CN116127840B - Data center load prediction method based on data driving - Google Patents
Data center load prediction method based on data driving Download PDFInfo
- Publication number
- CN116127840B CN116127840B CN202310015884.7A CN202310015884A CN116127840B CN 116127840 B CN116127840 B CN 116127840B CN 202310015884 A CN202310015884 A CN 202310015884A CN 116127840 B CN116127840 B CN 116127840B
- Authority
- CN
- China
- Prior art keywords
- neural network
- server
- data
- power consumption
- model
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/02—Data centres
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Computer Hardware Design (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Geometry (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Air Conditioning Control Device (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses a data center load prediction method based on data driving, which comprises the following steps: training and testing using LSTM neural networks; predicting the CPU utilization rate of the server by using the trained neural network model; fitting the CPU utilization rate of the server and the heating value of the server to obtain a function relationship of the CPU utilization rate and the heating value of the server; training and testing the neural network, and predicting the electricity consumption of the air conditioner in the machine room by using a trained neural network model; training and testing the neural network; predicting the power consumption of the server by using the trained neural network model; and summarizing the power consumption of the server, the power consumption of the air conditioner and other power consumption to obtain the power consumption of the whole data center. The application selects the physical mechanism model expression in the part with the clear mathematical relation expression, adopts the data driving model expression in the nonlinear relation and load prediction part, and greatly improves the modeling precision and the prediction accuracy.
Description
Technical Field
The application belongs to the technical field of data center energy efficiency management, and particularly relates to a data center load prediction method based on data driving.
Background
The construction speed of the data center in China is rapid, the total quantity of the data centers exceeds 40 ten thousand, and the annual power consumption of the data center exceeds 5% of the total power consumption of the society. The energy costs paid per year can be as high as tens of millions of dollars for a single data center. The research of the load energy consumption characteristic is a precondition for improving the electric energy utilization efficiency of the data center. The data center comprises electric and thermal loads, has high randomness and is difficult to model and predict. As shown in fig. 1, there are two general classes of load modeling methods: the method is a time sequence method, a neural network method and the like, and the method does not consider an energy consumption generation mechanism, and is essentially used for analyzing and processing historical data, so that the load prediction error is larger; yet another category is to build a mathematical model based on the physical mechanism of energy consumption generation, but many nonlinear features cannot be expressed.
Disclosure of Invention
In view of the above, the application provides a data-driven data center load prediction method, which is based on a data center energy consumption generation mechanism, and on the basis, a data-driven technology is utilized for modeling, so that the load modeling and prediction accuracy can be effectively improved.
Specifically, the application discloses a data center load prediction method based on data driving, which comprises the following steps:
randomly selecting 80% -90% of the collected first historical data as a training set, and training by using an LSTM neural network; randomly selecting 10% -20% of historical data as a test set to test the trained neural network to see whether the error meets the requirement; if the error requirement is not met, the neural network is regulated by changing the super parameters of the neural network; finally, predicting the CPU utilization rate of the server by using the trained LSTM neural network model;
fitting the CPU utilization rate of the server and the heating value of the server to obtain a function relationship between the CPU utilization rate of the server and the heating value of the server, and establishing a mathematical model; secondly, randomly selecting 80% -90% of the collected second historical data as a training set to train the neural network; randomly selecting 10% -20% of historical data as a test set to test the trained neural network to see whether the error meets the requirement; if the error requirement is not met, the neural network is regulated by changing the super parameters of the neural network; predicting the electricity consumption of the air conditioner in the machine room by using the trained RNN neural network model;
randomly selecting 80% -90% of the collected third historical data as a training set to train the neural network; randomly selecting 10% -20% of historical data as a test set to test the trained RNN neural network to see whether the error meets the requirement; if the error requirement is not met, the neural network is regulated by changing the super parameters of the neural network; predicting the power consumption of the server by using the trained neural network model;
and summarizing the power consumption of the server, the power consumption of the air conditioner and other power consumption to obtain the power consumption of the whole data center.
Further, the input data in the first historical data is the current region, season, time period and user number; the output data is the CPU utilization rate of the server; the input data in the second historical data is the heat productivity of the server, and the output data is the air conditioner electric quantity of the machine room; and the input data in the third historical data is the CPU utilization rate of the server, and the output data is the power consumption of the server.
Further, the super parameters of the neural network include the number of layers of the neural network, the number of neurons, and the learning rate.
Further, the server CPU utilization and the server calorific value are fitted, and the established mathematical model is as follows:
wherein a, b, c are parameters to be fitted, CPU usage For CPU utilization, hot is the server heating value.
Further, the RNN neural network model is one of a unidirectional RNN model, a bidirectional RNN model and a depth RNN model.
The beneficial effects of the application are as follows:
the model expression from the historical data of the general system to the electric energy consumption in the traditional modeling mode is refined;
selecting a physical mechanism model expression in a part with an explicit mathematical relation expression, and adopting a data driving model expression in a nonlinear relation and load prediction part;
compared with the traditional model, the obtained data center energy consumption model has the advantages that modeling precision and prediction accuracy are improved greatly.
Drawings
FIG. 1 is a diagram of a conventional data center energy consumption prediction model;
FIG. 2 is a graph of a data center energy consumption prediction model of the present application.
Detailed Description
The application is further described below with reference to the accompanying drawings, without limiting the application in any way, and any alterations or substitutions based on the teachings of the application are intended to fall within the scope of the application.
The data center load prediction method comprises the following steps:
s1: firstly, 80% -90% of collected historical data (input: current region, season, period, number of users and the like; output: server CPU utilization rate) is randomly selected as a training set to train the LSTM neural network. Then, 10% -20% of the historical data are randomly selected as a test set to test the trained neural network, and whether the error meets the requirement is judged; if the error requirement is not met, the neural network is regulated by changing the number of layers of the neural network, the number of neurons, the learning rate and other factors. And finally, predicting the CPU utilization rate of the server by using the trained LSTM neural network model.
The long short-term memory network LSTM (long short-term memory) of the present application is a variant of RNN, whose core concept is the cellular status and "gate" structure. The cell state corresponds to the path of information transmission, allowing information to be transferred in sequence. It can be regarded as a "memory" of the network. In theory, the cell state can always convey relevant information during sequence processing. Thus, even the information of the earlier time step can be carried into the cells of the later time step, which overcomes the influence of the short-term memory. The addition and removal of information is accomplished through a "gate" structure that learns what information is saved or forgotten during the training process.
S2: firstly, fitting the utilization rate of a CPU of a server and the heating value of the server to obtain a function relation of the utilization rate of the CPU of the server and the heating value of the server, and establishing a mathematical model. And secondly, randomly selecting 80% -90% of collected historical data (input: server heating value and output: machine room air-conditioning electric quantity) as a training set to train the neural network. Then, 10% -20% of the historical data are randomly selected as a test set to test the trained neural network, and whether the error meets the requirement is judged; if the error requirement is not met, the neural network is regulated by changing the number of layers of the neural network, the number of neurons, the learning rate and other factors. And finally, predicting the electricity consumption of the air conditioner in the machine room by using the trained RNN neural network model.
The CPU utilization rate of the server and the calorific value of the server are fitted, and the established mathematical model is as follows:
wherein a, b, c are parameters to be fitted, CPU usage For CPU utilization, hot is the server heating value.
S3: firstly, the collected historical data (input: server CPU utilization rate; output: server power consumption) is randomly selected from 80% -90% as training set to train the neural network. Then, 10% -20% of the historical data are randomly selected as a test set to test the trained RNN neural network, and whether the error meets the requirement is judged; if the error requirement is not met, the neural network is regulated by changing the number of layers of the neural network, the number of neurons, the learning rate and other factors. And finally, predicting the power consumption of the server by using the trained neural network model.
And S4, finally, summarizing the power consumption of the server, the power consumption of the air conditioner and other power consumption to obtain the power consumption of the whole data center. The other power consumption is relatively fixed and is obtained through historical data estimation.
The whole structure of the RNN circulating neural network in the application is divided into 3 layers: an input layer, a hidden layer and an output layer. Wherein the state of the hidden layer at time t is not only related to the input but also to the state of the hidden layer at the previous time.
Since the hidden layer has one more input from itself, this layer is called a loop layer. There are also various types of recurrent neural networks, the direction based on the circulation is divided into: unidirectional circulating neural network and bidirectional circulating neural network, the depth based on circulation is divided into: a recurrent neural network and a deep recurrent neural network.
The working principle of the application is as follows:
training the LSTM neural network by using historical data (region, season, time period, user number and the like) to obtain a CPU utilization rate prediction model;
performing functional relation fitting by means of CPU utilization rate and server heating value data to obtain mathematical relation expressions of the CPU utilization rate and the server heating value data;
training the RNN neural network by utilizing historical data (server heating and machine room air conditioner electric quantity) to obtain an air conditioner electric quantity model;
training the RNN neural network by utilizing historical data (CPU utilization rate and server power consumption) to obtain a server power model;
finally, summarizing the power consumption of the server, the power consumption of the air conditioner and other power consumption to obtain the power consumption of the whole data center;
when the model runs online, information such as user data, CPU utilization rate, server heating value, air conditioner electricity consumption, server electricity consumption and the like is collected in real time, and each sub-model is corrected.
The application also discloses a data center load prediction system, which comprises:
server CPU utilization prediction module: firstly, 80% -90% of collected historical data (input: current region, season, period, number of users and the like; output: server CPU utilization rate) is randomly selected as a training set to train the LSTM neural network. Then, 10% -20% of the historical data are randomly selected as a test set to test the trained neural network, and whether the error meets the requirement is judged; if the error requirement is not met, the neural network is regulated by changing the number of layers of the neural network, the number of neurons, the learning rate and other factors. And finally, predicting the CPU utilization rate of the server by using the trained LSTM neural network model.
The power consumption prediction module of the air conditioner in the machine room comprises: firstly, fitting the utilization rate of a CPU of a server and the heating value of the server to obtain a function relation of the utilization rate of the CPU of the server and the heating value of the server, and establishing a mathematical model. And secondly, randomly selecting 80% -90% of collected historical data (input: server heating value and output: machine room air-conditioning electric quantity) as a training set to train the neural network. Then, 10% -20% of the historical data are randomly selected as a test set to test the trained neural network, and whether the error meets the requirement is judged; if the error requirement is not met, the neural network is regulated by changing the number of layers of the neural network, the number of neurons, the learning rate and other factors. And finally, predicting the electricity consumption of the air conditioner in the machine room by using the trained RNN neural network model.
And a server electricity consumption prediction module: firstly, the collected historical data (input: server CPU utilization rate; output: server power consumption) is randomly selected from 80% -90% as training set to train the neural network. Then, 10% -20% of the historical data are randomly selected as a test set to test the trained RNN neural network, and whether the error meets the requirement is judged; if the error requirement is not met, the neural network is regulated by changing the number of layers of the neural network, the number of neurons, the learning rate and other factors. And finally, predicting the power consumption of the server by using the trained neural network model.
And the comprehensive prediction module is used for: and summarizing the power consumption of the server, the power consumption of the air conditioner and other power consumption to obtain the power consumption of the whole data center.
The beneficial effects of the application are as follows:
the model expression from the historical data of the general system to the electric energy consumption in the traditional modeling mode is refined;
selecting a physical mechanism model expression in a part with an explicit mathematical relation expression, and adopting a data driving model expression in a nonlinear relation and load prediction part;
compared with the traditional model, the obtained data center energy consumption model has the advantages that modeling precision and prediction accuracy are improved greatly.
The word "preferred" is used herein to mean serving as an example, instance, or illustration. Any aspect or design described herein as "preferred" is not necessarily to be construed as advantageous over other aspects or designs. Rather, use of the word "preferred" is intended to present concepts in a concrete fashion. The term "or" as used in this disclosure is intended to mean an inclusive "or" rather than an exclusive "or". That is, unless specified otherwise or clear from the context, "X uses a or B" is intended to naturally include any of the permutations. That is, if X uses A; x is B; or X uses both A and B, then "X uses A or B" is satisfied in any of the foregoing examples.
Moreover, although the disclosure has been shown and described with respect to one or more implementations, equivalent alterations and modifications will occur to others skilled in the art based upon a reading and understanding of this specification and the annexed drawings. The present disclosure includes all such modifications and alterations and is limited only by the scope of the following claims. In particular regard to the various functions performed by the above described components (e.g., elements, etc.), the terms used to describe such components are intended to correspond, unless otherwise indicated, to any component which performs the specified function of the described component (e.g., that is functionally equivalent), even though not structurally equivalent to the disclosed structure which performs the function in the herein illustrated exemplary implementations of the disclosure. Furthermore, while a particular feature of the disclosure may have been disclosed with respect to only one of several implementations, such feature may be combined with one or other features of the other implementations as may be desired and advantageous for a given or particular application. Moreover, to the extent that the terms "includes," has, "" contains, "or variants thereof are used in either the detailed description or the claims, such terms are intended to be inclusive in a manner similar to the term" comprising.
The functional units in the embodiment of the application can be integrated in one processing module, or each unit can exist alone physically, or a plurality of or more than one unit can be integrated in one module. The integrated modules may be implemented in hardware or in software functional modules. The integrated modules may also be stored in a computer readable storage medium if implemented in the form of software functional modules and sold or used as a stand-alone product. The above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, or the like. The above-mentioned devices or systems may perform the storage methods in the corresponding method embodiments.
In summary, the foregoing embodiment is an implementation of the present application, but the implementation of the present application is not limited to the embodiment, and any other changes, modifications, substitutions, combinations, and simplifications made by the spirit and principles of the present application should be equivalent to the substitution manner, and all the changes, modifications, substitutions, combinations, and simplifications are included in the protection scope of the present application.
Claims (3)
1. A data center load prediction method based on data driving, comprising the steps of:
taking the collected current region, season, time period and user data as first historical data, randomly selecting a part of the first historical data as a training set, and training by using an LSTM neural network; randomly selecting a part of the first historical data as a test set to test the trained neural network to see whether the error meets the requirement; if the error requirement is not met, the neural network is regulated by changing the super parameters of the neural network; finally, predicting the utilization rate of the CPU of the server by using the trained LSTM neural network model, and outputting data as the utilization rate of the CPU of the server;
fitting the CPU utilization rate of the server and the heating value of the server to obtain a functional relation of the CPU utilization rate of the server and the heating value of the server, and establishing a mathematical model as follows:
wherein a, b, c are parameters to be fitted, CPU usage For CPU utilization, hot is the heating value of the server;
secondly, taking the collected server heating value data as second historical data, and randomly selecting a part of the second historical data as a training set to train the neural network; randomly selecting a part of the second historical data as a test set to test the trained neural network to see whether the error meets the requirement; if the error requirement is not met, the neural network is regulated by changing the super parameters of the neural network; predicting the electricity consumption of the air conditioner of the machine room by using the trained RNN neural network model, and outputting the electricity consumption of the air conditioner of the machine room;
taking the collected CPU utilization rate of the server as third historical data, and randomly selecting a part of the third historical data as a training set to train the neural network; randomly selecting a part of the third historical data as a test set to test the trained RNN neural network to see whether the error meets the requirement; if the error requirement is not met, the neural network is regulated by changing the super parameters of the neural network; predicting the power consumption of the server by using the trained neural network model, and outputting the power consumption of the server;
summarizing the power consumption of the server, the power consumption of the air conditioner and other power consumption to obtain the power consumption of the whole data center;
and correcting each model according to the user data, the CPU utilization rate, the server heating value data, the air conditioner electric quantity and the server electric quantity.
2. The data-driven based data center load prediction method according to claim 1, wherein the super parameters of the neural network include the number of layers of the neural network, the number of neurons, and the learning rate.
3. The data-driven data center load prediction method according to claim 1, wherein the RNN neural network model is one of a unidirectional RNN model, a bidirectional RNN model, and a depth RNN model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310015884.7A CN116127840B (en) | 2023-01-05 | 2023-01-05 | Data center load prediction method based on data driving |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310015884.7A CN116127840B (en) | 2023-01-05 | 2023-01-05 | Data center load prediction method based on data driving |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116127840A CN116127840A (en) | 2023-05-16 |
| CN116127840B true CN116127840B (en) | 2023-09-22 |
Family
ID=86307560
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310015884.7A Active CN116127840B (en) | 2023-01-05 | 2023-01-05 | Data center load prediction method based on data driving |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116127840B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117370323A (en) * | 2023-10-09 | 2024-01-09 | 中国电信股份有限公司技术创新中心 | Complementary methods and related equipment for computer room electricity consumption |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108595301A (en) * | 2018-03-26 | 2018-09-28 | 中国科学院计算技术研究所 | A kind of server energy consumption prediction technique and system based on machine learning |
| CN109189190A (en) * | 2018-10-16 | 2019-01-11 | 西安交通大学 | A kind of data center's thermal management method based on temperature prediction |
| KR20190090356A (en) * | 2018-01-24 | 2019-08-01 | 고려대학교 산학협력단 | Machine learning based CPU temperature prediction method and apparatus |
| CN110619389A (en) * | 2019-09-23 | 2019-12-27 | 山东大学 | Load prediction method and system of combined cooling heating and power system based on LSTM-RNN |
| CN113887801A (en) * | 2021-09-29 | 2022-01-04 | 西安建筑科技大学 | Building cold load prediction method, system, equipment and readable storage medium |
| CN113962142A (en) * | 2021-09-26 | 2022-01-21 | 西安交通大学 | Data center temperature prediction method and system based on two-segment type LSTM |
| US11423327B2 (en) * | 2018-10-10 | 2022-08-23 | Oracle International Corporation | Out of band server utilization estimation and server workload characterization for datacenter resource optimization and forecasting |
| CN115099135A (en) * | 2022-06-16 | 2022-09-23 | 桂林理工大学 | Improved artificial neural network multi-type operation power consumption prediction method |
| CN115204360A (en) * | 2022-06-10 | 2022-10-18 | 南京有嘉科技有限公司 | Method for predicting power consumption of home user based on deep learning |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7502768B2 (en) * | 2004-02-27 | 2009-03-10 | Siemens Building Technologies, Inc. | System and method for predicting building thermal loads |
-
2023
- 2023-01-05 CN CN202310015884.7A patent/CN116127840B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20190090356A (en) * | 2018-01-24 | 2019-08-01 | 고려대학교 산학협력단 | Machine learning based CPU temperature prediction method and apparatus |
| CN108595301A (en) * | 2018-03-26 | 2018-09-28 | 中国科学院计算技术研究所 | A kind of server energy consumption prediction technique and system based on machine learning |
| US11423327B2 (en) * | 2018-10-10 | 2022-08-23 | Oracle International Corporation | Out of band server utilization estimation and server workload characterization for datacenter resource optimization and forecasting |
| CN109189190A (en) * | 2018-10-16 | 2019-01-11 | 西安交通大学 | A kind of data center's thermal management method based on temperature prediction |
| CN110619389A (en) * | 2019-09-23 | 2019-12-27 | 山东大学 | Load prediction method and system of combined cooling heating and power system based on LSTM-RNN |
| CN113962142A (en) * | 2021-09-26 | 2022-01-21 | 西安交通大学 | Data center temperature prediction method and system based on two-segment type LSTM |
| CN113887801A (en) * | 2021-09-29 | 2022-01-04 | 西安建筑科技大学 | Building cold load prediction method, system, equipment and readable storage medium |
| CN115204360A (en) * | 2022-06-10 | 2022-10-18 | 南京有嘉科技有限公司 | Method for predicting power consumption of home user based on deep learning |
| CN115099135A (en) * | 2022-06-16 | 2022-09-23 | 桂林理工大学 | Improved artificial neural network multi-type operation power consumption prediction method |
Non-Patent Citations (3)
| Title |
|---|
| Online Energy Estimation of Relational Operations in Database Systems;Zichen Xu 等;IEEE TRANSACTIONS ON COMPUTERS;第3223-3236页 * |
| 基于长短期记忆网络(LSTM)的数据中心温度预测算法;徐一轩;伍卫国;王思敏;胡壮;崔舜;;计算机技术与发展;第29卷(第12期);第1-7页 * |
| 数据中心电能管理及参与需求侧资源调度的展望;高赐威 等;电力系统自动化;第41卷(第23期);第1-7页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116127840A (en) | 2023-05-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113128113B (en) | Lean information building load prediction method based on deep learning and transfer learning | |
| CN112232575B (en) | A comprehensive energy system control method and device based on multi-element load forecasting | |
| Bi et al. | Accurate prediction of workloads and resources with multi-head attention and hybrid LSTM for cloud data centers | |
| CN110889603A (en) | An economic dispatch method for power system considering wind power correlation based on PCA-Copula theory | |
| Ye et al. | Optimizing AIGC services by prompt engineering and edge computing: A generative diffusion model-based contract theory approach | |
| CN108053077A (en) | A Short-Term Wind Speed Prediction Method and System Based on Interval Type 2 T-S Fuzzy Model | |
| CN117973644B (en) | A distributed photovoltaic power virtual collection method considering reference power station optimization | |
| Hu et al. | Optimal operation of power systems with energy storage under uncertainty: A scenario-based method with strategic sampling | |
| CN117035055A (en) | A device RUL prediction method and system based on BFL and semantics | |
| Abdelaziz et al. | Convolutional neural network with genetic algorithm for predicting energy consumption in public buildings | |
| CN116127840B (en) | Data center load prediction method based on data driving | |
| CN118211847A (en) | A short-term prediction method for carbon emissions from multi-energy loads | |
| CN117649070A (en) | A distribution network grid planning method considering investment and operation costs and photovoltaic consumption | |
| Yu | [Retracted] Research on Optimization Strategy of Task Scheduling Software Based on Genetic Algorithm in Cloud Computing Environment | |
| Yuan et al. | Personalized federated transfer learning based on global synthetic data for battery state of health estimation | |
| CN116933860A (en) | Transient stability evaluation model updating method and device, electronic equipment and storage medium | |
| CN119624153B (en) | A rapid assessment method for provincial power balance margin considering source-load uncertainty and market mechanism | |
| CN119009945B (en) | Comprehensive energy system load prediction method based on multi-energy demand response and ICPO-BiLSTM model | |
| Zheng et al. | Improved adaptive war strategy optimization algorithm assisted-adaptive multi-head graph attention mechanism network for remaining useful life of complex equipment | |
| CN116526582B (en) | Combined dispatching method and system for electric power unit based on artificial intelligence combined driving | |
| CN115660187B (en) | Low-carbon town ground source heat pump capacity optimization configuration method | |
| CN115528750B (en) | Power grid safety and stability oriented data model hybrid drive unit combination method | |
| CN113222216B (en) | Method, device and system for predicting cooling, heating and electricity loads | |
| Xie et al. | Vector ordinal optimization theory based large‐scale multi‐objective unit commitment considering stochastic wind power | |
| CN118569456B (en) | New energy and load forecasting method based on training simple cyclic unit network model |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |