CN110689324A - Auxiliary judging method for detection result of distribution transformer - Google Patents
Auxiliary judging method for detection result of distribution transformer Download PDFInfo
- Publication number
- CN110689324A CN110689324A CN201910933689.6A CN201910933689A CN110689324A CN 110689324 A CN110689324 A CN 110689324A CN 201910933689 A CN201910933689 A CN 201910933689A CN 110689324 A CN110689324 A CN 110689324A
- Authority
- CN
- China
- Prior art keywords
- data
- detection
- transformer
- sample
- distribution transformer
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 71
- 238000000034 method Methods 0.000 title claims abstract description 18
- 230000002159 abnormal effect Effects 0.000 claims abstract description 30
- 238000012360 testing method Methods 0.000 claims abstract description 24
- 238000012549 training Methods 0.000 claims abstract description 22
- 238000011156 evaluation Methods 0.000 claims abstract description 9
- 238000004140 cleaning Methods 0.000 claims abstract description 8
- 238000010606 normalization Methods 0.000 claims abstract description 8
- 238000012545 processing Methods 0.000 claims abstract description 8
- 238000012706 support-vector machine Methods 0.000 claims abstract description 8
- 238000005070 sampling Methods 0.000 claims abstract description 7
- 230000006870 function Effects 0.000 claims description 22
- 238000011161 development Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 claims description 4
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000002759 z-score normalization Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 abstract description 3
- 238000013102 re-test Methods 0.000 abstract description 3
- 238000011425 standardization method Methods 0.000 abstract 1
- 239000000463 material Substances 0.000 description 6
- 230000002547 anomalous effect Effects 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010835 comparative analysis Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000013179 statistical model Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- Quality & Reliability (AREA)
- Artificial Intelligence (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Operations Research (AREA)
- Primary Health Care (AREA)
- Life Sciences & Earth Sciences (AREA)
- Water Supply & Treatment (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention provides an auxiliary judging method of a distribution transformer detection result, S1, sorting and extracting qualified distribution transformer sampling detection data, and sorting the qualified distribution transformer sampling detection data into a data characteristic sample set; s2, performing data cleaning on the data characteristic sample set; s3, carrying out normalization processing on the data characteristic quantity by using a Z-score standardization method; s4, establishing a single-class support vector machine judgment model to obtain a judgment model parameter; and S5, after training is finished, judging by using the established judgment model according to the test data set, marking abnormal points and outputting results. The method comprises the steps of preprocessing detection data of all indexes of a plurality of detection reports, training normal data by using an OCSVM (online closed system virtual machine) to obtain reasonable model parameters, and distinguishing a test set by using a trained model, so that an abnormal detection report suspected of being wrongly evaluated can be found out. And an experimenter can refer to the evaluation result to determine whether to perform retest on the transformer corresponding to the abnormal detection report.
Description
Technical Field
The invention belongs to the field of material detection of power systems, and provides an auxiliary judgment method for a detection result of a distribution transformer.
Background
The power grid material detection is the key work of putting good power grid materials into a gateway and ensuring the intrinsic safety of a power grid. With the advance of the upgrading and transforming work of distribution networks and rural networks, the coverage and working requirements of power grid material detection are constantly improved, the detection traffic volume is increasing day by day, the existing material quality detection system has some defects in the aspects of bearing capacity, working efficiency, resource utilization, deep analysis and the like, can not adapt to the practical needs of rapid development of power grids and intensive material resource management, and needs to be improved in technology and management ways urgently.
The 10kV distribution transformer is a main device in a distribution network, and has a plurality of detection indexes and procedures in the detection process. In the actual detection process, due to special conditions such as errors of a measuring instrument, negligence of inspection workers and the like, unqualified transformer products can be judged to be qualified products, so that the unqualified products flow into a power grid, and adverse consequences are caused. In order to avoid the consequence as much as possible, in the detection process, if some 10kV transformers are found to have partial indexes higher or close to the qualified limit value, but are within the qualified range; in this case, the inspector may determine whether or not to recheck the transformer based on experience and intuition.
Disclosure of Invention
In view of the above problems, the present invention provides an auxiliary evaluation method for a detection result of a distribution transformer, which is used for auxiliary evaluation of the detection result of the distribution transformer. The method and the device can find out the suspected misjudged detection report for the reference of detection personnel so as to determine whether to perform the retest on the transformer corresponding to the suspected misjudged detection report.
The technical scheme of the invention is as follows:
an auxiliary evaluation method for a detection result of a distribution transformer comprises the following specific steps:
s1, sorting and extracting qualified sampling detection data of distribution network transformers, classifying the obtained data according to transformer capacity, sorting the transformer data with the capacity of 10kV, which accounts for the highest proportion of the data, into a data characteristic sample set, regarding the transformer as a sample, and regarding data generated by various tests performed on the transformer as the characteristic attribute of the sample;
s2, performing data cleaning on the data characteristic sample set, removing blank or repeated detection data, and converting the data characteristic sample set subjected to data cleaning into a vector data set;
s3, carrying out normalization processing on the data characteristic quantity by using a Z-score normalization method, wherein the processed data conform to standard normal distribution, namely the mean value is 0, the standard deviation is 1, and the conversion function is as follows:
in the formula, x*Is the value of Z, x is the normalized random variable, u is the sample mean, σ is the sample standard deviation;
s4, establishing a single-class support vector machine judgment model, and training the data set subjected to normalization processing by using the judgment model to obtain a judgment model parameter;
and S5, after training is finished, judging by using the established judgment model according to the test data set, marking abnormal points, outputting results, and verifying the effect of auxiliary judgment of the judgment model.
The detection data in the step S1 include the dc resistance, the short-circuit resistance, the ground resistance, the unbalance rate, the no-load loss experimental data, the short-circuit loss experimental data, the induced withstand voltage and the applied withstand voltage experimental data, and the temperature rise experimental data of the transformer, and can relatively comprehensively reflect the quality level of the transformer.
In step S4, a single-type support vector machine model is directly established by using a python development platform of a skearn platform according to the following statements:
OCSVM(nu=,kernel=,gamma=)
wherein nu represents the number of abnormal points in the training set, and the value range is [0, 1 ]; kernel represents the selected kernel function, and various kernel functions of 'rbf', 'poly' and 'sigmoid' are available; gamma is a parameter of the 'rbf', 'poly' or 'sigmoid' kernel function for controlling the continuity of the boundary between the normal point and the abnormal point, the lower the gamma, the higher the continuity of the boundary, the more likely the continuity is to surround all the normal data points with one hypersphere, and vice versa, the more likely the hypersphere is to surround the normal data points respectively, but overfitting is easy to occur.
The test data set in step S5 is a artificially constructed test set in which the abnormal points are distributed.
Compared with the prior art, the invention has the beneficial effects that: the method comprises the steps of preprocessing detection data of all indexes of a plurality of detection reports, training normal data by using an OCSVM (online closed system virtual machine) to obtain reasonable model parameters, and distinguishing a test set by using a trained model, so that an abnormal detection report suspected of being wrongly evaluated can be found out. And an experimenter can refer to the evaluation result to determine whether to perform retest on the transformer corresponding to the abnormal detection report.
Drawings
FIG. 1 is a flow chart of an implementation of the present invention;
FIG. 2 is a graph of the distance of each data point of the test data set from the boundary hypersphere.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
According to the statistical probability, the detection results close to most of the detection results are normal and credible, and although each item of detection data of few detection results meets the standard, the value distribution and the proportional relation between the values are different from most of the normal detection results, so that the detection results are considered to be unreliable and possibly unqualified products which are misjudged to be normal products. Moreover, the distribution of such unreliable detection results in the data space should be in a region where the data points are sparse, most of the data is normal data, and in the feature space, the normal data and abnormal data occupy different spatial regions. If the area occupied by normal data in the feature space can be estimated, points not in the area can be considered as abnormal points, and the normal data and the abnormal data are separated by a hyperplane by using a single-class Support vector machine (OCSVM) and compressing the area occupied by the data set in the feature space by maximizing the classification boundary.
Aiming at the auxiliary judgment requirement of the detection result of the power distribution transformer of the power grid, the invention takes the qualified detection result (positive sample) of the sampling of the 10kV transformer of a certain power-saving network as a research object and adopts an OCSVM algorithm to evaluate the detection result. The method comprises the steps of preprocessing detection data of all indexes of a plurality of detection reports, training normal data by using an OCSVM (online closed system virtual machine) to obtain reasonable model parameters, and distinguishing a test set by using a trained model, so that an abnormal detection report suspected of being wrongly evaluated can be found out.
OCSVM-based detection result auxiliary judgment method for abnormal point identification
Anomaly detection is an important component in data mining. Anomalous data refers to small portions of data in a data set that deviate from a large portion of data or are not subject to the same statistical model as other large portions of data in the data set. And the anomaly detection is to identify anomalous data. Anomaly detection machine learning algorithms can be divided into 2 major categories: supervised learning algorithms and unsupervised learning algorithms. The OCSVM algorithm detection result in the unsupervised learning algorithm is used for judging.
The principle is that after data cleaning is carried out on detection data (positive samples) of a batch of normal distribution network transformer detection reports, normalization processing is carried out, and then an OCSVM is used for training a normal report data set to obtain reasonable model parameters. After training, the model is used for testing the detection report set to be tested, so that abnormal points, namely detection reports suspected of misjudgment, are obtained.
OCSVM principle
The OCSVM is a refinement of a two-classification support vector machine, and is a novel machine learning algorithm developed on the basis of a statistical learning theory. The OCSVM adopts a structure minimization principle, has minimized risk experience and confidence range, has the advantages of high fitting precision, few selected parameters, strong popularization capability, global optimization and the like, and belongs to an important classical small sample learning algorithm in the field of anomaly detection. The detection of anomalous data by the OCSVM may provide good generalization capability when determining the appropriate parameter configuration.
OCSVM utilizes a sample set { xi,i=1,2,..,l},xi∈RnAnd mapping the non-linear kernel function phi (x) in a high-dimensional feature space H, and calculating a minimum hypersphere containing as many samples as possible as a decision boundary. Wherein a is the center of the hyper-sphere and R is the radius of the hyper-sphere. The solution of the minimum hypersphere containing all training samples can be transformed into the following convex optimization problem:
where C is a penalty factor, ξi> 0 is a relaxation variable. The inner product operation in the high-dimensional space optimization is replaced by a kernel function meeting the conditions of the mercer, namely, a kernel function is found so that:
K(x,y)=<Φ(x),Φ(y)>(2)
the dual problem of equation (1) can be expressed as:
the lagrange method can be used for solving the problem formula (3) to obtain aiOnly a small part aiAnd the corresponding point is a support vector when the value is more than or equal to 0. The center a of the hyper-sphere can be represented byThe radius R can be determined by substituting any support vector into the following equation:
the final form of the decision function is:
for a given x, a normal point is determined when f (x) > 0 and an outlier is determined when f (x) ≦ 0.
Referring to fig. 1, the present invention provides a technical solution:
an auxiliary evaluation method for distribution network transformer inspection results comprises the following specific steps:
s1, sorting and extracting qualified sampling detection data of the distribution transformer, classifying the obtained data according to transformer capacity, sorting the transformer data with the capacity of 10kV, which accounts for the highest proportion of the data, into a data characteristic sample set, regarding the transformer as a sample, and regarding data generated by various tests performed on the transformer as the characteristic attribute of the sample;
s2, performing data cleaning on the data characteristic sample set, removing blank or repeated detection data, and converting the data characteristic sample set subjected to data cleaning into a vector data set;
s3, carrying out normalization processing on the data characteristic quantity by using a Z-score normalization method, wherein the processed data conform to standard normal distribution, namely the mean value is 0, the standard deviation is 1, and the conversion function is as follows:
in the formula, x*Is the value of Z, x is the normalized random variable, u is the sample mean, σ is the sample standard deviation;
s4, establishing a single-class support vector machine judgment model, and training the data set subjected to normalization processing by using the judgment model to obtain a judgment model parameter;
and S5, after training is finished, judging by using the established judgment model according to the test data set, marking abnormal points, outputting results, and verifying the effect of auxiliary judgment of the judgment model.
The technical scheme of the invention is further specifically explained by specific calculation examples.
(1) Data samples
And sorting and extracting qualified transformer sampling detection data of 2016, 2017 and 2018 of a certain power saving network, namely all positive samples. The detection data comprise direct current resistance, short-circuit resistance, ground resistance, unbalance rate, no-load loss experimental data, short-circuit loss experimental data, induction voltage-withstanding and external voltage-withstanding experimental data, temperature rise experimental data and the like of the transformer, and the quality level of the transformer can be comprehensively reflected.
The obtained data are classified according to the transformer capacity, wherein the transformer data with the capacity of 10kV (100KVA) which accounts for the highest proportion of the data are collected and are arranged into a data characteristic sample set. The data generated by tests performed on a transformer is taken as a characteristic attribute of a sample. The data is cleaned and removed incompletely, and the finally obtained training set has 107 samples in total, and each sample contains 59 characteristic attributes.
And determining the abnormal point detection effect of each model, and manually constructing a test set in which the abnormal points are distributed for testing. The test set used herein contains 97 samples, each sample contains 59 attributes, wherein 5 points numbered 92 to 96 are artificially constructed abnormal samples, and the detection data of the abnormal samples are in accordance with the power grid transformer standard, but have a certain difference with the detection data of the normal transformer.
(2) Model training
① model implementation
Using the python development platform of the sklern platform, the Oneclass SVM model can be directly established by the following statements:
OneclassSVM(nu=,kernel=,gamma=)
wherein nu represents the number of abnormal points in the training set, the value range is (0, 1), kernel represents the selected kernel function, various kernel functions such as 'rbf', 'poly' and 'sigmoid' can be selected, gamma is a parameter of the kernel functions such as 'rbf', 'poly' or 'sigmoid' and is used for controlling the continuity of the boundary between the normal points and the abnormal points, the lower the gamma, the higher the continuity of the boundary, the higher the continuity, the more the normal data points are surrounded by a hypersphere, otherwise, the more the hypersphere respectively surround the normal data points, but overfitting is easy to occur.
② OCSVM parameter tuning
For training data, all the transformers are defaulted to be positive samples according to the detection result of a certain provincial power grid, namely all the transformers are qualified products, and the parameter nu cannot be 0, so that the parameter nu is set to be 0.01 which is closer to the actual failure rate.
Setting kernel function parameters:
the commonly used kernel functions of the OCSVM model include polynomial kernel functions, RBF kernel functions, gaussian kernel functions, etc., and the optimal kernel functions and parameters are usually selected according to the trial calculation result. By comparative analysis, the mean of the gaussian kernel was chosen to be 0 and the variance was chosen to be 500.
Because the training sample data used for training are all qualified transformers which pass through the power grid spot check, when the proportion of abnormal points is set to be 0.01 for the output results of the machine learning algorithms, the closer the corresponding quantity of the theoretically detected abnormal points is to the proportion, the better the fitting degree of the model is; when the abnormal point ratio is set to 0, theoretically, the fewer abnormal points in the training data are detected, the better the fitting degree of the model is. From the program tuning operation result, the parameter configuration is nu equal to 0.01, kernel equal to 'rbf', and the fitting degree of gamma equal to 0.0011 is very good.
(3) Testing
The test samples described above were tested using an OCSVM model with parameters nu 0.01, kernel rbf, and gamma 0.0011, the set of test samples comprising 97 samples, each sample containing 59 attributes, of which 5 points numbered 92 to 96 are artificially constructed abnormal samples.
In the test results, the distance of each data point from the boundary hypersphere is shown in FIG. 2,
it can be seen that there are 5 negative distance samples, numbered 92, 93, 94, 95, 96, i.e. 5 outliers are detected.
Therefore, the suspected abnormality detection report can be effectively found out by using the OCSVM algorithm.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (4)
1. An auxiliary evaluation method for detection results of a distribution transformer is characterized by comprising the following specific steps:
s1, sorting and extracting qualified sampling detection data of the distribution transformer, classifying the obtained data according to transformer capacity, sorting the transformer data with the capacity of 10kV, which accounts for the highest proportion of the data, into a data characteristic sample set, regarding the transformer as a sample, and regarding data generated by various tests performed on the transformer as the characteristic attribute of the sample;
s2, performing data cleaning on the data characteristic sample set, removing blank or repeated detection data, and converting the data characteristic sample set subjected to data cleaning into a vector data set;
s3, carrying out normalization processing on the data characteristic quantity by using a Z-score normalization method, wherein the processed data conform to standard normal distribution, namely the mean value is 0, the standard deviation is 1, and the conversion function is as follows:
in the formula, x*Is the value of Z, x is the normalized random variable, u is the sample mean, σ is the sample standard deviation;
s4, establishing a single-class support vector machine judgment model, and training the data set subjected to normalization processing by using the judgment model to obtain a judgment model parameter;
and S5, after training is finished, judging by using the established judgment model according to the test data set, marking abnormal points, outputting results, and verifying the effect of auxiliary judgment of the judgment model.
2. The auxiliary evaluation method for the distribution transformer detection result according to claim 1, wherein the detection data in the step S1 includes dc resistance, short-circuit resistance, ground resistance, unbalance rate, no-load loss experimental data, short-circuit loss experimental data, induced withstand voltage and applied withstand voltage experimental data, and temperature rise experimental data of the transformer, and can relatively comprehensively reflect the quality level of the transformer.
3. The auxiliary evaluation method for the distribution transformer detection result according to claim 1, wherein in step S4, a python development platform of a sklern platform is used, and the single-class support vector machine model is directly built by the following statements:
OCSVM(nu=,kernel=,gamma=)
wherein nu represents the number of abnormal points in the training set, and the value range is [0, 1 ]; kernel represents the selected kernel function, and various kernel functions of 'rbf', 'poly' and 'sigmoid' are available; gamma is a parameter of the 'rbf', 'poly' or 'sigmoid' kernel function for controlling the continuity of the boundary between the normal point and the abnormal point, the lower the gamma, the higher the continuity of the boundary, the more likely the continuity is to surround all the normal data points with one hypersphere, and vice versa, the more likely the hypersphere is to surround the normal data points respectively, but overfitting is easy to occur.
4. The method of claim 1, wherein the test data set in step S5 is a artificially constructed test set with known distribution of outliers.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910933689.6A CN110689324A (en) | 2019-09-29 | 2019-09-29 | Auxiliary judging method for detection result of distribution transformer |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910933689.6A CN110689324A (en) | 2019-09-29 | 2019-09-29 | Auxiliary judging method for detection result of distribution transformer |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110689324A true CN110689324A (en) | 2020-01-14 |
Family
ID=69110964
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910933689.6A Pending CN110689324A (en) | 2019-09-29 | 2019-09-29 | Auxiliary judging method for detection result of distribution transformer |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110689324A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111553396A (en) * | 2020-04-21 | 2020-08-18 | 广东电网有限责任公司电力科学研究院 | Online detection method and device for distribution automation terminal |
CN113945329A (en) * | 2021-10-14 | 2022-01-18 | 西安西电开关电气有限公司 | SF (sulfur hexafluoride)6Gas leakage defect studying and judging method and system |
CN116703245A (en) * | 2023-08-02 | 2023-09-05 | 宜兴市旭航电子有限公司 | System and method for monitoring and early warning of flow production process |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101907088A (en) * | 2010-05-27 | 2010-12-08 | 中国人民解放军国防科学技术大学 | Fault diagnosis method based on one-class support vector machines |
CN102339389A (en) * | 2011-09-14 | 2012-02-01 | 清华大学 | Fault detection method for one-class support vector machine based on density parameter optimization |
CN107229849A (en) * | 2016-03-24 | 2017-10-03 | 全球能源互联网研究院 | Towards the database user behavior safety auditing method on power information intranet and extranet border |
CN107728028A (en) * | 2017-11-07 | 2018-02-23 | 武汉大学 | GIS partial discharge fault distinguishing method based on one-class support vector machines |
CN110207997A (en) * | 2019-07-24 | 2019-09-06 | 中国人民解放军国防科技大学 | Liquid rocket engine fault detection method based on convolution self-encoder |
CN110287447A (en) * | 2019-06-18 | 2019-09-27 | 浙江工业大学 | A kind of networking multi-shaft motion control system sine attack detection method based on one-class support vector machines |
-
2019
- 2019-09-29 CN CN201910933689.6A patent/CN110689324A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101907088A (en) * | 2010-05-27 | 2010-12-08 | 中国人民解放军国防科学技术大学 | Fault diagnosis method based on one-class support vector machines |
CN102339389A (en) * | 2011-09-14 | 2012-02-01 | 清华大学 | Fault detection method for one-class support vector machine based on density parameter optimization |
CN107229849A (en) * | 2016-03-24 | 2017-10-03 | 全球能源互联网研究院 | Towards the database user behavior safety auditing method on power information intranet and extranet border |
CN107728028A (en) * | 2017-11-07 | 2018-02-23 | 武汉大学 | GIS partial discharge fault distinguishing method based on one-class support vector machines |
CN110287447A (en) * | 2019-06-18 | 2019-09-27 | 浙江工业大学 | A kind of networking multi-shaft motion control system sine attack detection method based on one-class support vector machines |
CN110207997A (en) * | 2019-07-24 | 2019-09-06 | 中国人民解放军国防科技大学 | Liquid rocket engine fault detection method based on convolution self-encoder |
Non-Patent Citations (1)
Title |
---|
郭利花: ""基于新异类检测技术的变压器振动异常检测"", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111553396A (en) * | 2020-04-21 | 2020-08-18 | 广东电网有限责任公司电力科学研究院 | Online detection method and device for distribution automation terminal |
CN113945329A (en) * | 2021-10-14 | 2022-01-18 | 西安西电开关电气有限公司 | SF (sulfur hexafluoride)6Gas leakage defect studying and judging method and system |
CN113945329B (en) * | 2021-10-14 | 2024-01-23 | 西安西电开关电气有限公司 | SF (sulfur hexafluoride) 6 Method and system for judging gas leakage defect |
CN116703245A (en) * | 2023-08-02 | 2023-09-05 | 宜兴市旭航电子有限公司 | System and method for monitoring and early warning of flow production process |
CN116703245B (en) * | 2023-08-02 | 2023-10-27 | 宜兴市旭航电子有限公司 | System and method for monitoring and early warning of flow production process |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110336534B (en) | Fault diagnosis method based on photovoltaic array electrical parameter time series feature extraction | |
CN110763929A (en) | Intelligent monitoring and early warning system and method for convertor station equipment | |
CN104809658B (en) | A kind of rapid analysis method of low-voltage distribution network taiwan area line loss | |
CN110689324A (en) | Auxiliary judging method for detection result of distribution transformer | |
CN113378990B (en) | Flow data anomaly detection method based on deep learning | |
CN112149873B (en) | Low-voltage station line loss reasonable interval prediction method based on deep learning | |
Jia et al. | Defect prediction of relay protection systems based on LSSVM-BNDT | |
CN111582406A (en) | Power equipment state monitoring data clustering method and system | |
CN110889441A (en) | Distance and point density based substation equipment data anomaly identification method | |
CN110930057A (en) | Quantitative evaluation method for reliability of distribution transformer test result based on LOF algorithm | |
CN110705859A (en) | PCA-self-organizing neural network-based method for evaluating running state of medium and low voltage distribution network | |
CN112417763A (en) | Defect diagnosis method, device and equipment for power transmission line and storage medium | |
CN114580934A (en) | Early warning method for food detection data risk based on unsupervised anomaly detection | |
CN115511387A (en) | Oil immersed transformer state evaluation method and system based on multi-source information fusion | |
CN116467658A (en) | Equipment fault tracing method based on Markov chain | |
CN116933010A (en) | Load rate analysis and evaluation method and system based on multi-source data fusion and deep learning | |
CN111027841A (en) | Low-voltage transformer area line loss calculation method based on gradient lifting decision tree | |
CN110096723B (en) | High-voltage switch cabinet insulation state analysis method based on operation and maintenance detection big data | |
CN117932501A (en) | Electric energy meter running state management method and system | |
CN110232399A (en) | The transmission facility defect analysis method and system clustered based on Set Pair Analysis and K-means | |
CN113484817A (en) | Intelligent electric energy meter automatic verification system abnormity detection method based on TSVM model | |
CN113705920B (en) | Method for generating water data sample set for thermal power plant and terminal equipment | |
CN116861232A (en) | Air quality data anomaly detection model based on DBN-OCSVM | |
CN111209955A (en) | Airplane power supply system fault identification method based on deep neural network and random forest | |
CN116151799A (en) | BP neural network-based distribution line multi-working-condition fault rate rapid assessment method |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200114 |