short-time memory network (LSTM) is used for predicting future load and energy consumption trend, and external factors including weather and service cycle are considered by the model.
Preferably, the S1.1 sensor collects data in real time, including key parameters of temperature (T), humidity (H), and power consumption (P), and generates time series data
$D＝{(t_1,T_1,H_1,P_1),(t_2,T_2,H_2,P_2),……}$；
The step S1.2 of data preprocessing is based on a machine learning algorithm, and abnormal values are removed;
applying a smoothing algorithm, using a moving average, to reduce the effect of noise on the model;
normalizing the data, mapping the data with different scales to the same numerical range, and normalizing parameters such as temperature, humidity and the like to a [0,1] interval;
the intelligent predictive model of step S1.3 forms a training set using historical data,
$D_{train}＝{(t_1,T_1,H_1,P_1),……}$；
extracting characteristics affecting load and energy consumption, and considering external factors such as time correlation, seasonality, weather and the like;
building a long-time memory network (LSTM) model by using a deep learning framework;
training an LSTM model by using gradient descent through an optimization algorithm, and adjusting model parameters to minimize a prediction error;
and verifying the performance of the model by using the verification set data, adjusting the model structure and the super parameters to improve the prediction accuracy, and embedding the trained model into a real-time data stream so as to continuously predict future loads and energy consumption.
Preferably, the step S2 includes:
s2.1, dynamic resource scheduling utilizes a reinforcement learning algorithm to monitor the work load of a data center in real time and dynamically adjust the opening and closing states of a server so as to adapt to actual demands;
s2.2, temperature regulation and control are based on a PID control algorithm, and the temperature of the data center is regulated in real time by regulating the cooling equipment, the fan and other systems, so that the cooling efficiency is improved.
Preferably, the step S2.1 reinforcement learning algorithm selects a Deep Q Network (DQN), defines executable actions, opens or closes a server, and possible load migration according to the relationship between the modeling state and the actions, so that the server can encourage to reduce energy consumption, improve resource utilization rate, consider performance indexes of a data center, train by using historical data in actual operation, enable a reinforcement learning model to adapt to different workload situations, make decisions in real time by using the reinforcement learning model on the basis of monitoring the workload of the data center in real time, and adjust the state of the server to achieve energy saving effect;
the step S2.2 is to arrange a temperature sensor, collect temperature information of a data center in real time, establish a PID control model by utilizing historical temperature data, select a proportional-integral-derivative (PID) control algorithm for adjusting systems such as cooling equipment and fans in real time, establish the PID control model by utilizing the historical temperature data, determine a proportional coefficient, an integral coefficient and a derivative coefficient, design a temperature control strategy, set a target range of temperature according to a target range of set temperature, adjust working states of the cooling equipment and the fans according to actual temperature, monitor the temperature in real time when the data center is operated, and adjust the cooling equipment and the fans by utilizing the PID control algorithm to maintain stable temperature.
Preferably, the step S3 includes:
s3.1, monitoring and feeding back in real time, continuously monitoring the performance and the energy consumption of the data center, analyzing in real time through a supervised learning algorithm, and adjusting algorithm parameters according to real-time data to provide feedback information;
s3.2, optimizing energy management, namely using a deep reinforcement learning algorithm to find the best balance between performance and energy cost by means of adjusting the voltage frequency of equipment, adopting a sleep mode and the like.
Preferably, the step S3.1 monitors the energy consumption of the data center in real time by using the power monitoring device, records the power consumption of each device, and performs preprocessing on the collected data, including outlier removal, data smoothing and normalization to ensure the quality and consistency of the data, and the supervised learning algorithm selects a Support Vector Machine (SVM) for analyzing the performance and energy consumption data of the data center in real time, and provides feedback information, possibly including performance adjustment suggestions, optimization strategies, etc., for subsequent adjustment and optimization according to the real-time analysis result;
the S3.2 deep reinforcement learning algorithm selects a deep reinforcement learning network (DRL) to model decision processes of voltage frequency adjustment, sleep mode and the like of equipment, define a state space of equipment adjustment, including voltage frequency, working state and the like of the equipment, define actions which can be executed, including voltage frequency adjustment, sleep mode switching and the like, design a reward function, consider factors of equipment performance, energy cost, overall data center performance and the like, make real-time decisions by using a deep reinforcement learning model, dynamically adjust the voltage frequency and the working state of the equipment to find the balance of the optimal performance and the energy cost, and make real-time decisions by using the deep reinforcement learning model in actual operation, dynamically adjust the voltage frequency and the working state of the equipment to find the balance of the optimal performance and the energy cost.
Preferably, the step S4 includes:
s4.1, adapting application scenes, and adjusting algorithm parameters and strategies according to different application scenes so as to adapt to energy-saving control requirements under different workloads.
Preferably, the step S4.1 uses a machine learning algorithm to identify the workload and the application scenario of the current data center in real time, understand the working environment where the current data center is located, and adjust relevant parameters in the AI algorithm and state space, action space and the like in the deep reinforcement learning model according to different application scenarios so as to ensure that the algorithm is better adapted to the current workload.
Compared with the prior art, the invention has the beneficial effects that: according to the data center energy-saving control method based on the AI algorithm, future load and energy consumption trend can be predicted according to application scene adaptation, corresponding energy consumption adjustment preparation is facilitated to be made in advance, through real-time monitoring and prediction, the system can intelligently adjust the resource configuration of the data center, including the on-off state of the server, so as to adapt to actual requirements, and therefore energy consumption is reduced, enterprises can reduce electricity consumption cost, profitability is improved, and through AI intelligent scheduling and management mechanisms, the data center can more efficiently utilize hardware resources, calculation efficiency is improved, and waste is reduced.
Drawings
FIG. 1 is a schematic flow chart of a data center energy-saving control method based on an AI algorithm;
FIG. 2 is a flow chart of step S1 of the present invention;
FIG. 3 is a flow chart of step S2 of the present invention;
fig. 4 is a flow chart of step S3 of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1-4, the present invention provides a data center energy-saving control method based on AI algorithm, comprising the following steps:
s1, data acquisition monitoring and data preprocessing, and performing an intelligent prediction model according to an AI algorithm;
s2, carrying out temperature regulation and control according to the intelligent prediction model structure dynamic resource scheduling;
s3, monitoring and feeding back in real time to optimize energy management;
s4, adapting according to the application scene.
Further, the step S1 includes:
s1.1, data acquisition and monitoring, namely deploying sensors such as a temperature sensor, a humidity sensor, power monitoring equipment and the like in a data center, and acquiring key parameter data in real time;
s1.2, preprocessing data, namely processing the acquired data by utilizing a preprocessing algorithm, wherein the preprocessing algorithm comprises outlier removal, data smoothing and normalization, so that the quality and consistency of the data are ensured;
s1.3, an intelligent prediction model is used for establishing a machine learning model, a long-short-time memory network (LSTM) is used for predicting future load and energy consumption trend, and the model considers external factors including weather and service period;
through the data acquisition and monitoring of S1.1 step, the system can acquire key parameter data in real time, so that the monitoring of the state of the data center is more comprehensive and accurate, the data preprocessing in S1.2 ensures that the acquired data is subjected to outlier removal, smoothing and normalization processing, so that the quality and consistency of the data are improved, the subsequent analysis is more reliable, the intelligent prediction model in S1.3 utilizes machine learning, particularly long and short time memory network (LSTM), future load and energy consumption trend can be predicted, corresponding energy consumption adjustment preparation is facilitated in advance, and through the real-time monitoring and prediction, the system can intelligently adjust the resource configuration of the data center, including the opening and closing states of a server, so as to adapt to the actual demands, and thereby the energy consumption is reduced.
Further, the S1.1 sensor collects data in real time, including key parameters of temperature (T), humidity (H) and power consumption (P), and generates time series data
$D＝{(t_1,T_1,H_1,P_1),(t_2,T_2,H_2,P_2),……}$；
The step S1.2 of data preprocessing is based on a machine learning algorithm, and abnormal values are removed;
applying a smoothing algorithm, using a moving average, to reduce the effect of noise on the model;
normalizing the data, mapping the data with different scales to the same numerical range, and normalizing parameters such as temperature, humidity and the like to a [0,1] interval;
the intelligent predictive model of step S1.3 forms a training set using historical data,
$D_{train}＝{(t_1,T_1,H_1,P_1),……}$；
extracting characteristics affecting load and energy consumption, and considering external factors such as time correlation, seasonality, weather and the like;
building a long-time memory network (LSTM) model by using a deep learning framework;
training an LSTM model by using gradient descent through an optimization algorithm, and adjusting model parameters to minimize a prediction error;
and verifying the performance of the model by using the verification set data, adjusting the model structure and the super parameters to improve the prediction accuracy, and embedding the trained model into a real-time data stream so as to continuously predict future loads and energy consumption.
Further, the step S2 includes:
s2.1, dynamic resource scheduling utilizes a reinforcement learning algorithm to monitor the work load of a data center in real time and dynamically adjust the opening and closing states of a server so as to adapt to actual demands;
s2.2, temperature regulation and control are based on a PID control algorithm, and the temperature of the data center is regulated in real time by regulating systems such as cooling equipment and fans, so that the cooling efficiency is improved;
the reinforcement learning algorithm in S2.1 monitors the work load of the data center in real time, dynamically adjusts the opening and closing states of the server, and can flexibly configure resources according to actual demands, thereby being beneficial to reducing unnecessary energy consumption, and the reinforcement learning algorithm and the PID control algorithm are applied, so that the system has more intellectualization and stability, can better adapt to the change and fluctuation of the data center, and improves the stability of the whole system.
Further, the step S2.1 reinforcement learning algorithm selects a Deep Q Network (DQN) to model the relationship between states and actions, define actions that can be performed, turn on or off servers, and possibly load migration, so that it can encourage reduction of energy consumption, increase of resource utilization, and consider performance indexes of the data center;
and step S2.2, deploying a temperature sensor, collecting temperature information of the data center in real time, utilizing historical temperature data, establishing a PID control model, designing a temperature control strategy, setting a target range of temperature, adjusting working states of cooling equipment and fans according to actual temperature, monitoring the temperature in real time when the data center is operated, and adjusting the cooling equipment and the fans by utilizing a PID control algorithm so as to maintain stable temperature.
Further, the step S3 includes:
s3.1, monitoring and feeding back in real time, continuously monitoring the performance and the energy consumption of the data center, analyzing in real time through a supervised learning algorithm, and adjusting algorithm parameters according to real-time data to provide feedback information;
s3.2, optimizing energy management, namely, using a deep reinforcement learning algorithm, and finding the best balance between performance and energy cost by means of adjusting the voltage frequency of equipment, adopting a sleep mode and the like;
the energy management optimization is carried out by adopting a deep reinforcement learning algorithm, and the optimal balance between the performance and the energy cost is found by means of adjusting the voltage frequency of equipment, adopting a sleep mode and the like. This results in a data center that reduces energy consumption while maintaining performance and improves energy efficiency.
Further, the step S3.1 is to use power monitoring equipment to monitor the energy consumption condition of the data center in real time, record the power consumption of each equipment, and select a Support Vector Machine (SVM) for monitoring the learning algorithm to analyze the performance and the energy consumption data of the data center in real time;
and 3.2, selecting a deep reinforcement learning algorithm and a deep reinforcement learning network (DRL), carrying out real-time decision by using the deep reinforcement learning model according to decision processes such as voltage frequency adjustment and sleep mode of modeling equipment, and dynamically adjusting the voltage frequency and working state of the equipment so as to find the balance of the optimal performance and energy cost.
Further, the step S4 includes:
s4.1, adapting the application scene, and adjusting algorithm parameters and strategies according to different application scenes so as to adapt to energy-saving control requirements under different workloads. The personalized optimization enables the data center to more effectively perform energy-saving management under different use situations, and improves the overall performance.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (8)
1. The data center energy-saving control method based on the AI algorithm is characterized by comprising the following steps of: the control method comprises the following steps:
s1, data acquisition monitoring and data preprocessing, and performing an intelligent prediction model according to an AI algorithm;
s2, carrying out temperature regulation and control according to the intelligent prediction model structure dynamic resource scheduling;
s3, monitoring and feeding back in real time to optimize energy management;
s4, adapting according to the application scene.
2. The AI algorithm-based data center energy-saving control method of claim 1, wherein: the step S1 includes:
s1.1, data acquisition and monitoring, namely deploying sensors such as a temperature sensor, a humidity sensor, power monitoring equipment and the like in a data center, and acquiring key parameter data in real time;
s1.2, preprocessing data, namely processing the acquired data by utilizing a preprocessing algorithm, wherein the preprocessing algorithm comprises outlier removal, data smoothing and normalization, so that the quality and consistency of the data are ensured;
s1.3, an intelligent prediction model is used for establishing a machine learning model, a long-short-time memory network (LSTM) is used for predicting future load and energy consumption trend, and external factors including weather and service cycle are considered by the model.
3. The AI algorithm-based data center energy-saving control method of claim 2, wherein: the S1.1 sensor collects data in real time, wherein the data comprise key parameters of temperature (T), humidity (H) and power consumption (P), and generates time sequence data
$D＝{(t_1,T_1,H_1,P_1),(t_2,T_2,H_2,P_2),……}$；
The step S1.2 of data preprocessing is based on a machine learning algorithm, and abnormal values are removed;
applying a smoothing algorithm, using a moving average, to reduce the effect of noise on the model;
normalizing the data, mapping the data with different scales to the same numerical range, and normalizing parameters such as temperature, humidity and the like to a [0,1] interval;
the intelligent predictive model of step S1.3 forms a training set using historical data,
$D_{train}＝{(t_1,T_1,H_1,P_1),……}$；
extracting characteristics affecting load and energy consumption, and considering external factors such as time correlation, seasonality, weather and the like;
building a long-time memory network (LSTM) model by using a deep learning framework;
training an LSTM model by using gradient descent through an optimization algorithm, and adjusting model parameters to minimize a prediction error;
and verifying the performance of the model by using the verification set data, adjusting the model structure and the super parameters to improve the prediction accuracy, and embedding the trained model into the real-time data stream.
4. The AI algorithm-based data center energy-saving control method of claim 1, wherein: the step S2 includes:
s2.1, dynamic resource scheduling utilizes a reinforcement learning algorithm to monitor the work load of a data center in real time and dynamically adjust the opening and closing states of a server so as to adapt to actual demands;
s2.2, temperature regulation and control are based on a PID control algorithm, and the temperature of the data center is regulated in real time by regulating the cooling equipment, the fan and other systems, so that the cooling efficiency is improved.
5. The AI-algorithm-based data center energy-saving control method of claim 4, wherein: the step S2.1 reinforcement learning algorithm selects a Deep Q Network (DQN) to model the relationship between state and actions, define actions that can be performed, turn on or off servers, and possibly load migration;
and step S2.2, deploying a temperature sensor, collecting temperature information of the data center in real time, utilizing historical temperature data, establishing a PID control model, designing a temperature control strategy, setting a target range of temperature, adjusting working states of cooling equipment and fans according to actual temperature, monitoring the temperature in real time when the data center is operated, and adjusting the cooling equipment and the fans by utilizing a PID control algorithm.
6. The AI algorithm-based data center energy-saving control method of claim 1, wherein: the step S3 includes:
s3.1, monitoring and feeding back in real time, continuously monitoring the performance and the energy consumption of the data center, analyzing in real time through a supervised learning algorithm, and adjusting algorithm parameters according to real-time data to provide feedback information;
s3.2, optimizing energy management, namely using a deep reinforcement learning algorithm to find the best balance between performance and energy cost by means of adjusting the voltage frequency of equipment, adopting a sleep mode and the like.
7. The AI-algorithm-based data center energy-saving control method of claim 6, wherein: step S3.1, monitoring the energy consumption condition of the data center in real time by using power monitoring equipment, recording the power consumption of each equipment, and selecting a Support Vector Machine (SVM) by using a supervised learning algorithm for analyzing the performance and energy consumption data of the data center in real time;
the 3.2 deep reinforcement learning algorithm selects a deep reinforcement learning network (DRL) to model the voltage frequency adjustment, sleep mode and other decision processes of the equipment, and makes real-time decisions by using the deep reinforcement learning model to dynamically adjust the voltage frequency and the working state of the equipment.
8. The AI algorithm-based data center energy-saving control method of claim 1, wherein: the step S4 includes:
s4.1, adapting application scenes, and adjusting algorithm parameters and strategies according to different application scenes so as to adapt to energy-saving control requirements under different workloads.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311720120.4A CN117850494A (en) | 2023-12-14 | 2023-12-14 | Data center energy-saving control method based on AI algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311720120.4A CN117850494A (en) | 2023-12-14 | 2023-12-14 | Data center energy-saving control method based on AI algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117850494A true CN117850494A (en) | 2024-04-09 |
Family
ID=90542702
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311720120.4A Pending CN117850494A (en) | 2023-12-14 | 2023-12-14 | Data center energy-saving control method based on AI algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117850494A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118278620A (en) * | 2024-05-29 | 2024-07-02 | 湘江实验室 | Infrastructure electric energy consumption monitoring and management system |
| CN118446839A (en) * | 2024-04-28 | 2024-08-06 | 中交一公局集团有限公司 | Energy consumption real-time monitoring system and method based on green energy-saving building |
-
2023
- 2023-12-14 CN CN202311720120.4A patent/CN117850494A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118446839A (en) * | 2024-04-28 | 2024-08-06 | 中交一公局集团有限公司 | Energy consumption real-time monitoring system and method based on green energy-saving building |
| CN118278620A (en) * | 2024-05-29 | 2024-07-02 | 湘江实验室 | Infrastructure electric energy consumption monitoring and management system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117850494A (en) | Data center energy-saving control method based on AI algorithm | |
| CN109800066B (en) | Energy-saving scheduling method and system for data center | |
| US10429921B2 (en) | Datacenter power management optimizations | |
| CN102326134B (en) | Energy-aware server admin | |
| CN118231883B (en) | Battery management method and system for liquid cooling energy storage system | |
| CN117076882A (en) | Dynamic prediction management method for cloud service resources | |
| CN118364887B (en) | Intelligent scheduling method and system for energy management monitoring | |
| US20160313780A1 (en) | Multivariable control for power-latency management to support optimization of data centers or other systems | |
| CN103279184A (en) | Cloud server energy-saving system based on data mining | |
| CN117650628B (en) | Energy efficiency management system based on self-adaptive preconditioning scene | |
| CN113225994B (en) | Intelligent air conditioner control method facing data center | |
| CN117891681A (en) | Storage device dynamic configuration method and system based on energy consumption prediction | |
| CN111914000B (en) | Server power capping method and system based on power consumption prediction model | |
| KR20220061709A (en) | Reward generation method to reduce peak load of electric power and action control apparatus performing the same method | |
| CN116880976A (en) | AI-identified computing power distribution system | |
| CN113632132A (en) | Computer-aided energy management method and energy management system | |
| CN118093301A (en) | Method and device for regulating temperature of server cluster | |
| CN118502885B (en) | Thermomigration method, equipment, program product and medium | |
| CN118446384B (en) | LSTM-based energy-saving optimization method, system and device for building heating and ventilation system | |
| CN118132244B (en) | Energy-saving regulation and control method, system, equipment and medium thereof for intelligent data center | |
| Rueb et al. | 6G and the Sustainability Aspect: Exploiting Surplus Renewable Energy for Distributed Learning Clusters in 6G Networks | |
| CN117891559A (en) | Method for intelligently predicting K8s resource scheduling | |
| CN118466238A (en) | Intelligent household system | |
| Liu | Constructing a Cloud Computing Elastic Scaling Technology Based on Load Prediction | |
| CN118783428A (en) | Power grid equipment scheduling control method and system based on artificial intelligence |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |