CN117540344A - Distribution transformer anomaly monitoring and operation trend prediction method, equipment and medium - Google Patents
Distribution transformer anomaly monitoring and operation trend prediction method, equipment and medium Download PDFInfo
- Publication number
- CN117540344A CN117540344A CN202311482890.XA CN202311482890A CN117540344A CN 117540344 A CN117540344 A CN 117540344A CN 202311482890 A CN202311482890 A CN 202311482890A CN 117540344 A CN117540344 A CN 117540344A
- Authority
- CN
- China
- Prior art keywords
- distribution transformer
- data
- index
- sampling
- indexes
- 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 title claims abstract description 31
- 238000000034 method Methods 0.000 title claims abstract description 30
- 230000002159 abnormal effect Effects 0.000 claims abstract description 50
- 238000005070 sampling Methods 0.000 claims abstract description 38
- 238000012417 linear regression Methods 0.000 claims abstract description 22
- 239000011159 matrix material Substances 0.000 claims abstract description 22
- 238000001514 detection method Methods 0.000 claims abstract description 16
- 238000011156 evaluation Methods 0.000 claims abstract description 14
- 238000007781 pre-processing Methods 0.000 claims abstract description 8
- 238000003066 decision tree Methods 0.000 claims description 9
- 238000002955 isolation Methods 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 6
- 239000006185 dispersion Substances 0.000 claims description 6
- 230000015654 memory Effects 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 5
- 238000012935 Averaging Methods 0.000 claims description 4
- 230000005856 abnormality Effects 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 230000000452 restraining effect Effects 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 3
- 238000012423 maintenance Methods 0.000 description 7
- 230000008901 benefit Effects 0.000 description 4
- 230000008439 repair process Effects 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 230000008034 disappearance Effects 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 238000004880 explosion Methods 0.000 description 2
- 230000036541 health Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/27—Regression, e.g. linear or logistic regression
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- 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
- G06Q10/20—Administration of product repair or maintenance
-
- 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
-
- 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)
- Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Economics (AREA)
- Evolutionary Computation (AREA)
- Human Resources & Organizations (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Strategic Management (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Marketing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Molecular Biology (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Entrepreneurship & Innovation (AREA)
- Computing Systems (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a distribution transformer anomaly monitoring and running trend prediction method, equipment and medium, which comprise the following specific steps: collecting distribution transformer operation data of a target detection area, and preprocessing the data based on over sampling and under sampling to obtain a sample set; constructing a linear regression model by adopting kernel ridge regression, carrying out anomaly monitoring on the sample set, and determining anomaly indexes; constructing an average matrix based on the abnormal data indexes, and determining the weight of the abnormal data indexes; calculating the primary score of the running state evaluation of the distribution transformer in real time based on the index weight; and evaluating the final score of the operation state of the distribution transformer based on the preliminary score of the operation state of the distribution transformer to be evaluated. By adopting the nuclear ridge regression to construct a linear regression model, abnormal detection is carried out on the sample set, multiple prediction is carried out on the running state of the distribution transformer, accurate evaluation and prediction are carried out on the running state of the distribution transformer, the running state of the distribution transformer can be timely repaired by hearing, and power grid faults are reduced.
Description
Technical Field
The invention relates to the technical field of power grid anomaly monitoring, in particular to a distribution transformer anomaly monitoring and operation trend prediction method, equipment and medium.
Background
The distribution transformer is used as the most important electrical equipment in a distribution network system, and can reduce the voltage level and transmit continuous and high-quality electric energy to different users, so that once the transformer has serious faults, the transformer is light to cause regional power failure events, and the electricity utilization experience of the users is seriously affected; the running condition of the distribution transformer is sensed in real time, the running health degree of the distribution transformer at the current running node is mastered, and the potential abnormal distribution transformer which possibly breaks down can be aimed at, so that the running trend of the distribution transformer in a future period of time is deduced in a rolling way based on the running and collecting data of the current time node, the possibly broken-down transformer is found, and the running and maintenance team is arranged for timely rush repair, so that the distribution network has great significance in improving the power supply reliability of the distribution network and the electricity utilization reliability of users.
If the running state of the distribution transformer can be monitored in real time, timely rush repair and predictive maintenance are carried out on equipment which is likely to fail, the traditional mode of on-site inspection and failure and rush repair is changed, and the power supply reliability of the low-voltage distribution network and the power consumption reliability of users are greatly improved.
However, the operation and maintenance means of the low-voltage distribution network operation and maintenance team on the distribution transformer at present is mainly on site inspection and fault first-aid repair, when faults occur and maintenance is carried out again, power failure accidents are inevitably caused, and an existing distribution information management system does not have an on-line monitoring module or system specially aiming at the operation state of the distribution transformer, so that operation and maintenance staff cannot master the operation health state of the distribution transformer in real time, and purposeful and targeted prediction maintenance is carried out. Therefore, under the rapid promotion trend of ubiquitous power internet of things construction, a power distribution internet of things system framework facing to the application scene of distribution transformer running state evaluation and trend prediction is built, a set of running state evaluation and prediction system of a distribution transformer is designed and realized, real-time accurate evaluation is carried out on the running state of the distribution transformer by one station, timely deduction is carried out on the running state of the distribution transformer under abnormal working conditions, safe, stable and reliable green running of a power distribution network can be ensured, the power supply reliability of the power distribution network and the power utilization reliability of users can be improved, considerable economic benefit and social benefit are brought, and meanwhile, the power distribution network system has important practical significance and engineering significance for the construction of a strong intelligent power grid.
Disclosure of Invention
The invention aims to provide a method, equipment and medium for monitoring abnormity of a distribution transformer and predicting operation trend, which aims to solve the technical problems that the monitoring accuracy of data is low and the problems cannot be found in time.
The invention is realized by the following technical scheme:
the first aspect of the invention provides a distribution transformer abnormality monitoring and operation trend prediction method, which comprises the following specific steps:
collecting distribution transformer operation data of a target detection area, and preprocessing the data based on over sampling and under sampling to obtain a sample set;
constructing a linear regression model by adopting kernel ridge regression, carrying out anomaly monitoring on the sample set, and determining anomaly indexes;
constructing an average matrix based on the abnormal data indexes, and determining the weight of the abnormal data indexes;
calculating the primary score of the running state evaluation of the distribution transformer in real time based on the index weight;
and evaluating the final score of the operation state of the distribution transformer based on the preliminary score of the operation state of the distribution transformer to be evaluated.
According to the invention, the distribution transformer operation data of the target detection area are collected, the data are preprocessed based on over sampling and under sampling, the effectiveness of the collected data is improved, a kernel-ridge regression is adopted to construct a linear regression model, abnormal detection is carried out on a sample set, abnormal data index weights are obtained to carry out operation state estimation, the operation state and the operation trend of the distribution transformer are predicted, the index weights are obtained, the preliminary score of the operation state estimation of the distribution transformer is calculated in real time, the final score of the operation state of the distribution transformer is estimated through the preliminary score, the operation trend prediction is carried out, the operation state multidimensional estimation of the distribution transformer can be realized, the operation state of the distribution transformer is accurately estimated and predicted, the operation state of the distribution transformer can be timely reacted, and the power grid faults are reduced.
Further, the preprocessing of the data based on the over-sampling and the under-sampling specifically includes:
sampling the acquired data by adopting over sampling and under sampling to obtain sampled data;
the number of decision trees constructed based on the sampled data;
and carrying out data specification on the number of the decision trees by adopting a maximum and minimum method to obtain a sample set.
Further, the number of decision trees constructed based on the sampled data specifically includes:
acquiring characteristic attributes and definition values of the sampling data, and recursively dividing the sampling data until each data object is represented by a binary tree;
constructing an isolation number, extracting a sample number set from original sampling data, and determining a root node;
selecting an isolation attribute and an isolation value, calculating an attribute value of a sample point in the set, and placing nodes according to the attribute value;
the construction of new leaf nodes is repeated until the leaf nodes have reached a preset height with only one sample or tree.
Further, the performing anomaly monitoring on the sample set specifically includes:
constructing a linear regression model by adopting kernel ridge regression, and inputting a sample set into the linear regression model;
regularization is introduced into the regression model, and the complexity of the linear regression model is calculated;
and (3) acquiring historical abnormal data indexes, inputting the historical abnormal data indexes into a linear regression model, and determining the abnormal data indexes by combining the complexity to obtain the current abnormal data indexes.
Further, the obtaining the historical abnormal data index includes:
constructing an LSTM model, and inputting historical operation data of the distribution transformer in a target detection area into the LSTM;
performing time sequence feature extraction based on LSTM, and outputting a time sequence feature sequence of the real-time index;
setting the time scale of the acquired historical operation data, and counting the predicted value of the statistical type index from the predicted value of the real-time type index according to the time scale;
and collecting real-time index and statistical index predicted values to obtain historical abnormal data indexes.
Further, the constructing an average matrix based on the abnormal data index, and determining the abnormal data index weight specifically includes:
constructing a plurality of judgment matrixes based on the abnormal data indexes, and comparing every two judgment matrixes;
averaging the plurality of judgment matrixes to obtain an average matrix of the plurality of judgment matrixes;
calculating the dispersion, the maximum eigenvalue and the eigenvector of the average matrix;
and carrying out normalization verification on the matrix based on the dispersion, the maximum eigenvalue and the eigenvector to obtain the abnormal data index weight.
Further, the calculating the preliminary score of the running state evaluation of the distribution transformer in real time specifically includes:
calculating the score of each single index in real time according to the weight of the abnormal data index;
and counting to obtain single index scores of historical abnormal data indexes, and obtaining preliminary scores of the running states of the distribution transformer to be evaluated.
Further, the final score for evaluating the operational status of the distribution transformer specifically includes:
constructing a preliminary scoring matrix of the operation state of the distribution transformer to be evaluated, and obtaining a matrix characteristic vector;
traversing the feature vector, and restraining the feature vector by using a nonnegative linear least square algorithm;
and setting a constraint threshold until the difference between the two iteration values converges, and outputting a final score of the operation state of the distribution transformer to be evaluated.
In a second aspect, the present invention provides an electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing a distribution transformer anomaly monitoring and operational trend prediction method when executing the program.
A third aspect of the present invention provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements a distribution transformer anomaly monitoring and operational trend prediction method.
Compared with the prior art, the invention has the following advantages and beneficial effects:
the method comprises the steps of collecting distribution transformer operation data of a target detection area, preprocessing the data based on over sampling and under sampling, improving the effectiveness of the collected data, constructing a linear regression model by adopting kernel-ridge regression, carrying out anomaly detection on a sample set, acquiring an anomaly data index weight, carrying out operation state estimation, predicting the operation state and the operation trend of the distribution transformer, acquiring the index weight, calculating an initial evaluation score of the operation state of the distribution transformer in real time, evaluating the final score of the operation state of the distribution transformer through the initial score, carrying out operation trend prediction, realizing multidimensional evaluation of the operation state of the distribution transformer, accurately evaluating and predicting the operation state of the distribution transformer, timely reacting to the operation state of the distribution transformer, and reducing power grid faults.
Drawings
In order to more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the drawings that are needed in the examples will be briefly described below, it being understood that the following drawings only illustrate some examples of the present invention and therefore should not be considered as limiting the scope, and that other related drawings may be obtained from these drawings without inventive effort for a person skilled in the art. In the drawings:
FIG. 1 is a flow chart of a method according to an embodiment of the invention.
Detailed Description
For the purpose of making apparent the objects, technical solutions and advantages of the present invention, the present invention will be further described in detail with reference to the following examples and the accompanying drawings, wherein the exemplary embodiments of the present invention and the descriptions thereof are for illustrating the present invention only and are not to be construed as limiting the present invention.
Example 1
As shown in fig. 1, a first aspect of the present embodiment provides a method for monitoring abnormality and predicting operation trend of a distribution transformer, including the following specific steps:
collecting distribution transformer operation data of a target detection area, and preprocessing the data based on over sampling and under sampling to obtain a sample set;
constructing a linear regression model by adopting kernel ridge regression, carrying out anomaly monitoring on the sample set, and determining anomaly indexes;
constructing an average matrix based on the abnormal data indexes, and determining the weight of the abnormal data indexes;
calculating the primary score of the running state evaluation of the distribution transformer in real time based on the index weight;
and evaluating the final score of the operation state of the distribution transformer based on the preliminary score of the operation state of the distribution transformer to be evaluated.
According to the embodiment, the distribution transformer operation data of the target detection area are collected, the data are preprocessed based on over sampling and under sampling, the effectiveness of the collected data is improved, a kernel-ridge regression is adopted to construct a linear regression model, abnormal detection is carried out on a sample set, abnormal data index weights are obtained to carry out operation state estimation, the operation state and the operation trend of the distribution transformer are predicted, index weights are obtained, the initial score of the operation state estimation of the distribution transformer is calculated in real time, the final score of the operation state of the distribution transformer is estimated through the initial score, the operation trend prediction is carried out, the operation state multidimensional estimation of the distribution transformer can be realized, accurate estimation and prediction are carried out on the operation state of the distribution transformer, the operation state of the distribution transformer can be timely responded, and the power grid faults are reduced.
In some possible embodiments, during the data collection process of the distribution transformer area, due to inaccurate recorded data caused by factors such as equipment and human factors, abnormal data exist, and these data cannot reflect the objective real operation situation of the distribution transformer area, if the abnormal detection and the operation trend evaluation are directly performed by using these data, the judgment capability during the detection will be interfered, and the prediction accuracy will not be high, so that the collected operation data needs to be subjected to unbalanced processing, and the preprocessing of the data based on over sampling and under sampling specifically includes:
sampling the acquired data by adopting over sampling and under sampling to obtain sampled data;
the number of decision trees constructed based on the sampled data;
and carrying out data specification on the number of the decision trees by adopting a maximum and minimum method to obtain a sample set.
In some possible embodiments, the number of decision trees constructed based on the sampled data specifically includes:
acquiring characteristic attributes and definition values of the sampling data, and recursively dividing the sampling data until each data object is represented by a binary tree;
constructing an isolation number, extracting a sample number set from original sampling data, and determining a root node;
selecting an isolation attribute and an isolation value, calculating an attribute value of a sample point in the set, and placing nodes according to the attribute value;
the construction of new leaf nodes is repeated until the leaf nodes have reached a preset height with only one sample or tree.
In some possible embodiments, the anomaly monitoring of the sample set specifically includes:
constructing a linear regression model by adopting kernel ridge regression, and inputting a sample set into the linear regression model;
regularization is introduced into the regression model, and the complexity of the linear regression model is calculated;
and (3) acquiring historical abnormal data indexes, inputting the historical abnormal data indexes into a linear regression model, and determining the abnormal data indexes by combining the complexity to obtain the current abnormal data indexes. The data in the sample space is converted into high-dimensional characteristic space data through a nonlinear transformation by using a kernel method, so that a linear function of an original space constructed by the corresponding kernel and the data can be learned, and the nonlinear function in the original space of the corresponding nonlinear kernel has better nonlinear fitting performance. Data that cannot be modeled by a linear regression method in the original sample space can be modeled by a linear regression method in a high-dimensional feature space.
In some possible embodiments, obtaining the historical anomaly data indicator includes:
constructing an LSTM model, and inputting historical operation data of the distribution transformer in a target detection area into the LSTM;
performing time sequence feature extraction based on LSTM, and outputting a time sequence feature sequence of the real-time index;
setting the time scale of the acquired historical operation data, and counting the predicted value of the statistical type index from the predicted value of the real-time type index according to the time scale;
and collecting real-time index and statistical index predicted values to obtain historical abnormal data indexes.
The LSTM can solve the problems of gradient disappearance and gradient explosion in the traditional cyclic neural network long-sequence training process, and the input gate, the output gate and the forgetting gate are added into neurons, so that the problems of gradient disappearance and gradient explosion can be effectively relieved through the structure. Compared with a common neural network, the LSTM is a popular time series prediction model, and can better process the distribution transformer series change electrical data with long-term dependence. When predicting the running trend of the distribution transformer, the prediction of the real-time index is a multidimensional nonlinear regression task, the time sequence distribution of the bottom data acquired by the real-time index shows a certain degree of non-stationarity, and the change trend of the sequence data shows a certain degree of regularity. Therefore, the embodiment selects the time sequence data sensitive and efficient long and short term memory neural network LSTM to extract the time sequence characteristics of the real-time index to obtain the historical abnormal data index.
Constructing an average matrix based on the abnormal data index, wherein determining the weight of the abnormal data index specifically comprises:
constructing a plurality of judgment matrixes based on the abnormal data indexes, and comparing every two judgment matrixes;
averaging the multiple judgment matrixes, and averaging other elements except the main diagonal element of each pair of judgment matrixes to obtain n corresponding average judgment matrixes, so as to obtain an average matrix of the multiple judgment matrixes;
calculating the dispersion, the maximum eigenvalue and the eigenvector of the average matrix;
and carrying out normalization verification on the matrix based on the dispersion, the maximum eigenvalue and the eigenvector to obtain the abnormal data index weight.
In some possible embodiments, calculating the distribution transformer operating state evaluation preliminary score in real time specifically includes:
calculating the score of each single index in real time according to the weight of the abnormal data index;
and counting to obtain single index scores of historical abnormal data indexes, and obtaining preliminary scores of the running states of the distribution transformer to be evaluated.
Constructing a preliminary scoring matrix of the to-be-evaluated distribution transformer running state based on the preliminary scoring of the to-be-evaluated distribution transformer running state, and obtaining a matrix characteristic vector;
traversing the feature vector, and restraining the feature vector by using a nonnegative linear least square algorithm;
and setting a constraint threshold until the difference between the two iteration values converges, and outputting a final score of the operation state of the distribution transformer to be evaluated.
A second aspect of the present embodiment provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements a distribution transformer anomaly monitoring and operational trend prediction method when executing the program.
A third aspect of the present embodiment provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements a distribution transformer anomaly monitoring and operational trend prediction method.
The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the invention, and is not meant to limit the scope of the invention, but to limit the invention to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the invention are intended to be included within the scope of the invention.
Claims (10)
1. The distribution transformer abnormality monitoring and operation trend predicting method is characterized by comprising the following specific steps:
collecting distribution transformer operation data of a target detection area, and preprocessing the data based on over sampling and under sampling to obtain a sample set;
constructing a linear regression model by adopting kernel ridge regression, carrying out anomaly monitoring on the sample set, and determining anomaly indexes;
constructing an average matrix based on the abnormal data indexes, and determining the weight of the abnormal data indexes;
calculating the primary score of the running state evaluation of the distribution transformer in real time based on the index weight;
and evaluating the final score of the operation state of the distribution transformer based on the preliminary score of the operation state of the distribution transformer to be evaluated.
2. The method for monitoring anomalies and predicting operational trends of a distribution transformer according to claim 1, wherein the pre-processing the data based on over-sampling and under-sampling specifically comprises:
sampling the acquired data by adopting over sampling and under sampling to obtain sampled data;
the number of decision trees constructed based on the sampled data;
and carrying out data specification on the number of the decision trees by adopting a maximum and minimum method to obtain a sample set.
3. The method for monitoring abnormality and predicting operation trend of distribution transformer according to claim 2, wherein the number of decision trees constructed based on the sampled data specifically comprises:
acquiring characteristic attributes and definition values of the sampling data, and recursively dividing the sampling data until each data object is represented by a binary tree;
constructing an isolation number, extracting a sample number set from original sampling data, and determining a root node;
selecting an isolation attribute and an isolation value, calculating an attribute value of a sample point in the set, and placing nodes according to the attribute value;
the construction of new leaf nodes is repeated until the leaf nodes have reached a preset height with only one sample or tree.
4. The method for monitoring anomalies and predicting operational trends of a distribution transformer according to claim 2, wherein said monitoring anomalies of a sample set specifically comprises:
constructing a linear regression model by adopting kernel ridge regression, and inputting a sample set into the linear regression model;
regularization is introduced into the regression model, and the complexity of the linear regression model is calculated;
and (3) acquiring historical abnormal data indexes, inputting the historical abnormal data indexes into a linear regression model, and determining the abnormal data indexes by combining the complexity to obtain the current abnormal data indexes.
5. The method for monitoring and predicting operational trends of a distribution transformer according to claim 4, wherein said obtaining historical anomaly data metrics comprises:
constructing an LSTM model, and inputting historical operation data of the distribution transformer in a target detection area into the LSTM;
performing time sequence feature extraction based on LSTM, and outputting a time sequence feature sequence of the real-time index;
setting the time scale of the acquired historical operation data, and counting the predicted value of the statistical type index from the predicted value of the real-time type index according to the time scale;
and collecting real-time index and statistical index predicted values to obtain historical abnormal data indexes.
6. The method for monitoring anomalies and predicting operational trends of a distribution transformer according to claim 1, wherein said constructing an average matrix based on the anomalies data indexes, determining the anomalies data index weights, specifically comprises:
constructing a plurality of judgment matrixes based on the abnormal data indexes, and comparing every two judgment matrixes;
averaging the plurality of judgment matrixes to obtain an average matrix of the plurality of judgment matrixes;
calculating the dispersion, the maximum eigenvalue and the eigenvector of the average matrix;
and carrying out normalization verification on the matrix based on the dispersion, the maximum eigenvalue and the eigenvector to obtain the abnormal data index weight.
7. The method for monitoring anomalies and predicting operational trends of a distribution transformer according to claim 1, wherein said calculating in real time a preliminary score for an evaluation of the operational state of the distribution transformer comprises:
calculating the score of each single index in real time according to the weight of the abnormal data index;
and counting to obtain single index scores of historical abnormal data indexes, and obtaining preliminary scores of the running states of the distribution transformer to be evaluated.
8. The method for monitoring anomalies and predicting operational trends of a distribution transformer according to claim 7, wherein said evaluating a final score of an operational state of a distribution transformer comprises:
constructing a preliminary scoring matrix of the operation state of the distribution transformer to be evaluated, and obtaining a matrix characteristic vector;
traversing the feature vector, and restraining the feature vector by using a nonnegative linear least square algorithm;
and setting a constraint threshold until the difference between the two iteration values converges, and outputting a final score of the operation state of the distribution transformer to be evaluated.
9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the distribution transformer anomaly monitoring and operational trend prediction method of any one of claims 1 to 8 when the program is executed by the processor.
10. A computer-readable storage medium having stored thereon a computer program which when executed by a processor implements a distribution transformer anomaly monitoring and operational trend prediction method as claimed in any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311482890.XA CN117540344A (en) | 2023-11-07 | 2023-11-07 | Distribution transformer anomaly monitoring and operation trend prediction method, equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311482890.XA CN117540344A (en) | 2023-11-07 | 2023-11-07 | Distribution transformer anomaly monitoring and operation trend prediction method, equipment and medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117540344A true CN117540344A (en) | 2024-02-09 |
Family
ID=89783359
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311482890.XA Pending CN117540344A (en) | 2023-11-07 | 2023-11-07 | Distribution transformer anomaly monitoring and operation trend prediction method, equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117540344A (en) |
-
2023
- 2023-11-07 CN CN202311482890.XA patent/CN117540344A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108320043B (en) | Power distribution network equipment state diagnosis and prediction method based on electric power big data | |
| CN111368890A (en) | Fault detection method and device and information physical fusion system | |
| CN108761377A (en) | A kind of electric energy metering device method for detecting abnormality based on long memory models in short-term | |
| CN107807860B (en) | Power failure analysis method and system based on matrix decomposition | |
| CN112633421A (en) | Method and device for detecting abnormal electricity utilization behavior of user | |
| Zhang et al. | A fault early warning method for auxiliary equipment based on multivariate state estimation technique and sliding window similarity | |
| CN117408162B (en) | Power grid fault control method based on digital twin | |
| CN112149873A (en) | Low-voltage transformer area line loss reasonable interval prediction method based on deep learning | |
| CN117134507B (en) | Online monitoring method and system for full-station capacitive equipment based on intelligent group association | |
| CN114386537A (en) | Lithium battery fault diagnosis method and device based on Catboost and electronic equipment | |
| CN118051827A (en) | Power grid fault prediction method based on deep learning | |
| CN112069666B (en) | Power grid short-term reliability evaluation method based on probabilistic power flow method | |
| CN117394529A (en) | SCADA-based auxiliary decision method and system for main distribution network loop-closing reverse power supply control conditions | |
| CN117674119A (en) | Power grid operation risk assessment method, device, computer equipment and storage medium | |
| CN116467658A (en) | Equipment fault tracing method based on Markov chain | |
| CN118017502A (en) | Digital twinning-based power distribution calculation power prediction method, system and medium | |
| CN118193954A (en) | Power distribution network abnormal data detection method and system based on edge calculation | |
| Dong et al. | Fault diagnosis and classification in photovoltaic systems using scada data | |
| ul Hassan et al. | Online static security assessment for cascading failure using stacked De-noising Auto-encoder | |
| CN115146715B (en) | Method, device, equipment and storage medium for diagnosing potential safety hazard of electricity | |
| Ren et al. | Research on causes of transmission line fault based on decision tree classification | |
| CN117540344A (en) | Distribution transformer anomaly monitoring and operation trend prediction method, equipment and medium | |
| CN114548701B (en) | Full-measurement-point-oriented coupling structure analysis and estimation process early warning method and system | |
| CN115936663A (en) | Maintenance method and device for power system | |
| CN114548762A (en) | Real-time power system cascading failure risk assessment method and system based on space-time diagram neural network |
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 |