CN113466630A - RSSPN model-based power distribution network fault reason classification method - Google Patents

RSSPN model-based power distribution network fault reason classification method Download PDF

Info

Publication number
CN113466630A
CN113466630A CN202110811543.1A CN202110811543A CN113466630A CN 113466630 A CN113466630 A CN 113466630A CN 202110811543 A CN202110811543 A CN 202110811543A CN 113466630 A CN113466630 A CN 113466630A
Authority
CN
China
Prior art keywords
model
rsspn
waveform data
distribution network
power distribution
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.)
Granted
Application number
CN202110811543.1A
Other languages
Chinese (zh)
Other versions
CN113466630B (en
Inventor
杨帆
方健
莫文雄
王勇
张敏
刘振东
陈创升
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangzhou Power Supply Bureau of Guangdong Power Grid Co Ltd
Original Assignee
Guangzhou Power Supply Bureau of Guangdong Power Grid Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Guangzhou Power Supply Bureau of Guangdong Power Grid Co Ltd filed Critical Guangzhou Power Supply Bureau of Guangdong Power Grid Co Ltd
Priority to CN202110811543.1A priority Critical patent/CN113466630B/en
Publication of CN113466630A publication Critical patent/CN113466630A/en
Application granted granted Critical
Publication of CN113466630B publication Critical patent/CN113466630B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/08—Locating faults in cables, transmission lines, or networks
    • G01R31/088—Aspects of digital computing
    • 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
    • Y04S10/52—Outage or fault management, e.g. fault detection or location

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Supply And Distribution Of Alternating Current (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The invention discloses a power distribution network fault reason classification method based on an RSSPN model, which comprises the steps of obtaining fault current and voltage waveform data by using a distributed digital fault recorder, and carrying out class marking on the fault current and voltage waveform data; training an RSSPN model by using unlabeled waveform data and labeled waveform data to obtain a learned model function and a decision radius function; fine-tuning the trained RSSPN model to generate a new classifier; the classification of the fault reasons of the power distribution network is completed through a new classifier; according to the invention, information is mined from unmarked data by utilizing a prototype network structure and semi-supervised learning, so that the accuracy of distribution network fault reason classification is improved.

Description

RSSPN model-based power distribution network fault reason classification method
Technical Field
The invention relates to the technical field of power distribution network fault analysis, in particular to a power distribution network fault reason classification method based on an RSSPN model.
Background
The distribution network is the final stage of delivering power from the distribution system to individual consumers, and its safe and stable operation is directly related to the interests of each consumer.
However, fault cause classification in power distribution networks is a challenging problem. First, the mechanism is not clear, and there is no rapid and accurate manual classification method at present. Second, due to the lack of labeled data (manually labeled error waveforms with root cause of failure), it is difficult for the machine learning method to achieve high accuracy and good generalization performance. Third, power distribution networks are becoming more complex due to the continuous access of low current grounding systems and distributed power sources in the power distribution networks, such that traditional classifiers (e.g., multi-layer perceptrons and convolutional neural networks) cannot achieve high-precision classification.
Disclosure of Invention
This section is for the purpose of summarizing some aspects of embodiments of the invention and to briefly introduce some preferred embodiments. In this section, as well as in the abstract and the title of the invention of this application, simplifications or omissions may be made to avoid obscuring the purpose of the section, the abstract and the title, and such simplifications or omissions are not intended to limit the scope of the invention.
The present invention has been made in view of the above-mentioned conventional problems.
Therefore, the invention provides a distribution network fault reason classification method based on an RSSPN model, which can solve the problem that high-precision classification cannot be realized due to insufficient labeled data.
In order to solve the technical problems, the invention provides the following technical scheme: comprises the following steps: acquiring fault current and voltage waveform data by using a distributed digital fault recorder, and carrying out category marking on the fault current and voltage waveform data; training an RSSPN model by using unlabeled waveform data and labeled waveform data to obtain a learned model function and a decision radius function; fine-tuning the trained RSSPN model to generate a new classifier; and finishing the classification of the fault reasons of the power distribution network through the new classifier.
As an optimal scheme of the RSSPN model-based power distribution network fault cause classification method of the present invention, wherein: the method is characterized in that: further comprising that the quantity ratio of the waveform data of the unmarked type to the waveform data of the marked type is 4: 1.
as an optimal scheme of the RSSPN model-based power distribution network fault cause classification method of the present invention, wherein: further comprising training the RSSPN model using a mixing loss function; the hybrid loss function is composed of a supervised loss function, a pseudo-supervised loss function and an unsupervised loss function.
As an optimal scheme of the RSSPN model-based power distribution network fault cause classification method of the present invention, wherein: the supervisory loss function includes at least one of,
Figure BDA0003168414660000021
wherein L issFor monitoring the loss value, T is a preset total number of training iterations,
Figure BDA0003168414660000022
for the ith output value of the RSSPN model input layer,
Figure BDA0003168414660000023
is the ith input value, k, of the input layer of the RSSPN modelcIs the model function.
As an optimal scheme of the RSSPN model-based power distribution network fault cause classification method of the present invention, wherein: further comprising, said pseudo-supervised loss function comprises,
Figure BDA0003168414660000024
wherein L isPIn order to be a pseudo-supervised loss value,
Figure BDA0003168414660000025
for the jth waveform data of the labeled class,
Figure BDA0003168414660000026
is the unlabeled jth waveform data.
As an optimal scheme of the RSSPN model-based power distribution network fault cause classification method of the present invention, wherein: further comprising, the unsupervised loss function comprises,
Lu＝-logpw(y＝l|x)
wherein L isuTo monitor the loss value, pw(y ═ l | x) is the probability of each of the labeled categories of waveform data, y is the classification label, x is the vector representation of the labeled categories of waveform data, and l is the number of labeled categories of waveform data.
As an optimal scheme of the RSSPN model-based power distribution network fault cause classification method of the present invention, wherein: the model function may include a function of the model,
Figure BDA0003168414660000027
wherein c is the fault cause category of the power distribution network, h (x)i) Is a vector of unlabeled waveform data, h (x)j) Is a vector of waveform data of the labeled class, zi,cIs the ith category in the category c; z is a radical ofj,cIs the jth category in category c.
As an optimal scheme of the RSSPN model-based power distribution network fault cause classification method of the present invention, wherein: the decision radius function comprises that assuming that only the unlabeled waveform data belongs to known fault categories in the decision radius, calculating the decision radius by using the decision radius function, and updating the RSSPN model with the unlabeled waveform data falling in the decision radius; the decision radius function is as follows:
rc＝MLP([min(dj,c),max(dj,c),mean(dj,c),var(dj,c),skew(dj,c),kurt(dj,c)])
wherein r iscIs the decision radius, dj,cIs the distance of the unlabeled waveform data from class c.
As an optimal scheme of the RSSPN model-based power distribution network fault cause classification method of the present invention, wherein: the fine adjustment comprises the steps of loading the RSSPN model which is trained through a Caffe framework; deleting the fully connected layers in the RSSPN model after training is finished, and keeping the weight of the rest layers unchanged; and constructing a group of new full connection layers, and initializing the weights of the new full connection layers to obtain the new classifier.
The invention has the beneficial effects that: according to the invention, information is mined from unmarked data by utilizing a prototype network structure and semi-supervised learning, so that the accuracy of distribution network fault reason classification is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive exercise. Wherein:
fig. 1 is a schematic diagram of a newly-built full-connection layer fc8 of a power distribution network fault cause classification method based on an RSSPN model according to a first embodiment of the present invention.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, specific embodiments accompanied with figures are described in detail below, and it is apparent that the described embodiments are a part of the embodiments of the present invention, not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making creative efforts based on the embodiments of the present invention, shall fall within the protection scope of the present invention.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways than those specifically described and will be readily apparent to those of ordinary skill in the art without departing from the spirit of the present invention, and therefore the present invention is not limited to the specific embodiments disclosed below.
Furthermore, reference herein to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one implementation of the invention. The appearances of the phrase "in one embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments.
The present invention will be described in detail with reference to the drawings, wherein the cross-sectional views illustrating the structure of the device are not enlarged partially in general scale for convenience of illustration, and the drawings are only exemplary and should not be construed as limiting the scope of the present invention. In addition, the three-dimensional dimensions of length, width and depth should be included in the actual fabrication.
Meanwhile, in the description of the present invention, it should be noted that the terms "upper, lower, inner and outer" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, and are only for convenience of describing the present invention and simplifying the description, but do not indicate or imply that the referred device or element must have a specific orientation, be constructed in a specific orientation and operate, and thus, cannot be construed as limiting the present invention. Furthermore, the terms first, second, or third are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
The terms "mounted, connected and connected" in the present invention are to be understood broadly, unless otherwise explicitly specified or limited, for example: can be fixedly connected, detachably connected or integrally connected; they may be mechanically, electrically, or directly connected, or indirectly connected through intervening media, or may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
Example 1
Referring to fig. 1, a first embodiment of the present invention provides a method for classifying power distribution network fault causes based on RSSPN model, including:
s1: and acquiring fault current and voltage waveform data by using a distributed digital fault recorder, and carrying out category marking on the fault current and voltage waveform data.
The current comprises three-phase current and neutral point current, and the voltage is three-phase voltage.
S2: the RSSPN model is trained using unlabeled waveform data and labeled class waveform data and a learned model function is obtained as well as a decision radius function.
The RSSPN (road Semi-super powered type Network) model is trained using a large amount of unlabeled waveform data and a small amount of labeled class waveform data, where the ratio of the number of unlabeled waveform data to the number of labeled class waveform data is 4: 1.
specifically, the RSSPN model is trained by using the mixed loss function, so as to obtain a model function kcAnd a decision radius function rc(ii) a Wherein the mixed loss function is composed of supervision loss, pseudo supervision loss and unsupervised loss.
(1) Supervision loss function:
Figure BDA0003168414660000051
wherein L issFor monitoring the loss value, T is a preset total number of training iterations,
Figure BDA0003168414660000052
for the ith output value of the RSSPN model input layer,
Figure BDA0003168414660000053
is the ith input value, k, of the input layer of the RSSPN modelcIs a model function.
(2) A function of the pseudo-supervised loss is used,
Figure BDA0003168414660000054
wherein L isPIn order to be a pseudo-supervised loss value,
Figure BDA0003168414660000055
for the jth waveform data of the labeled class,
Figure BDA0003168414660000056
is the unlabeled jth waveform data.
(3) The function of the unsupervised loss,
Lu＝-logpw(y＝l|x)
wherein L isuTo monitor the loss value, pw(y ═ l | x) is the probability of each of the labeled categories of waveform data, y is the classification label, x is the vector representation of the labeled categories of waveform data, and l is the number of labeled categories of waveform data.
Model function:
Figure BDA0003168414660000057
wherein c is the fault cause category of the power distribution network, h (x)i) Is a vector of unlabeled waveform data, h (x)j) Is a vector of waveform data of the labeled class, zi,cIs the ith category in the category c; z is a radical ofj,cIs the jth category in category c.
Assuming that only the unmarked waveform data belongs to known fault categories in the decision radius, calculating the decision radius by using a decision radius function, and updating the unmarked waveform data in the decision radius into an RSSPN model; the decision radius function is as follows:
rc＝MLP([min(dj,c),max(dj,c),mean(dj,c),var(dj,c),skew(dj,c),kurt(dj,c)])
wherein r iscTo decide the radius, dj,cOf unlabeled waveform data and class cA distance; MLP (Multi-layer perceptron) represents a Multi-layer perceptron.
S3: and fine-tuning the trained RSSPN model to generate a new classifier.
The fine tuning steps are as follows:
(1) loading the RSSPN model which is trained through a Caffe framework;
caffe, called the comprehensive Architecture for Fast Feature Embedding, is a deep learning framework with expressiveness, speed and thinking modularity.
(2) Deleting the fully connected layers in the RSSPN model after training is finished, and keeping the weight of the rest layers unchanged;
(3) and constructing a group of new full connection layers, initializing the weights of the new full connection layers, and obtaining a new classifier.
Referring to fig. 1, a fully-connected layer fc8 is newly created, the setting step size is 2000, the fine-tuning learning rate is 0.01, the maximum iteration number is 5000, and the GPU is used to initialize the weight of the new fully-connected layer.
S4: and finishing the classification of the fault reasons of the power distribution network through a new classifier.
Example 2
In order to verify and explain the technical effects adopted in the method, the embodiment selects the SVM, the prototype network and the convolutional neural network and adopts the method to perform comparison test, and compares the test results by means of scientific demonstration to verify the real effect of the method.
With the increase of samples, the classification precision obtained by the SVM and the convolutional neural network is continuously reduced; prototype networks, while capable of achieving higher accuracy, are unable to distinguish new classes from old ones.
In order to verify that the method has higher classification precision for a large number of samples relative to the SVM, the prototype network and the convolutional neural network, the SVM, the prototype network, the convolutional neural network and the method are adopted to classify and compare fault waveform data (the number of unmarked data is 800, and the number of marked data is 200) collected by a power grid respectively, and the fault reasons of the power distribution network include type A (lightning arrester fault, burning of a flexible cable of a transformer and damage of an elbow of the transformer) and type B (dry land, wet land, dry concrete, wet concrete and asphalt material).
And (3) testing environment: fault waveform data are input into the SVM, the prototype network, the convolutional neural network and the new classifier constructed by the method by utilizing matlab, and the experimental result is as follows:
table 1: and (5) comparing classification results.
Figure BDA0003168414660000061
Figure BDA0003168414660000071
Table 2: the method has the advantages that the classification accuracy is high under the condition that different quantities of labeled data are different.
Number of unmarked data Class A (not fine tuned) Class A (Fine tuning) Class B (not fine tuned) Class B (trimming)
0 78.61％ 85.14％ 62.45％ 72.48％
200 86.45％ 91.49％ 68.17％ 79.14％
400 90.62％ 96.78％ 76.82％ 88.60％
600 92.12％ 98.63％ 79.15％ 94.58％
800 93.46％ 99.45％ 84.94％ 95.67％
As can be seen from Table 1, the performance of the method is far superior to that of the traditional classifier and neural network; referring to table 2, it can be seen that the classification accuracy gradually increases with the increase of the number of unlabeled data, proving that the method effectively improves the classification accuracy.
It should be noted that the above-mentioned embodiments are only for illustrating the technical solutions of the present invention and not for limiting, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, which should be covered by the claims of the present invention.

Claims (9)

1. A power distribution network fault reason classification method based on an RSSPN model is characterized by comprising the following steps: comprises the steps of (a) preparing a mixture of a plurality of raw materials,
acquiring fault current and voltage waveform data by using a distributed digital fault recorder, and carrying out category marking on the fault current and voltage waveform data;
training an RSSPN model by using unlabeled waveform data and labeled waveform data to obtain a learned model function and a decision radius function;
fine-tuning the trained RSSPN model to generate a new classifier;
and finishing the classification of the fault reasons of the power distribution network through the new classifier.
2. The RSSPN model-based power distribution network fault cause classification method of claim 1, wherein: also comprises the following steps of (1) preparing,
the quantity ratio of the waveform data of the unlabeled class to the waveform data of the labeled class is 4: 1.
3. the RSSPN model-based power distribution network fault cause classification method of claim 1, wherein: further comprising training the RSSPN model using a mixing loss function; the hybrid loss function is composed of a supervised loss function, a pseudo-supervised loss function and an unsupervised loss function.
4. The RSSPN model-based power distribution network fault cause classification method of claim 3, wherein: the supervisory loss function includes at least one of,
Figure FDA0003168414650000011
wherein L issFor monitoring the loss value, T is a preset total number of training iterations,
Figure FDA0003168414650000012
is RSSPThe ith output value of the N model input layer,
Figure FDA0003168414650000013
is the ith input value, k, of the input layer of the RSSPN modelcIs the model function.
5. The RSSPN model-based power distribution network fault cause classification method of claim 3, wherein: further comprising, said pseudo-supervised loss function comprises,
Figure FDA0003168414650000014
wherein L isPIn order to be a pseudo-supervised loss value,
Figure FDA0003168414650000015
for the jth waveform data of the labeled class,
Figure FDA0003168414650000016
is the unlabeled jth waveform data.
6. The RSSPN model-based power distribution network fault cause classification method of claim 3, wherein: further comprising, the unsupervised loss function comprises,
Lu＝-log pw(y＝l|x)
wherein L isuTo monitor the loss value, pw(y ═ l | x) is the probability of each of the labeled categories of waveform data, y is the classification label, x is the vector representation of the labeled categories of waveform data, and l is the number of labeled categories of waveform data.
7. The RSSPN model-based power distribution network fault cause classification method according to claim 1 or 4, wherein: the model function may include a function of the model,
Figure FDA0003168414650000021
wherein c is the fault cause category of the power distribution network, h (x)i) Is a vector of unlabeled waveform data, h (x)j) Is a vector of waveform data of the labeled class, zi,cIs the ith category in the category c; z is a radical ofj,cIs the jth category in category c.
8. The RSSPN model-based power distribution network fault cause classification method according to claim 1 or 2, wherein: the decision radius function includes a function of the radius of the decision,
assuming that only the unlabeled waveform data belong to known fault classes within a decision radius, calculating the decision radius by using the decision radius function, and updating the RSSPN model with the unlabeled waveform data within the decision radius; the decision radius function is as follows:
rc＝MLP([min(dj,c),max(dj,c),mean(dj,c),var(dj,c),skew(dj,c),kurt(dj,c)])
wherein r iscIs the decision radius, dj,cIs the distance of the unlabeled waveform data from class c.
9. The RSSPN model-based power distribution network fault cause classification method of claim 8, wherein: the fine-tuning may include the steps of,
loading the RSSPN model which is trained through a Caffe framework;
deleting the fully connected layers in the RSSPN model after training is finished, and keeping the weight of the rest layers unchanged;
and constructing a group of new full connection layers, and initializing the weights of the new full connection layers to obtain the new classifier.
CN202110811543.1A 2021-07-19 2021-07-19 Power distribution network fault cause classification method based on RSSPN model Active CN113466630B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110811543.1A CN113466630B (en) 2021-07-19 2021-07-19 Power distribution network fault cause classification method based on RSSPN model

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110811543.1A CN113466630B (en) 2021-07-19 2021-07-19 Power distribution network fault cause classification method based on RSSPN model

Publications (2)

Publication Number Publication Date
CN113466630A true CN113466630A (en) 2021-10-01
CN113466630B CN113466630B (en) 2023-10-24

Family

ID=77880936

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110811543.1A Active CN113466630B (en) 2021-07-19 2021-07-19 Power distribution network fault cause classification method based on RSSPN model

Country Status (1)

Country Link
CN (1) CN113466630B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114089108A (en) * 2021-11-17 2022-02-25 广东电网有限责任公司广州供电局 Cable sheath defect identification method, device, equipment and readable storage medium

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN207817133U (en) * 2018-01-13 2018-09-04 福州大学 Distributed power net fault wave recording device
CN209746068U (en) * 2019-01-07 2019-12-06 山东科技大学 Distribution network fault warning system based on waveform analysis
US20200134461A1 (en) * 2018-03-20 2020-04-30 Sri International Dynamic adaptation of deep neural networks
CN111160241A (en) * 2019-12-27 2020-05-15 华中科技大学 Power distribution network fault classification method, system and medium based on deep learning
CN112016473A (en) * 2020-08-31 2020-12-01 国网安徽省电力有限公司电力科学研究院 Power distribution network high-resistance grounding fault diagnosis method based on semi-supervised learning and attention mechanism
CN113050602A (en) * 2021-03-26 2021-06-29 杭州电子科技大学 Industrial process fault method based on robust semi-supervised discriminant analysis

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN207817133U (en) * 2018-01-13 2018-09-04 福州大学 Distributed power net fault wave recording device
US20200134461A1 (en) * 2018-03-20 2020-04-30 Sri International Dynamic adaptation of deep neural networks
CN209746068U (en) * 2019-01-07 2019-12-06 山东科技大学 Distribution network fault warning system based on waveform analysis
CN111160241A (en) * 2019-12-27 2020-05-15 华中科技大学 Power distribution network fault classification method, system and medium based on deep learning
CN112016473A (en) * 2020-08-31 2020-12-01 国网安徽省电力有限公司电力科学研究院 Power distribution network high-resistance grounding fault diagnosis method based on semi-supervised learning and attention mechanism
CN113050602A (en) * 2021-03-26 2021-06-29 杭州电子科技大学 Industrial process fault method based on robust semi-supervised discriminant analysis

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
MENGYE REN 等: "META-LEARNING FOR SEMI-SUPERVISED FEW-SHOT CLASSIFICATION", 《ICLR 2018》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114089108A (en) * 2021-11-17 2022-02-25 广东电网有限责任公司广州供电局 Cable sheath defect identification method, device, equipment and readable storage medium
CN114089108B (en) * 2021-11-17 2022-08-02 广东电网有限责任公司广州供电局 Cable sheath defect identification method, device, equipment and readable storage medium

Also Published As

Publication number Publication date
CN113466630B (en) 2023-10-24

Similar Documents

Publication Publication Date Title
CN111476435B (en) Charging pile load prediction method based on density peak value
CN112039687A (en) Small sample feature-oriented fault diagnosis method based on improved generation countermeasure network
CN103324980A (en) Wind power station wind speed prediction method
CN113011084B (en) Intelligent identification method for power transmission line fault reasons based on correlation vector machine
CN108336739A (en) A kind of Probabilistic Load Flow on-line calculation method based on RBF neural
CN105471647B (en) A kind of power communication network fault positioning method
US20240014651A1 (en) Probability estimation method for photovoltaic power based on optimized copula function and photovoltaic power system
CN110738232A (en) grid voltage out-of-limit cause diagnosis method based on data mining technology
CN111652478B (en) Umbrella algorithm-based power system voltage stability evaluation misclassification constraint method
CN112700121A (en) Distribution network fault risk level prediction method based on random forest
CN107527071A (en) A kind of sorting technique and device that k nearest neighbor is obscured based on flower pollination algorithm optimization
CN111091141B (en) Photovoltaic backboard fault diagnosis method based on layered Softmax
CN113379116A (en) Cluster and convolutional neural network-based line loss prediction method for transformer area
CN110991689B (en) Distributed photovoltaic power generation system short-term prediction method based on LSTM-Morlet model
CN113466630A (en) RSSPN model-based power distribution network fault reason classification method
CN113887729A (en) Fault diagnosis method for low-voltage power line carrier communication system
CN112052544A (en) Wind power plant current collection network design method and system, storage medium and computing device
Syu et al. Ultra-short-term wind speed forecasting for wind power based on gated recurrent unit
CN111080001A (en) Deep neural network prediction method applied to wind speed of wind power plant
CN105160598A (en) Power grid service classification method based on improved EM algorithm
CN117077546B (en) Power system load modeling method and system based on data driving
CN112686404B (en) Power distribution network fault first-aid repair-based collaborative optimization method
CN113627674A (en) Distributed photovoltaic power station output prediction method and device and storage medium
Wang et al. Complex fault source identification method for high-voltage trip-offs of wind farms based on SU-MRMR and PSO-SVM
CN108038518A (en) A kind of photovoltaic generation power based on meteorological data determines method and system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant