CN111104298A - LSTM-based power grid server running state prediction device - Google Patents
LSTM-based power grid server running state prediction device Download PDFInfo
- Publication number
- CN111104298A CN111104298A CN201911159122.4A CN201911159122A CN111104298A CN 111104298 A CN111104298 A CN 111104298A CN 201911159122 A CN201911159122 A CN 201911159122A CN 111104298 A CN111104298 A CN 111104298A
- Authority
- CN
- China
- Prior art keywords
- server
- data
- lstm
- monitoring
- power grid
- 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
- 238000012544 monitoring process Methods 0.000 claims abstract description 43
- 238000012545 processing Methods 0.000 claims abstract description 14
- 238000013135 deep learning Methods 0.000 claims abstract description 10
- 238000007781 pre-processing Methods 0.000 claims abstract description 9
- 238000004458 analytical method Methods 0.000 claims abstract description 8
- 238000011156 evaluation Methods 0.000 claims abstract description 8
- 238000010606 normalization Methods 0.000 claims abstract description 6
- 238000012549 training Methods 0.000 claims abstract description 6
- 230000006870 function Effects 0.000 claims description 14
- 230000004913 activation Effects 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 4
- 238000003062 neural network model Methods 0.000 claims description 2
- 238000005457 optimization Methods 0.000 claims description 2
- 238000000034 method Methods 0.000 abstract description 5
- 230000008859 change Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 238000012300 Sequence Analysis Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
- G06F11/3447—Performance evaluation by modeling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
- G06F11/3466—Performance evaluation by tracing or monitoring
- G06F11/3476—Data logging
-
- 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
-
- 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
- G06N3/045—Combinations of networks
-
- 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
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Hardware Design (AREA)
- Quality & Reliability (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
A power grid server running state prediction device based on LSTM relates to the field of deep learning big data and aims to solve the problem that an existing real-time monitoring system cannot complete server evaluation and early warning. The server data monitoring module is used for acquiring server parameter indexes, extracting monitoring data and constructing a training set; the monitoring data refers to the utilization rate data of the CPU, the memory bank and the hard disk; the server data preprocessing module is used for carrying out normalization processing on the monitoring data acquired by the server data monitoring module and then carrying out feature selection; and the server running state prediction module is used for constructing an LSTM model to carry out big data deep learning processing analysis on the monitoring data after the characteristic selection so as to complete prediction evaluation and early warning of the running state of the power grid server. The method has the advantages that the problem that the prediction accuracy is reduced or prediction is difficult due to the fact that monitoring data information exceeds the computing capacity due to the fact that a single model is used in the existing method is solved.
Description
Technical Field
The invention relates to the field of deep learning big data.
Background
In the smart grid system, big data are generated in each link of the whole system; especially, the monitoring center can generate mass data information every day; the data volume maintained by data centers of many companies and enterprises in China is very large, professional operation and maintenance personnel are lacked, and non-professionals do not know and understand the operation and maintenance technology in place; many times, after some problems occur in the server, the server is usually not careful, and a long time is needed from the occurrence of the problems to the correct positioning of the problems.
Currently, in the related art, real-time monitoring cannot find problems in advance, and only can passively cope with the problems.
Disclosure of Invention
The invention aims to solve the problem that the existing real-time monitoring system cannot complete the evaluation and early warning of a server, and provides an LSTM-based power grid server operation state prediction device.
The LSTM-based power grid server running state prediction device comprises a server data monitoring module, a server data preprocessing module and a server running state prediction module;
the server data monitoring module is used for acquiring server parameter indexes, extracting monitoring data and constructing a training set;
the monitoring data refers to the utilization rate data of a CPU, a memory bank and a hard disk;
the server data preprocessing module is used for carrying out normalization processing on the monitoring data acquired by the server data monitoring module and then carrying out feature selection;
and the server running state prediction module is used for constructing an LSTM model to carry out big data deep learning processing analysis on the monitoring data after the characteristic selection so as to complete prediction evaluation and early warning of the running state of the power grid server.
The method has the advantages that the server parameter indexes are obtained through the server data monitoring module, the monitoring data are extracted, and a training set is constructed; after normalization processing is carried out by a server data preprocessing module, feature selection is carried out; finally, an LSTM model is built through a server running state prediction module to carry out big data deep learning processing analysis on the monitoring data after feature selection, and accurate prediction on the monitoring data of the power grid server is achieved; the method solves the problems that the prediction precision is reduced or difficult to predict because the monitoring data information exceeds the computing power due to the use of a single model in the prior art, and has the effects of dynamic and self-adaptive transformation; and the threshold can be flexibly selected, so that the calculation complexity is simplified, the implementation cost of a prediction algorithm is reduced, the manual investment is reduced, and the operation is more flexible and simpler to change.
Drawings
Fig. 1 is a block diagram of an LSTM-based power grid server operation state prediction apparatus according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of the structure of LSTM cells according to the second embodiment.
Detailed Description
The first embodiment is as follows: the present embodiment is described with reference to fig. 1, and the LSTM-based power grid server operation state prediction apparatus in the present embodiment includes a server data monitoring module 1, a server data preprocessing module 2, and a server operation state prediction module 3;
the server data monitoring module 1 is used for acquiring server parameter indexes, extracting monitoring data and constructing a training set;
the monitoring data refers to the utilization rate data of a CPU, a memory bank and a hard disk;
the server data preprocessing module 2 is used for performing normalization processing on the monitoring data acquired by the server data monitoring module 1 and then performing feature selection;
and the server running state prediction module 3 is used for constructing an LSTM model to perform big data deep learning processing analysis on the monitoring data after the characteristic selection so as to complete prediction evaluation and early warning of the running state of the power grid server.
In this embodiment, LSTM is a long-short term memory recurrent neural network; acquiring server parameter indexes through a server data monitoring module 1, extracting monitoring data and constructing a training set; after normalization processing is carried out by the server data preprocessing module 2, feature selection is carried out; finally, an LSTM model is built through a server running state prediction module 3 to carry out big data deep learning processing analysis on the monitoring data after feature selection, and accurate prediction on the monitoring data of the power grid server is achieved; (ii) a The problem that the prediction precision is reduced or is difficult to predict due to the fact that monitoring data information exceeds the computing capacity due to the fact that a single model is used in the existing method is solved, and the power grid server operation state prediction device has the effects of dynamic and self-adaptive transformation; and the threshold can be flexibly selected, so that the calculation complexity is simplified, the implementation cost of a prediction algorithm is reduced, the manual investment is reduced, and the operation is more flexible and simpler to change.
The second embodiment is as follows: the present embodiment is described with reference to fig. 2, and is further limited to the LSTM-based power grid server operation state prediction apparatus according to the first embodiment, in the present embodiment, the constructed LSTM model has an output dimension of 64, an input activation function of tan h, an output activation function of tan h, and tan h is a hyperbolic tangent function; the gate activation function is sigma, and sigma is a sigmoid function; the LSTM model structure is a neural network model with 1 layer of LSTM +1 layer of DENSE.
In this embodiment, the output dimension of the LSTM layer, i.e., the number of cells included in the LSTM layer, increases the number of cells, and increases the internal parameters.
In the embodiment, the DENSE layer is a hidden layer, the inherent rule in the time sequence is deeply excavated through the unique unit structure of the LSTM, the LSTM layer and the DENSE layer are combined to realize the running state prediction of the CPU data of the power grid server, and finally, the time sequence analysis prediction of the power grid server is realized.
The third concrete implementation mode: in the second embodiment, the LSTM model performs a big data deep learning processing analysis on the monitoring data after feature selection by using a BPTT algorithm;
the BPTT algorithm is realized by the following steps;
step one, calculating an output value of the LSTM cell according to a forward calculation method formula; wherein, the forward calculation method has the formula from formula (1) to formula (5);
it=σ(Wxixt+Whiht-1+Wcict-1+bi) (1)
ft=σ(Wxfxt+Whfht-1+Wcfct-1+bf) (2)
ct=ftct-1+ittanh(Wxcxt+Whcht-1+bc) (3)
Ot=σ(Wxoxt+Whoht-1+Wcoct+bo) (4)
ht=ottanh(ct) (5)
in the formula: i is an input gate, f is a forgetting gate, c is a cell state, and O is an output gate; w is a corresponding weight coefficient matrix; such as WxfRepresenting input layers toA weight coefficient matrix of the forgetting gate; b is the corresponding bias term; sigma is sigmoid function; tan h is a hyperbolic tangent activation function; x is a given sequence of an input layer, namely the utilization rate of a CPU, a memory bank and a hard disk under a normal condition; h is a hidden layer sequence, namely a prediction evaluation and early warning result of the operation state of the power grid server; t is the current moment, and t-1 is the last moment;
the relationship between x and h is as follows: h ist=tanh(Wxhxt+Whhht-1+bh);
Step two, reversely calculating an error term of each LSTM cell, wherein the error term comprises 2 reverse propagation directions according to time and network levels;
step three, calculating the gradient of each weight according to the corresponding error term;
and step four, updating the weight by applying an optimization algorithm based on gradient.
Claims (3)
1. The LSTM-based power grid server operation state prediction device is characterized by comprising a server data monitoring module (1), a server data preprocessing module (2) and a server operation state prediction module (3);
the server data monitoring module (1) is used for acquiring server parameter indexes, extracting monitoring data and constructing a training set;
the monitoring data refers to the utilization rate data of a CPU, a memory bank and a hard disk;
the server data preprocessing module (2) is used for carrying out normalization processing on the monitoring data acquired by the server data monitoring module (1) and then carrying out feature selection;
and the server running state prediction module (3) is used for constructing an LSTM model to carry out big data deep learning processing analysis on the monitoring data after the characteristic selection so as to complete prediction evaluation and early warning of the running state of the power grid server.
2. The LSTM-based power grid server operation state prediction device of claim 1, wherein the constructed LSTM model has an output dimension of 64, an input activation function of tanh, an output activation function of tanh, and a hyperbolic tangent function of tanh; the gate activation function is sigma, and sigma is a sigmoid function; the LSTM model structure is a neural network model with 1 layer of LSTM +1 layer of DENSE.
3. The LSTM-based power grid server operation state prediction device according to claim 2, wherein the LSTM model performs big data deep learning processing analysis on the monitoring data after feature selection by adopting a BPTT algorithm;
the BPTT algorithm is realized by the following steps;
step one, calculating an output value of the LSTM cell according to a forward calculation method formula; wherein, the forward calculation method has the formula from formula (1) to formula (5);
it=σ(Wxixt+Whiht-1+Wcict-1+bi) (1)
ft=σ(Wxfxt+Whfht-1+Wcfct-1+bf) (2)
ct=ftct-1+ittanh(Wxcxt+Whcht-1+bc) (3)
Ot=σ(Wxoxt+Whoht-1+Wcoct+bo) (4)
ht=ottanh(ct) (5)
in the formula: i is an input gate, f is a forgetting gate, c is a cell state, and O is an output gate; w is a corresponding weight coefficient matrix; such as WxfA weight coefficient matrix representing the input layer to the forgetting gate; b is the corresponding bias term; sigma is sigmoid function; tan h is a hyperbolic tangent activation function; x is a given sequence of an input layer, namely the utilization rate of a CPU, a memory bank and a hard disk under a normal condition; h is a hidden layer sequence, namely a prediction evaluation and early warning result of the operation state of the power grid server; t is the current moment, and t-1 is the last moment;
the relationship between x and h is as follows:ht=tanh(Wxhxt+Whhht-1+bh);
step two, reversely calculating an error term of each LSTM cell, wherein the error term comprises 2 reverse propagation directions according to time and network levels;
step three, calculating the gradient of each weight according to the corresponding error term;
and step four, updating the weight by applying an optimization algorithm based on gradient.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911159122.4A CN111104298A (en) | 2019-11-22 | 2019-11-22 | LSTM-based power grid server running state prediction device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911159122.4A CN111104298A (en) | 2019-11-22 | 2019-11-22 | LSTM-based power grid server running state prediction device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111104298A true CN111104298A (en) | 2020-05-05 |
Family
ID=70421359
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911159122.4A Pending CN111104298A (en) | 2019-11-22 | 2019-11-22 | LSTM-based power grid server running state prediction device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111104298A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112101519A (en) * | 2020-08-07 | 2020-12-18 | 中国南方电网有限责任公司超高压输电公司广州局 | Converter valve cooling margin prediction operation and maintenance method and system based on LSTM |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106909989A (en) * | 2017-02-24 | 2017-06-30 | 国网河南省电力公司电力科学研究院 | A kind of grid disturbance Forecasting Methodology and device |
CN108416690A (en) * | 2018-01-19 | 2018-08-17 | 中国矿业大学 | Load Forecasting based on depth LSTM neural networks |
CN110135637A (en) * | 2019-05-13 | 2019-08-16 | 武汉科技大学 | Micro-capacitance sensor short-term load forecasting method based on shot and long term memory and adaptive boosting |
CN110222901A (en) * | 2019-06-13 | 2019-09-10 | 河海大学常州校区 | A kind of electric load prediction technique of the Bi-LSTM based on deep learning |
CN110334879A (en) * | 2019-07-11 | 2019-10-15 | 华北电力大学 | Power grid bus reactive load forecasting method and device |
CN110427261A (en) * | 2019-08-12 | 2019-11-08 | 电子科技大学 | A kind of edge calculations method for allocating tasks based on the search of depth Monte Carlo tree |
-
2019
- 2019-11-22 CN CN201911159122.4A patent/CN111104298A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106909989A (en) * | 2017-02-24 | 2017-06-30 | 国网河南省电力公司电力科学研究院 | A kind of grid disturbance Forecasting Methodology and device |
CN108416690A (en) * | 2018-01-19 | 2018-08-17 | 中国矿业大学 | Load Forecasting based on depth LSTM neural networks |
CN110135637A (en) * | 2019-05-13 | 2019-08-16 | 武汉科技大学 | Micro-capacitance sensor short-term load forecasting method based on shot and long term memory and adaptive boosting |
CN110222901A (en) * | 2019-06-13 | 2019-09-10 | 河海大学常州校区 | A kind of electric load prediction technique of the Bi-LSTM based on deep learning |
CN110334879A (en) * | 2019-07-11 | 2019-10-15 | 华北电力大学 | Power grid bus reactive load forecasting method and device |
CN110427261A (en) * | 2019-08-12 | 2019-11-08 | 电子科技大学 | A kind of edge calculations method for allocating tasks based on the search of depth Monte Carlo tree |
Non-Patent Citations (2)
Title |
---|
代杰杰 等: "采用LSTM网络的电力变压器运行状态预测方法研究", 《高电压技术》 * |
温惠英 等: "GA-LSTM 模型在高速公路交通流预测中的应用", 《哈尔滨工业大学学报》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112101519A (en) * | 2020-08-07 | 2020-12-18 | 中国南方电网有限责任公司超高压输电公司广州局 | Converter valve cooling margin prediction operation and maintenance method and system based on LSTM |
CN112101519B (en) * | 2020-08-07 | 2022-04-08 | 中国南方电网有限责任公司超高压输电公司广州局 | Converter valve cooling margin prediction operation and maintenance method and system based on LSTM |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109492822B (en) | Air pollutant concentration time-space domain correlation prediction method | |
CN111458748B (en) | Performance earthquake motion risk analysis method based on three-layer data set neural network | |
CN106951695A (en) | Plant equipment remaining life computational methods and system under multi-state | |
CN112001110B (en) | Structural damage identification monitoring method based on vibration signal space real-time recurrent graph convolutional neural network | |
CN109920248B (en) | Bus arrival time prediction method based on GRU neural network | |
CN113391211B (en) | Method for predicting remaining life of lithium battery under small sample condition | |
CN115438726A (en) | Device life and fault type prediction method and system based on digital twin technology | |
CN115564114A (en) | Short-term prediction method and system for airspace carbon emission based on graph neural network | |
Wang et al. | A remaining useful life prediction model based on hybrid long-short sequences for engines | |
Zeng et al. | Short-term load forecasting of smart grid systems by combination of general regression neural network and least squares-support vector machine algorithm optimized by harmony search algorithm method | |
CN115952724A (en) | Method, system, equipment and medium for predicting residual life of aircraft engine | |
CN115080795A (en) | Multi-charging-station cooperative load prediction method and device | |
CN111104298A (en) | LSTM-based power grid server running state prediction device | |
Tyagi et al. | Analysis of infectious disease transmission and prediction through SEIQR epidemic model | |
CN113688770A (en) | Long-term wind pressure missing data completion method and device for high-rise building | |
CN105808962A (en) | Assessment method considering voltage probabilities of multiple electric power systems with wind power output randomness | |
Zong et al. | Embedded software fault prediction based on back propagation neural network | |
CN117371303A (en) | Prediction method for effective wave height under sea wave | |
CN117154680A (en) | Wind power prediction method based on non-stationary transducer model | |
CN115965160A (en) | Data center energy consumption prediction method and device, storage medium and electronic equipment | |
CN115034432A (en) | Wind speed prediction method for wind generating set of wind power plant | |
CN116226654A (en) | Machine learning data forgetting method based on mask gradient | |
CN112183847B (en) | QRNN improved Stacking algorithm-based train running wind speed probability prediction method | |
CN114841063A (en) | Aero-engine residual life prediction method based on deep learning | |
CN111476408B (en) | Power communication equipment state prediction method and system |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20200505 |