CN110334726A - A kind of identification of the electric load abnormal data based on Density Clustering and LSTM and restorative procedure - Google Patents
A kind of identification of the electric load abnormal data based on Density Clustering and LSTM and restorative procedure Download PDFInfo
- Publication number
- CN110334726A CN110334726A CN201910331551.9A CN201910331551A CN110334726A CN 110334726 A CN110334726 A CN 110334726A CN 201910331551 A CN201910331551 A CN 201910331551A CN 110334726 A CN110334726 A CN 110334726A
- Authority
- CN
- China
- Prior art keywords
- data
- lstm
- value
- abnormal data
- density clustering
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- 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
-
- 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
- G06N3/084—Backpropagation, e.g. using gradient descent
Abstract
A method of the identification and reparation of the electric load abnormal data based on Density Clustering and LSTM belong to power quality analysis method and technology field.This method is identified and is repaired to abnormal data in such a way that density-based algorithms (Density-Based Spatial Clustering of Applications with Noise) and shot and long term Memory Neural Networks (Long Short-Term Memory) combine.First with DSCAN algorithm logarithm, day is that unit carries out Density Clustering to this method accordingly, obtains abnormal data；Followed by shot and long term Memory Neural Networks LSTM, it will be determined as that abnormal time series data as its input, predicts next sequence data using preceding n sequence data；Finally, the threshold value to float up and down is arranged using the predicted value of LSTM as exact value, if measured value exceeds threshold range, it is considered as exceptional value, and using the predicted value of LSTM as correction value.This method has fully considered the timing and regularity of electric energy quality monitoring system data in actual electric network, can precisely detect specific abnormal numerical value and repair, have good practical application value.
Description
Technical field
A method of the identification and reparation of the electric load abnormal data based on Density Clustering and LSTM belong to electric energy matter
Analysis method technical field.
Background technique
With Chinese society rapid economic development, Living consumption is significantly improved, the load data amount in smart grid
It is increasing.However, by the hardware devices such as transformer, smart grid test environment, parameter configuration and manual record fault etc. because
Element influences, and always has some abnormal datas and occurs.These abnormal datas interfere data analysis, in some instances it may even be possible to bring the analysis of mistake
As a result.For the accuracy for ensuring post analysis, before Develop Data analysis, need to carry out initial data disorder data recognition and
It repairs.
The side that this method uses density-based algorithms and shot and long term Memory Neural Networks (LSTM) to combine for the first time
Formula identifies abnormal data, this method fully considered in actual electric network the timing of electric energy quality monitoring system data and
Regularity can precisely detect specific abnormal numerical value, while avoiding the extraction of conventional method manual features and being easy to happen information
The problem of loss, has good recognition effect and practical application value.
Summary of the invention
The algorithm first automatically analyzes analysis and the area based process Hou Tai load data using DBSCAN algorithm, to one
Annual data carries out Density Clustering as unit of day, obtains outlier therein, i.e. abnormal data；Then, it is cut using normal data
It is disconnected that preceding 8 sequence datas is taken to predict next sequence data to train LSTM neural network；It will be determined as followed by LSTM different
Normal time series data takes preceding 8 sequence datas to predict next sequence data as input, truncation；Finally, during processing
Using the predicted value of LSTM as exact value, the threshold value to float up and down is set, and sentences to the corresponding measured value of its sequence data
It is disconnected, if it exceeds threshold range, it is considered as exceptional value, and using the predicted value of LSTM as correction value, continuation is predicted backward, until
Sequence data end of run.
Compared with prior art, the method for the present invention has the advantage that
1) present invention is using depth learning technology, it is possible to prevente effectively from traditional abnormal data identification is inefficient and accurate
Rate is low；
2) anomalous identification density clustering algorithm can be according to cluster carrying out sample to load data using density clustering algorithm
The Density Distribution difference automatic cluster of object in space is at different clusters, and as a result there is no bias and can be to arbitrary shape
Dense data set clustered:
3) LSTM is suitble to study to have forward-backward correlation, successional time series data, can be responsible for by LSTM to load number
It according to progress sample interior anomalous identification and is modified, accuracy rate greatly improves.
Detailed description of the invention
Fig. 1 is the method model figure of the identification and reparation of Density Clustering and the electric load abnormal data of LSTM.
Specific embodiment
Embodiment:
DBSCAN clustering algorithm process
The design that density clustering algorithm of the present invention is realized is as follows:
(1) k-dist for calculating each point shows the variation tendency of k-dist with scatter plot in Excel, and determines half
The value of diameter Eps；
(2) initialization data indicates that the point is not visited to the attribute of all data points setting unvisited；
(3) it is concentrated in all properties labeled as the point of unvisited, looks for a point p at random, and be marked as
Visited checks whether the point is kernel object.If it is not, then marking p is noise spot, and from labeled as unvisited attribute
Point concentration search out next point, until finding out core point；If so, the step of executing below.
(4) class (being denoted as C) is created, a Candidate Set Candidates is established.When initial, in Candidates only
One element, that is, the kernel object that previous step is found；
(5) for each of Candidates object (being denoted as q), following operation is done:
1. if class C is added q is not gathered any one class also；
2. if q is kernel object, by its ∈-neighborhood, except when the point except the point of preceding class C is added
Candidates；
3. if q is removed from noise set (because of certainly not noise) in noise set, by it；
4. if being labeled as visited the attribute of q is unvisited；
(6) circulation executes (5) step, until can not find kernel object in Candidates, this explanation it is current this
Class is looked for completely.
(7) (3) step is returned, starts to look for next cluster.Until all points all become visited.
Above-mentioned density clustering algorithm is carried out to load data, inputs the data as unit of day, Density Clustering can export with
Cluster is the data of unit, finds the data that only one in cluster is put, i.e. the noise spot of Density Clustering is exactly load abnormal data.
LSTM neural network
Neuron used in LSTM neural network model is more complex compared with for RNN, it contains 3 sigmoid
Activation primitive and 2 tanh activation primitive modules, and tradition RNN only includes a tanh activation primitive module.LSTM model relates to
And formula it is as follows:
ft=σ (ωf·[ht-1,xt]+bt) (1)
H in formula (1)t-1For the output at t-1 moment, xtFor the input of t moment, ωfT moment is reached with for the t-1 moment
The weight and biasing that door is forgotten corresponding to neuron, obtain Forgetting coefficient f finally by sigmoid functiont。
it=σ (ωi·[ht-1,xt]+bi) (2)
H in formula (2)t-1For the output at t-1 moment, xtFor the input of t moment, ωiAnd biT moment is reached for the t-1 moment
Neuron corresponding to input gate weight and biasing, obtain input coefficient i finally by sigmoid functiont。
H in formula (3)t-1For the output at t-1 moment, xtFor the input of t moment, ωcAnd bcT moment is reached for the t-1 moment
Neuron corresponding to input data weight and biasing, obtain input data finally by tanh function
C in formula (4)tFor the updated cell state of t moment, value is equal to what previous moment was obtained by forgetting algorithm
The data f retained in cell statet·Ct-1In addition the input data that t moment input gate determines
ot=σ (ωo·[ht-1,xt]+bo) (5)
It is the output at t-1 moment, x in formula (5)tFor the input of t moment, ωoAnd boThe mind of t moment is reached for the t-1 moment
Weight and biasing through out gate corresponding to member obtain output factor o finally by sigmoid functiont。
ht=ot·tanh(Ct) (6)
H in formula (6)tFor the output of t moment, CtFor the updated cell state of t moment, otGo out for what out gate calculated
Output factor, pass through tanh (Ct), otValue can be obtained t moment output data ht。
Before carrying out LSTM training, the training parameter of LSTM neural network is determined.The data chosen herein are DBSCAN
The outlier that module identifies, i.e. abnormal data constantly choose preceding 8 data to 96 point datas that each abnormal data includes
Input of the point as LSTM network, that is, the length of time series that input is arranged is 8, and input dimension is set as 1, output dimension setting
It is 1, the unit number in hidden layer is set as 10.The training process of LSTM neural network is such as RNN and backpropagation
Algorithm mainly has following three steps:
1) output valve of each neuron of forward calculation, for LSTM, i.e. ft、it、Ct、ot、htThe value of five vectors.
Shown in calculation method such as above formula (1)-(6).
2) value of the error term of each neuron of retrospectively calculate.As RNN, the backpropagation of LSTM error term is also packet
Include both direction: one is that backpropagation along the time calculates the error term at each moment that is, since current t moment；One
It is to propagate error term upper layer.The backpropagation formula of model is by taking out gate weight, biasing as an example, the formula of other coefficients
And so on:
3) according to corresponding error term, the gradient of each weight is calculated.δotIt is the correspondence error of propagated forward out gate
, e is by element multiplication, and weight, the more new formula of biasing are as follows:
Wherein η is learning rate, and tanh is hyperbolic tangent function.L is loss function, when the loss function of network structure is received
When holding back smaller range, just obtain with normal load data training LSTM neural network.Abnormal time series data is inputted LSTM
Neural network predicts next sequence data using preceding 8 sequence datas, the data and actual measurement number that LSTM neural network prediction goes out
According to being compared, a threshold value to float up and down is set, if predicted value and the difference of measured value exceed threshold range, depending on actual measurement
Data are exceptional value, and using the predicted value of LSTM neural network as correction value, continuation is predicted forward, until sequence data is run
Terminate.
The present invention solves abnormal number using in conjunction with the approach of density clustering algorithm and shot and long term Memory Neural Networks to find
The moving law for meeting power grid according to problem is a kind of effectively detection method.The present invention is suitably applied in the inspection of abnormal data
It surveys with reparation, there is good identification value and application effect.
Claims (5)
1. a kind of identification based on the electric load abnormal data based on Density Clustering and LSTM and the method repaired.Its feature exists
In the specific steps of this method are as follows:
Step 1: input data is normalized；
Step 2: dividing the label of normal data and abnormal data using DSCAN algorithm automatically, i.e. the data progress to 1 year is close
Degree cluster (as unit of day), obtains outlier, i.e., abnormal data point (includes one day all Temporal Sampling point)；
Step 3: constructing shot and long term Memory Neural Networks LSTM, and the abnormal data exported in step 2 is defeated as the timing of LSTM
Enter, it is 1 that optimal output neuron number, which is arranged, i.e., constantly inputs LSTM using preceding 8 sequence datas, predict next sequence number
According to；
Step 4: using the predicted value of each time phase of LSTM as exact value, the threshold value to float up and down is set, to the sequence number
The corresponding measured value in strong point judged, if it exceeds threshold range, is considered as exceptional value, and using the predicted value of LSTM as repairing
Positive value, continuation are predicted forward, until one day all sequences data run terminates.
2. a kind of identification based on based on Density Clustering and the electric load abnormal data of LSTM according to claim 1 with
The method of reparation, which is characterized in that data prediction described in step 1, the voltage dip original waveform data that will acquire, which is done, returns
One change processing, method for normalizing used are that min-max standardizes (Min-max normalization)/0-1 standardization (0-1
Normalization it) also makes deviation standardize, is the linear transformation to initial data, result is made to fall on [0,1] section, convert
Function is as follows:
Wherein max is the maximum value of sample data, and min is the minimum value of sample data.
3. a kind of identification based on based on Density Clustering and the electric load abnormal data of LSTM according to claim 1 with
The method of reparation, which is characterized in that step 2 carries out Density Clustering (as unit of day) to 1 year data, obtains outlier i.e.
Abnormal data.
4. a kind of identification based on based on Density Clustering and the electric load abnormal data of LSTM according to claim 1 with
The method of reparation, which is characterized in that step 3, can be according to preceding 8 point predictions using normal data training LSTM network
Next point out.
5. a kind of identification based on based on Density Clustering and the electric load abnormal data of LSTM according to claim 1 with
The method of reparation, which is characterized in that step 4 goes out electric load abnormal data and repair using LSTM neural network recognization.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910331551.9A CN110334726A (en) | 2019-04-24 | 2019-04-24 | A kind of identification of the electric load abnormal data based on Density Clustering and LSTM and restorative procedure |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910331551.9A CN110334726A (en) | 2019-04-24 | 2019-04-24 | A kind of identification of the electric load abnormal data based on Density Clustering and LSTM and restorative procedure |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110334726A true CN110334726A (en) | 2019-10-15 |
Family
ID=68139782
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910331551.9A Pending CN110334726A (en) | 2019-04-24 | 2019-04-24 | A kind of identification of the electric load abnormal data based on Density Clustering and LSTM and restorative procedure |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110334726A (en) |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111061714A (en) * | 2019-12-12 | 2020-04-24 | 清华大学 | Timestamp repairing method and device |
| CN111160603A (en) * | 2019-11-21 | 2020-05-15 | 中国电力科学研究院有限公司 | Method and system for guaranteeing reliability of end-to-end power communication service |
| CN111260198A (en) * | 2020-01-10 | 2020-06-09 | 广东电网有限责任公司 | Method and system for judging degree of rationality of line loss in transformer area synchronization and terminal equipment |
| CN111461400A (en) * | 2020-02-28 | 2020-07-28 | 国网浙江省电力有限公司 | Load data completion method based on Kmeans and T-L STM |
| CN111582542A (en) * | 2020-03-31 | 2020-08-25 | 国网上海市电力公司 | Power load prediction method and system based on abnormal restoration |
| CN111667135A (en) * | 2020-03-25 | 2020-09-15 | 国网天津市电力公司 | Load structure analysis method based on typical feature extraction |
| CN111738364A (en) * | 2020-08-05 | 2020-10-02 | 国网江西省电力有限公司供电服务管理中心 | Electricity stealing detection method based on combination of user load and electricity consumption parameter |
| CN112235043A (en) * | 2020-09-14 | 2021-01-15 | 上海大学 | Distributed optical fiber abnormal data restoration model based on self-adaptive long-term and short-term memory |
| CN112506899A (en) * | 2020-11-25 | 2021-03-16 | 东华理工大学 | PM2.5 data abnormal value detection method based on improved LSTM |
| CN112712189A (en) * | 2019-10-25 | 2021-04-27 | 北京市热力集团有限责任公司 | Heat supply demand load prediction method |
| CN112733417A (en) * | 2020-11-16 | 2021-04-30 | 南京邮电大学 | Abnormal load data detection and correction method and system based on model optimization |
| CN113010504A (en) * | 2021-03-04 | 2021-06-22 | 江苏方天电力技术有限公司 | Electric power data anomaly detection method and system based on LSTM and improved K-means algorithm |
| CN113205134A (en) * | 2021-04-30 | 2021-08-03 | 中国烟草总公司郑州烟草研究院 | Network security situation prediction method and system |
| CN113485986A (en) * | 2021-06-25 | 2021-10-08 | 国网江苏省电力有限公司信息通信分公司 | Electric power data restoration method |
| CN113554229A (en) * | 2021-07-23 | 2021-10-26 | 国网青海省电力公司信息通信公司 | Three-phase voltage unbalance abnormality detection method and device |
| CN113554079A (en) * | 2021-07-14 | 2021-10-26 | 中国地质大学（北京） | Electric power load abnormal data detection method and system based on secondary detection method |
| CN113780691A (en) * | 2020-06-09 | 2021-12-10 | 富泰华工业（深圳）有限公司 | Data testing method and device, electronic equipment and storage medium |
| CN115834424A (en) * | 2022-10-09 | 2023-03-21 | 国网甘肃省电力公司临夏供电公司 | Method for identifying and correcting abnormal data of line loss of power distribution network |
| CN115935284A (en) * | 2022-11-04 | 2023-04-07 | 贵州电网有限责任公司信息中心 | Power grid abnormal voltage detection method, device, equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104182803A (en) * | 2014-08-13 | 2014-12-03 | 东北电力大学 | Wind power data preprocessing method, wind power forecast method and system |
| CN107423769A (en) * | 2017-08-03 | 2017-12-01 | 四川大学 | Electric load curve adaptive clustering scheme based on morphological feature |
| US20180006900A1 (en) * | 2016-06-29 | 2018-01-04 | Microsoft Technology Licensing, Llc | Predictive anomaly detection in communication systems |
| CN108436050A (en) * | 2018-04-16 | 2018-08-24 | 大连理工大学 | A method of continuous cast mold bleed-out is forecast using space density clustering DBSCAN |
| CN109359786A (en) * | 2018-12-05 | 2019-02-19 | 国网江苏省电力有限公司扬州供电分公司 | A kind of power station area short-term load forecasting method |
-
2019
- 2019-04-24 CN CN201910331551.9A patent/CN110334726A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104182803A (en) * | 2014-08-13 | 2014-12-03 | 东北电力大学 | Wind power data preprocessing method, wind power forecast method and system |
| US20180006900A1 (en) * | 2016-06-29 | 2018-01-04 | Microsoft Technology Licensing, Llc | Predictive anomaly detection in communication systems |
| CN107423769A (en) * | 2017-08-03 | 2017-12-01 | 四川大学 | Electric load curve adaptive clustering scheme based on morphological feature |
| CN108436050A (en) * | 2018-04-16 | 2018-08-24 | 大连理工大学 | A method of continuous cast mold bleed-out is forecast using space density clustering DBSCAN |
| CN109359786A (en) * | 2018-12-05 | 2019-02-19 | 国网江苏省电力有限公司扬州供电分公司 | A kind of power station area short-term load forecasting method |
Non-Patent Citations (2)
| Title |
|---|
| YIFAN ZHOU 等: "A semi-supervised anomaly detection method for wind farm power data preprocessing", 《2017 IEEE POWER & ENERGY SOCIETY GENERAL MEETING》 * |
| 田力 等: "基于密度聚类技术的电力系统用电量异常分析算法", 《电力系统自动化》 * |
Cited By (27)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112712189A (en) * | 2019-10-25 | 2021-04-27 | 北京市热力集团有限责任公司 | Heat supply demand load prediction method |
| CN111160603A (en) * | 2019-11-21 | 2020-05-15 | 中国电力科学研究院有限公司 | Method and system for guaranteeing reliability of end-to-end power communication service |
| CN111061714A (en) * | 2019-12-12 | 2020-04-24 | 清华大学 | Timestamp repairing method and device |
| CN111260198A (en) * | 2020-01-10 | 2020-06-09 | 广东电网有限责任公司 | Method and system for judging degree of rationality of line loss in transformer area synchronization and terminal equipment |
| CN111461400A (en) * | 2020-02-28 | 2020-07-28 | 国网浙江省电力有限公司 | Load data completion method based on Kmeans and T-L STM |
| CN111461400B (en) * | 2020-02-28 | 2023-06-23 | 国网浙江省电力有限公司 | Kmeans and T-LSTM-based load data completion method |
| CN111667135A (en) * | 2020-03-25 | 2020-09-15 | 国网天津市电力公司 | Load structure analysis method based on typical feature extraction |
| CN111667135B (en) * | 2020-03-25 | 2023-07-28 | 国网天津市电力公司 | Load structure analysis method based on typical feature extraction |
| CN111582542B (en) * | 2020-03-31 | 2023-10-03 | 国网上海市电力公司 | Power load prediction method and system based on anomaly repair |
| CN111582542A (en) * | 2020-03-31 | 2020-08-25 | 国网上海市电力公司 | Power load prediction method and system based on abnormal restoration |
| CN113780691B (en) * | 2020-06-09 | 2023-09-29 | 富泰华工业（深圳）有限公司 | Data testing method and device, electronic equipment and storage medium |
| CN113780691A (en) * | 2020-06-09 | 2021-12-10 | 富泰华工业（深圳）有限公司 | Data testing method and device, electronic equipment and storage medium |
| CN111738364A (en) * | 2020-08-05 | 2020-10-02 | 国网江西省电力有限公司供电服务管理中心 | Electricity stealing detection method based on combination of user load and electricity consumption parameter |
| CN112235043A (en) * | 2020-09-14 | 2021-01-15 | 上海大学 | Distributed optical fiber abnormal data restoration model based on self-adaptive long-term and short-term memory |
| CN112235043B (en) * | 2020-09-14 | 2022-12-23 | 上海大学 | Distributed optical fiber abnormal data restoration device based on self-adaptive long-term and short-term memory |
| CN112733417B (en) * | 2020-11-16 | 2022-11-01 | 南京邮电大学 | Abnormal load data detection and correction method and system based on model optimization |
| CN112733417A (en) * | 2020-11-16 | 2021-04-30 | 南京邮电大学 | Abnormal load data detection and correction method and system based on model optimization |
| CN112506899A (en) * | 2020-11-25 | 2021-03-16 | 东华理工大学 | PM2.5 data abnormal value detection method based on improved LSTM |
| CN113010504B (en) * | 2021-03-04 | 2022-06-10 | 江苏方天电力技术有限公司 | Electric power data anomaly detection method and system based on LSTM and improved K-means algorithm |
| CN113010504A (en) * | 2021-03-04 | 2021-06-22 | 江苏方天电力技术有限公司 | Electric power data anomaly detection method and system based on LSTM and improved K-means algorithm |
| CN113205134A (en) * | 2021-04-30 | 2021-08-03 | 中国烟草总公司郑州烟草研究院 | Network security situation prediction method and system |
| CN113485986A (en) * | 2021-06-25 | 2021-10-08 | 国网江苏省电力有限公司信息通信分公司 | Electric power data restoration method |
| CN113554079A (en) * | 2021-07-14 | 2021-10-26 | 中国地质大学（北京） | Electric power load abnormal data detection method and system based on secondary detection method |
| CN113554229A (en) * | 2021-07-23 | 2021-10-26 | 国网青海省电力公司信息通信公司 | Three-phase voltage unbalance abnormality detection method and device |
| CN115834424A (en) * | 2022-10-09 | 2023-03-21 | 国网甘肃省电力公司临夏供电公司 | Method for identifying and correcting abnormal data of line loss of power distribution network |
| CN115834424B (en) * | 2022-10-09 | 2023-11-21 | 国网甘肃省电力公司临夏供电公司 | Identification and correction method for abnormal data of power distribution network line loss |
| CN115935284A (en) * | 2022-11-04 | 2023-04-07 | 贵州电网有限责任公司信息中心 | Power grid abnormal voltage detection method, device, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110334726A (en) | A kind of identification of the electric load abnormal data based on Density Clustering and LSTM and restorative procedure | |
| Chen et al. | Combining empirical mode decomposition and deep recurrent neural networks for predictive maintenance of lithium-ion battery | |
| Qu et al. | A neural-network-based method for RUL prediction and SOH monitoring of lithium-ion battery | |
| Khan et al. | Batteries state of health estimation via efficient neural networks with multiple channel charging profiles | |
| CN112036084B (en) | Similar product life migration screening method and system | |
| CN109116834B (en) | Intermittent process fault detection method based on deep learning | |
| Fu et al. | Correlation analysis and augmentation of samples for a bidirectional gate recurrent unit network for the remaining useful life prediction of bearings | |
| CN112289391B (en) | Anode aluminum foil performance prediction system based on machine learning | |
| Liu et al. | Lithium-ion battery remaining useful life prediction with long short-term memory recurrent neural network | |
| CN113780420B (en) | GRU-GCN-based method for predicting concentration of dissolved gas in transformer oil | |
| CN114861879A (en) | Modeling method for optimizing thermal error of electric spindle of Elman neural network based on longicorn whisker algorithm | |
| CN114897103A (en) | Industrial process fault diagnosis method based on neighbor component loss optimization multi-scale convolutional neural network | |
| Zhang et al. | Fault diagnosis of oil-immersed power transformer based on difference-mutation brain storm optimized catboost model | |
| Dorado-Rojas et al. | Time series-based small-signal stability assessment using deep learning | |
| CN113033898A (en) | Electrical load prediction method and system based on K-means clustering and BI-LSTM neural network | |
| Mazzi et al. | Lithium-ion battery state of health estimation using a hybrid model based on a convolutional neural network and bidirectional gated recurrent unit | |
| CN113536662A (en) | Electronic transformer error state prediction method based on firefly optimized LightGBM algorithm | |
| CN113126489A (en) | CNN-GRU-BINN-based intelligent BIT design method for heavy-duty gas turbine control system | |
| CN112766548A (en) | Order completion time prediction method based on GASA-BP neural network | |
| Zhao et al. | The grey theory and the preliminary probe into information acquisition technology | |
| CN113486926B (en) | Automatic change pier equipment anomaly detection system | |
| CN115577290A (en) | Distribution network fault classification and source positioning method based on deep learning | |
| CN115221520A (en) | Open set identification-based unknown attack detection method for industrial control network | |
| CN115423091A (en) | Conditional antagonistic neural network training method, scene generation method and system | |
| Tran-Quang et al. | Aquaculture environment prediction based on improved lstm deep learning 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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20191015 |
|
| WD01 | Invention patent application deemed withdrawn after publication |