CN116910668B - Lightning arrester fault early warning method, device, equipment and storage medium - Google Patents
Lightning arrester fault early warning method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN116910668B CN116910668B CN202311164420.9A CN202311164420A CN116910668B CN 116910668 B CN116910668 B CN 116910668B CN 202311164420 A CN202311164420 A CN 202311164420A CN 116910668 B CN116910668 B CN 116910668B
- Authority
- CN
- China
- Prior art keywords
- obtaining
- lightning arrester
- health state
- random forest
- preset
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 52
- 238000007637 random forest analysis Methods 0.000 claims abstract description 63
- 238000003066 decision tree Methods 0.000 claims abstract description 36
- 230000036541 health Effects 0.000 claims description 73
- 239000011159 matrix material Substances 0.000 claims description 55
- 238000011156 evaluation Methods 0.000 claims description 41
- 239000013598 vector Substances 0.000 claims description 27
- 238000012549 training Methods 0.000 claims description 18
- 238000004590 computer program Methods 0.000 claims description 13
- 230000008569 process Effects 0.000 claims description 12
- 230000006870 function Effects 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 9
- 230000032683 aging Effects 0.000 claims description 8
- 238000010276 construction Methods 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 7
- 230000007547 defect Effects 0.000 abstract description 5
- 238000012360 testing method Methods 0.000 description 14
- 230000009471 action Effects 0.000 description 8
- 230000007613 environmental effect Effects 0.000 description 8
- 238000012544 monitoring process Methods 0.000 description 7
- 230000008859 change Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 3
- 230000003862 health status Effects 0.000 description 3
- 230000007774 longterm Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000012706 support-vector machine Methods 0.000 description 3
- 238000011109 contamination Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- YHXISWVBGDMDLQ-UHFFFAOYSA-N moclobemide Chemical compound C1=CC(Cl)=CC=C1C(=O)NCCN1CCOCC1 YHXISWVBGDMDLQ-UHFFFAOYSA-N 0.000 description 2
- 230000000737 periodic effect Effects 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 101001095088 Homo sapiens Melanoma antigen preferentially expressed in tumors Proteins 0.000 description 1
- 102100037020 Melanoma antigen preferentially expressed in tumors Human genes 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 238000009421 internal insulation Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012067 mathematical method Methods 0.000 description 1
- 230000003449 preventive effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000003068 static effect 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
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- 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
- G01R31/003—Environmental or reliability tests
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- 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
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/01—Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
-
- 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)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Pure & Applied Mathematics (AREA)
- Computational Linguistics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Molecular Biology (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Environmental & Geological Engineering (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a lightning arrester fault early warning method, a device, equipment and a storage medium, and relates to the technical field of power systems, wherein the lightning arrester fault early warning method comprises the following steps: acquiring current key characteristic information and current weather information of a lightning arrester; fitting each piece of current key characteristic information with each piece of current weather information to obtain a plurality of pieces of fitting data; respectively inputting each fitting data into a corresponding preset long-short-period memory model to obtain corresponding predicted key feature information, and obtaining target feature quantity according to all the predicted key feature information; inputting the target characteristic quantity into a preset random forest model to obtain a predicted fault type of the lightning arrester, wherein the preset random forest model is used for predicting the fault type of the lightning arrester, and the preset random forest model is formed based on a plurality of decision trees; judging the predicted fault type, and obtaining the early warning state according to the judging result. The defect of the lightning arrester is found in time, early warning is achieved, and safety of the power system is improved.
Description
Technical Field
The invention relates to the technical field of power systems, in particular to a lightning arrester fault early warning method, device, equipment and storage medium.
Background
The lightning arrester is a powerful guarantee for safe operation of the power system, can influence the safety of the power system, and has a great influence on the economy of construction of the ultra-high voltage and ultra-high voltage systems. However, the lightning arrester has the defects of internal insulation, damp, aging of valve plates and the like during operation, so that the overvoltage resistance of the protected equipment is reduced, and personnel can be endangered and the safe operation of electric power can be influenced in severe cases.
In order to prevent the lightning arrester from causing more serious power grid accidents due to self faults, the conventional method is to periodically patrol and overhaul and carry out preventive tests. However, the protected equipment is required to be shut down in the periodic test, the test condition is not usually the real operation condition of the lightning arrester, the test result cannot fully reflect the real state of the lightning arrester during the operation, and the method of the periodic test cannot prevent faults occurring in the time between two tests. In addition, the actual working and running environment of the lightning arrester is complex, the accuracy of measured data can be greatly reduced due to the interference of environmental factors such as temperature, humidity and surface pollution, and the measured value is difficult to be used as the judgment basis of a final result, so that in most cases, the lightning arrester has a fault condition but is not found in time, and the continuity is directly or indirectly lost when the lightning arrester is struck next time.
Disclosure of Invention
The invention solves the problem of how to discover the defect of the lightning arrester in time so as to improve the safety of a power system.
In order to solve the above problems, the present invention provides a lightning arrester fault early warning method, including:
acquiring current key characteristic information and current weather information of a lightning arrester;
fitting the current key characteristic information with the current weather information to obtain a plurality of fitting data;
respectively inputting the fitting data into a corresponding preset long-short-period memory model to obtain corresponding prediction key feature information, and obtaining target feature quantity according to all the prediction key feature information, wherein the preset long-short-period memory model is used for obtaining the key feature information corresponding to the state of the lightning arrester;
inputting the target characteristic quantity into a preset random forest model to obtain a predicted fault type of the lightning arrester, wherein the preset random forest model is used for predicting the fault type of the lightning arrester, and the preset random forest model is formed based on a plurality of decision trees;
and judging the predicted fault type, and obtaining an early warning state according to a judging result.
Optionally, the construction process of the preset random forest model includes:
acquiring a historical health state of the lightning arrester, and acquiring corresponding historical key characteristic information according to the historical health state, wherein the health state comprises a normal state, an aging state, a damp state, surface pollution and other states;
constructing according to all the history key feature information corresponding to each history health state to obtain corresponding temporary feature quantity;
training an original random forest model according to all the temporary characteristic quantities to obtain corresponding temporary prediction fault types;
and carrying out loss calculation according to the temporary prediction fault type and the historical health state until the input of a loss function meets a preset condition, and taking the original random forest model after parameter adjustment as the preset random forest model.
Optionally, training the original random forest model according to all the temporary feature quantities to obtain a corresponding temporary prediction fault type, including:
obtaining corresponding health state evaluation indexes and temporary classification results according to all the temporary feature quantities and all the decision trees;
constructing a decision matrix according to all the health state evaluation indexes;
obtaining a weight matrix according to the decision matrix, wherein the weight matrix is used for representing the weight of each health state evaluation index, and constructing a voting matrix according to all temporary classification results;
and obtaining a voting result according to the weight matrix and the voting matrix, and obtaining a temporary fault type according to the voting result.
Optionally, the obtaining a weight matrix according to the decision matrix includes:
splitting the decision matrix to obtain a plurality of column vectors;
and comparing the maximum value in each column vector with a preset threshold value respectively, and obtaining a weight matrix according to a comparison result.
Optionally, comparing the maximum value in each column vector with a preset threshold value, and obtaining a weight matrix according to a comparison result, including:
when the maximum value is smaller than or equal to the preset threshold value, eliminating the corresponding decision tree;
when the maximum value is larger than the preset threshold value, obtaining an offset coefficient corresponding to each health state evaluation index according to the column vector;
and obtaining a weight coefficient corresponding to the health state evaluation index according to the offset coefficient, and constructing the weight matrix according to all the weight coefficients.
Optionally, the obtaining, according to the column vector, an offset coefficient corresponding to each health state evaluation index includes:
obtaining an offset coefficient corresponding to the health state evaluation index through a first formula according to the column vector;
the formula one is:
；
wherein M is ik Is x ik Corresponding to the offsetCoefficient, x ik The evaluation index of the kth tree on the ith type of health state is that i belongs to (1, 2 … …, N), N is the number of the health states, and x kmax Is the maximum value in the column vector.
Optionally, the obtaining the weight coefficient corresponding to the health state evaluation index according to the offset coefficient includes:
obtaining a weight coefficient of each decision tree through a second method according to all the offset coefficients;
the formula II is as follows:
；
wherein w is ik Is x ik Corresponding to the weight coefficient M ik Is x ik Corresponding to the offset coefficient, x ik The evaluation index of the kth tree on the ith type of health state is obtained.
Compared with the prior art, the lightning arrester fault early warning method has the advantages that: according to the current key characteristic information and the current weather information of the lightning arrester, the key characteristic information can comprise leakage current, resistive current and action times, the current key characteristic information and the current weather information are fitted, fitting data are predicted through a preset long-short-period memory model, and key characteristic information of a certain time point in the future is obtained, wherein the preset long-short-period memory model is used as a nonlinear model, long-term dependency relationship in sequence data can be better captured, further the predicted key characteristic information is more accurate, fault type prediction is carried out through a random forest model according to the predicted key characteristic information, corresponding warning is carried out according to the obtained predicted fault type, and the purposes of predicting the health state of the lightning arrester in advance and early warning in advance are achieved. Therefore, environmental factors are considered on the basis of on-line monitoring data, key characteristic information such as leakage current and resistive current is fitted with meteorological data, prediction is carried out through a preset long-short-period memory model, the hysteresis of the change of the leakage current and the resistive current can be considered under the influence of uncertain environmental factors, high prediction accuracy is achieved, fault type prediction is carried out on target characteristic quantity corresponding to the predicted key characteristic information through a random forest model, corresponding warning is carried out through the obtained predicted fault type, the healthy state of a lightning arrester is predicted in advance, early warning states are distributed, and the safety of a power system is improved.
In order to solve the technical problem, the invention also provides a lightning arrester fault early warning device, which comprises:
the acquisition unit is used for the current key characteristic information and the current weather information of the lightning arrester;
the fitting unit is used for fitting the current key characteristic information with the current weather information to obtain a plurality of fitting data;
the processing unit is used for respectively inputting the fitting data into a corresponding preset long-short-period memory model to obtain corresponding prediction key feature information, and obtaining target feature quantity according to all the prediction key feature information, wherein the preset long-short-period memory model is used for obtaining the key feature information corresponding to the state of the lightning arrester;
the processing unit is also used for obtaining target feature quantity according to all the predicted key feature information;
the processing unit is further used for inputting the target characteristic quantity into a preset random forest model to obtain a predicted fault type of the lightning arrester, the preset random forest model is used for predicting the fault type of the lightning arrester, and the preset random forest model is formed based on a plurality of decision trees;
and the judging unit is used for judging the predicted fault type and obtaining an early warning state according to a judging result.
The lightning arrester fault early warning device and the lightning arrester fault early warning method have the same advantages compared with the prior art, and are not described in detail herein.
In order to solve the technical problem, the invention also provides computer equipment, which comprises a memory and a processor:
the memory is used for storing a computer program;
the processor is used for realizing the lightning arrester fault early warning method when executing the computer program.
The remote sensing image segmentation model construction equipment and the lightning arrester fault early warning method have the same advantages compared with the prior art, and are not described in detail herein.
In order to solve the technical problem, the invention also provides a computer readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the lightning arrester fault early warning method is realized.
The computer readable storage medium and the lightning arrester fault early warning method have the same advantages compared with the prior art, and are not described in detail herein.
Drawings
FIG. 1 is a flow chart of a method for pre-warning a fault of an arrester in an embodiment of the present invention;
FIG. 2 is a flowchart of a method for constructing a random forest model preset in an embodiment of the present invention;
FIG. 3 is a diagram showing a construction of a fault early warning device for an arrester in an embodiment of the present invention;
fig. 4 is an internal structural diagram of a computer device in an embodiment of the present invention.
Detailed Description
In order that the above objects, features and advantages of the invention will be readily understood, a more particular description of the invention will be rendered by reference to specific embodiments thereof which are illustrated in the appended drawings.
As shown in fig. 1, in one embodiment, there is provided a lightning arrester fault early warning method, including the steps of:
step S1, current key characteristic information and current weather information of a lightning arrester are obtained;
specifically, the key characteristic information is on-line monitoring data, which can include leakage current, resistive current and action times, the weather information includes temperature, humidity, thunder, rain and haze, and the auxiliary key characteristic information, namely, the operation and maintenance data includes defect reasons and defect types, the material detection includes leakage current and resistive current, the standing account data includes jacket types, manufacturers and operational years, and the auxiliary key characteristic information can be used as auxiliary decision.
Step S2, fitting the current key characteristic information and the current weather information to obtain a plurality of fitting data;
specifically, fitting the current key feature information and the current weather information to obtain fitting data, for example, obtaining leakage current and resistive current, obtaining temperature and humidity of the current weather information, fitting the leakage current and the temperature and humidity to obtain first fitting data, and fitting the resistive current and the temperature and humidity to obtain second fitting data, wherein the data fitting is also called curve fitting, and is a representation mode of substituting the existing data into a formula through a mathematical method. For example, by obtaining discrete data such as by sampling, experimentation, etc., from which we often want a continuous function (i.e., curve) or more closely spaced discrete equations to fit to the known data, a process called fitting, which may be performed by least squares, spline interpolation, etc.
Step S3, respectively inputting the fitting data into a corresponding preset long-short-period memory model to obtain corresponding prediction key feature information, and obtaining target feature quantity according to all the prediction key feature information, wherein the preset long-short-period memory model is used for obtaining the key feature information corresponding to the state of the lightning arrester;
specifically, the key characteristic information of the lightning arrester comprises a plurality of data types, so that each data type corresponds to a preset long-period memory model, for example, leakage current is predicted, the input model is the corresponding preset long-period memory model of the leakage current, namely, the leakage current and weather information fitting data are input to the corresponding preset long-period memory model, the preset long-period memory model is a nonlinear model, sequence data are specially processed, a gate control mechanism is introduced, semantic association between long sequences can be effectively captured, and gradient disappearance or explosion phenomenon is relieved.
Specifically, for example, the construction process of the preset long-short-period memory model of the leakage current includes: acquiring historical leakage current data and corresponding meteorological data of a lightning arrester, fitting the historical leakage current and the corresponding meteorological data, and generating a corresponding time sequence data set, wherein the time sequence data set comprises a plurality of groups of sub-sequence data; inputting the subsequence data into an original long-short-term memory model for training to obtain a corresponding temporary prediction result, and carrying out loss calculation according to the temporary prediction result and the acquired actual leakage current data to obtain a loss function output; and adjusting model parameters of the original long-short-period memory model according to the loss function output until the loss function input meets the preset condition, and taking the original long-short-period memory model after parameter adjustment as a preset long-short-period memory model of leakage current. The specific prediction of the preset long-short-term memory model is a future time point, and the specific prediction can be set according to actual conditions, for example: the leakage current of the lightning arrester after 30 days is predicted, and the time for training the original long-short-period memory model is 30 days apart from the actual leakage current data and the historical leakage current data.
S4, inputting the target characteristic quantity into a preset random forest model to obtain a predicted fault type of the lightning arrester, wherein the preset random forest model is used for predicting the fault type of the lightning arrester and is formed based on a plurality of decision trees;
specifically, the predicted key feature information obtained through each preset long-short-term memory model is converted into a target feature vector, and the fault type of the target feature vector is predicted through a preset random forest model to obtain a predicted fault type; the fault type of the lightning arrester is the health state of the lightning arrester, including a normal state, an aging state, a damp state, a surface dirt state and other states.
And S5, judging the predicted fault type, and obtaining an early warning state according to a judging result.
Specifically, corresponding early warning is performed according to the predicted fault type, for example, when the predicted fault type is surface pollution, and according to auxiliary key characteristic information, for example, operational years, the early warning is determined to be several stages, and corresponding early warning lamps are turned on; specific early warning states can be set according to actual conditions, for example, three-level early warning can be set, when the health state of the lightning arrester is in a normal state, early warning is not carried out, a lamp green light is operated, when the health state of the lightning arrester is in a surface dirt state and other states, a first-level early warning state is issued, a first-level early warning lamp is lightened, when the health state of the lightning arrester is in a damp state and an aging state, a second-level early warning state is issued, the second-level early warning lamp is lightened, and corresponding treatment measures are carried out according to the early warning states.
According to the lightning arrester fault early warning method, according to the current key characteristic information and the current weather information of the lightning arrester, the key characteristic information can comprise leakage current, resistive current and action times, the key characteristic information and the current weather information are fitted, fitting data are carried out through a preset long-short-period memory model, and the key characteristic information of a certain time point in the future is obtained, wherein the preset long-short-period memory model is used as a nonlinear model, long-term dependency in sequence data can be better captured, further the predicted key characteristic information is more accurate, fault type prediction is carried out through a random forest model according to the predicted key characteristic information, corresponding warning is carried out according to the obtained predicted fault type, and accordingly the healthy state of the lightning arrester is predicted in advance and early warning is achieved. Therefore, environmental factors are considered on the basis of on-line monitoring data, key characteristic information such as leakage current and resistive current is fitted with meteorological data, prediction is carried out through a preset long-short-period memory model, the hysteresis of the change of the leakage current and the resistive current can be considered under the influence of uncertain environmental factors, high prediction accuracy is achieved, fault type prediction is carried out on target characteristic quantity corresponding to the predicted key characteristic information through a random forest model, corresponding warning is carried out through the obtained predicted fault type, the healthy state of a lightning arrester is predicted in advance, early warning states are distributed, and the safety of a power system is improved.
In some embodiments, as shown in fig. 2, the construction process of the preset random forest model includes:
a1, acquiring a historical health state of the lightning arrester, and acquiring corresponding historical key characteristic information according to the historical health state, wherein the health state comprises a normal state, an aging state, a damp state, a surface pollution and other states;
a2, constructing according to all the historical key characteristic information corresponding to each historical health state to obtain corresponding temporary characteristic quantity;
step A3, training an original random forest model according to all the temporary characteristic quantities to obtain corresponding temporary prediction fault types;
and step A4, carrying out loss calculation according to the temporary prediction fault type and the historical health state until the input of a loss function meets the preset condition, and taking the original random forest model after parameter adjustment as the preset random forest model.
Specifically, according to the historical health state of the lightning arrester, corresponding historical key characteristic information, namely leakage current, resistive current and action times, are obtained, all the historical key characteristic information corresponding to each historical health state is constructed to obtain corresponding temporary characteristic quantities, the temporary characteristic quantities are input into an original random forest model for training to obtain corresponding temporary prediction fault types, loss calculation is carried out according to the temporary prediction fault types and actual historical fault types, namely the historical health state until the input of a loss function meets preset conditions, and the original random forest model after parameter adjustment is used as a preset random forest model.
In some embodiments, in step A3, training the original random forest model according to the all temporary feature quantities to obtain a temporary predicted fault type includes:
step A31, obtaining corresponding health state evaluation indexes and temporary classification results according to all the temporary feature quantities and all the decision trees;
a32, constructing a decision matrix according to all the health state evaluation indexes;
a33, obtaining a weight matrix according to the decision matrix, wherein the weight matrix is used for representing the weight of each decision tree, and constructing a voting matrix according to all the temporary classification results;
and step A34, obtaining a voting result according to the weight matrix and the voting matrix, and obtaining a temporary fault type according to the voting result.
Specifically, the voting is a simple popular mechanism relative to the conventional random forest model, the weights of all decision trees are the same, and the strong classifier and the weak classifier occupy the same number of votes in the result due to the randomness generated by all decision trees, so that the deviation of the result is caused. The random forest model in the embodiment gives different weights to different classifiers according to the relevance difference of each decision tree to the result, so that the classification accuracy of the preset random forest model is improved.
In some embodiments, in step a33, the obtaining a weight matrix according to the decision matrix includes:
step A331, splitting the decision matrix to obtain a plurality of column vectors;
and step A332, comparing the maximum value in each column vector with a preset threshold value respectively, and obtaining a weight matrix according to the comparison result.
In some embodiments, in step a332, comparing the maximum value in each column vector with a preset threshold, and obtaining a weight matrix according to the comparison result includes:
step T1, eliminating the corresponding decision tree when the maximum value is smaller than or equal to the preset threshold value;
step T2, when the maximum value is larger than the preset threshold value, obtaining an offset coefficient corresponding to the health state evaluation index according to the column vector;
and step T3, obtaining a weight coefficient corresponding to the health state evaluation index according to the offset coefficient, and constructing the weight matrix according to all the weight coefficients.
In some embodiments, in step T2, the obtaining, according to the column vector, an offset coefficient corresponding to the health status evaluation index includes:
obtaining an offset coefficient corresponding to the health state evaluation index through a first formula according to the column vector;
the formula one is:
；
wherein M is ik Is x ik Corresponding to the offset coefficient, x ik The evaluation index of the kth tree on the ith type of health state is that i belongs to (1, 2, … … N), N is the number of the health states, and x kmax Is the maximum value in the column vector.
In some embodiments, in step T3, the obtaining the weight coefficient corresponding to the health status evaluation index according to the offset coefficient includes:
obtaining a weight coefficient of each decision tree through a second method according to all the offset coefficients;
the formula II is as follows:
；
wherein w is ik Is x ik Corresponding to the weight coefficient M ik Is x ik Corresponding to the offset coefficient, x ik The evaluation index of the kth tree on the ith type of health state is obtained.
In some preferred embodiments, the decision tree is calculated for each health state evaluation index based on 5 health states (aging state, damp state, surface contamination and other states) of the lightning arrester in the training set, and a decision matrix is constructed based on the calculated values:
A ；
wherein A is a decision matrix,indicate->Tree pair->Class (/ -)>=1, 2,3,4, 5) health status evaluation index, K is the total number of decision trees, K belongs to (1, … …, K);
will make a decision matrixSplit into->The column vectors: />Find->Maximum value->Setting a preset threshold value, and deleting the decision tree if the preset threshold value is smaller than the value; if the calculated value is larger than the value, calculating the offset coefficient of the decision tree, wherein the calculation formula is as follows:
；
wherein M is ik Is x ik Corresponding to the offset coefficient, x ik The evaluation index of the kth tree on the ith type of health state is that i belongs to (1, 2,3,4, 5), N is the number of the health states 5, x kmax Is the maximum value in the column vector;
the offset coefficient is used for reflecting the classification state of each decision tree andthe larger the deviation is, the worse the classification effect is, and the proportion of decision trees with larger deviation in the voting link is reduced.
Obtaining the weight coefficient corresponding to each health state evaluation index according to all the offset coefficients, namelyGiving weight coefficient->：
；
Wherein w is ik Is x ik Corresponding to the weight coefficient M ik Is x ik Corresponding to the offset coefficient, x ik The evaluation index of the kth tree on the ith type of health state is obtained.
Forming a weight matrix of the decision tree according to the ownership weight coefficient：
；
Respectively inputting all temporary characteristic quantities into each decision tree to obtain a corresponding temporary classification result, if the classification result of the decision tree is thatThere is->The voting result per tree is as follows:
；
wherein E is k (. Cndot.) is the operation of the decision tree,representing the input of training set X in a decision tree tr Output result is i, v ki Is a voting result;
according to all v ki Generating a voting matrix V:
；
finally, calculating the voting result of the ith class through the voting matrix V and the weight matrix W:
；
wherein K is the total number of decision trees;
finally, the identification result with the largest number of tickets is obtained and is used as the i-th health state of the lightning arrester, namely the predicted fault type:
；
specifically, the main two parts of the lightning arrester fault early warning method are prediction and fault classification, the method is realized through a long-short-period memory model and an improved random forest model (preset random forest model) respectively, the future change of the lightning arrester is predicted through key feature quantity of the lightning arrester to obtain a prediction result, and the specific fault type of the lightning arrester is predicted through the improved random forest model according to the obtained prediction result, so that the health state of equipment in a future time period is predicted, and early warning are realized.
According to the lightning arrester fault early warning method, according to the current key characteristic information and the current weather information of the lightning arrester, the key characteristic information can comprise leakage current, resistive current and action times, the key characteristic information and the current weather information are fitted, fitting data are carried out through a preset long-short-period memory model, and the key characteristic information of a certain time point in the future is obtained, wherein the preset long-short-period memory model is used as a nonlinear model, long-term dependency in sequence data can be better captured, further the predicted key characteristic information is more accurate, fault type prediction is carried out through a random forest model according to the predicted key characteristic information, corresponding warning is carried out according to the obtained predicted fault type, and accordingly the healthy state of the lightning arrester is predicted in advance and early warning is achieved. Therefore, environmental factors are considered on the basis of on-line monitoring data, key characteristic information such as leakage current and resistive current is fitted with meteorological data, prediction is carried out through a preset long-short-period memory model, the hysteresis of the change of the leakage current and the resistive current can be considered under the influence of uncertain environmental factors, high prediction accuracy is achieved, fault type prediction is carried out on target characteristic quantity corresponding to the predicted key characteristic information through a random forest model, corresponding warning is carried out through the obtained predicted fault type, the healthy state of a lightning arrester is predicted in advance, early warning states are distributed, and the safety of a power system is improved.
In some specific embodiments, the lightning arrester fault early warning method is experimentally verified, and the specific implementation manner is as follows:
and selecting operation monitoring data of the lightning arresters of 4 months to 7 months in 2022 as a data set, wherein the sample data are separated by 1h, and the total number of the sample data is 720. Firstly, fitting is carried out through temperature, leakage current and resistive current to obtain fitting data of the leakage current and the resistive current, then data formed by the fitting data are segmented, the first 708 items of data corresponding to a time sequence are used as training sets, and the remaining 12 items of data are used as test sets. In the constructed LSTM model, the final predicted sequence corresponding to the test set is obtained by iteratively predicting the values of 12 future time points, the same operation monitoring data is used as the data set, the ARIMA (Autoregressive Integrated Moving Average Model autoregressive integral moving average model) model is subjected to the same training and testing operation as the LSTM model, the true value and predicted value result of the corresponding test set of the lightning arrester are obtained, the average absolute error percentage (mean absolute percentage error, MAPE) and root mean square error (root mean square error, RMSE) are used as evaluation indexes, as shown in table 1,
TABLE 1 prediction accuracy for different models
The accuracy and the fitting degree of LSTM prediction are superior to ARIMA, LSTM can give consideration to the hysteresis quality of leakage current and resistive current change under the influence of uncertain environmental factors, high prediction accuracy is achieved, and the LSTM has strong adaptability in the aspect of lightning arrester monitoring variable prediction.
The experiment selects the recorded data of the lightning arresters in the transformer substation of 2018-2021, and the total number of the recorded data is 3670. The sample comprises normal, aged, wet, surface dirty and other five health states, wherein the aged sample accounts for the ratioMoisture sample ratio->Surface contamination sample ratio ∈>Other sample ratio->。
Randomly dividing normalized data, taking 80% data as a training set, taking 20% data as a testing set, and training and testing an original random forest model. Based on the test set data, the preset random forest model obtained by training is tested by using indexes such as overall accuracy, recall ratio, precision ratio and the like, and the prediction results of all insulation states are shown in a table 2,
table 2 classification result accuracy of preset random forest model
It can be seen that the "normal" recall and precision are highest, and the precision and recall of other fault types are over 78%, and the model has higher predictive power for each fault type. In addition, the average classification accuracy of the model was 92.6%.
In order to further verify the superiority of the random forest model, the model is compared with the model constructed by the decision tree and the support vector machine, the comparison process is repeated 10 times, 80% of data is randomly extracted from the original data each time to serve as training samples, the training samples are used for constructing the random forest model, the decision tree and the support vector machine model, the rest 20% of data are used as test samples, and the prediction accuracy of different models is calculated. The overall accuracy averages of the modified random forest model (random forest model), the decision tree model and the support vector machine model were 92.6%,85.23% and 84.75%, respectively. Therefore, the generalization capability of the random forest model is better, and the prediction accuracy is higher.
As shown in fig. 3, another embodiment of the present invention provides a lightning arrester fault early warning device, including:
the acquisition unit is used for the current key characteristic information and the current weather information of the lightning arrester;
the fitting unit is used for fitting the current key characteristic information with the current weather information to obtain a plurality of fitting data;
the processing unit is used for respectively inputting the fitting data into a corresponding preset long-short-period memory model to obtain corresponding prediction key feature information, and obtaining target feature quantity according to all the prediction key feature information, wherein the preset long-short-period memory model is used for obtaining the key feature information corresponding to the state of the lightning arrester;
the processing unit is also used for obtaining target feature quantity according to all the predicted key feature information;
the processing unit is further used for inputting the target characteristic quantity into a preset random forest model to obtain a predicted fault type of the lightning arrester, the preset random forest model is used for predicting the fault type of the lightning arrester, and the preset random forest model is formed based on a plurality of decision trees;
and the judging unit is used for judging the predicted fault type and obtaining an early warning state according to a judging result.
Yet another embodiment of the present invention provides a computer device comprising a memory and a processor: a memory for storing a computer program; and the processor is used for realizing the lightning arrester fault early warning method when executing the computer program.
It should be noted that the device may be a computer device such as a server, a mobile terminal, or the like.
FIG. 4 illustrates an internal block diagram of a computer device in one embodiment. The computer device includes a processor, a memory, a network interface, an input device, and a display screen connected by a system bus. The memory includes a nonvolatile storage medium and an internal memory. The non-volatile storage medium of the computer device stores an operating system, and may also store a computer program that, when executed by the processor, causes the processor to implement a lightning arrester fault warning method. The internal memory can also store a computer program which, when executed by the processor, can cause the processor to execute the lightning arrester fault early warning method. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, the input device of the computer equipment can be a touch layer covered on the display screen, can also be keys, a track ball or a touch pad arranged on the shell of the computer equipment, and can also be an external keyboard, a touch pad or a mouse and the like.
In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, which when executed by a processor implements the steps of the above-described arrester fault warning method.
Those skilled in the art will appreciate that all or part of the processes in the methods of the above embodiments may be implemented by a computer program for instructing relevant hardware, where the program may be stored in a non-volatile computer readable storage medium, and where the program, when executed, may include processes in the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the various embodiments provided herein may include non-volatile and/or volatile memory. The nonvolatile memory can include Read Only Memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link DRAM (SLDRAM), memory bus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), among others.
It should be noted that in this document, relational terms such as "first" and "second" and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The foregoing is only a specific embodiment of the invention to enable those skilled in the art to understand or practice the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Although the invention is disclosed above, the scope of the invention is not limited thereto. Various changes and modifications may be made by one skilled in the art without departing from the spirit and scope of the invention, and these changes and modifications will fall within the scope of the invention.
Claims (6)
1. The lightning arrester fault early warning method is characterized by comprising the following steps of:
acquiring current key characteristic information and current weather information of a lightning arrester;
fitting the current key characteristic information with the current weather information to obtain a plurality of fitting data;
respectively inputting the fitting data into a corresponding preset long-short-period memory model to obtain corresponding prediction key feature information, and obtaining target feature quantity according to all the prediction key feature information, wherein the preset long-short-period memory model is used for obtaining the key feature information corresponding to the state of the lightning arrester;
inputting the target characteristic quantity into a preset random forest model to obtain a predicted fault type of the lightning arrester, wherein the preset random forest model is used for predicting the fault type of the lightning arrester, the preset random forest model is formed based on a plurality of decision trees, and the construction process of the preset random forest model comprises the following steps:
acquiring a historical health state of the lightning arrester, and acquiring corresponding historical key characteristic information according to the historical health state, wherein the health state comprises a normal state, an aging state, a damp state, surface pollution and other states;
constructing according to all the history key feature information corresponding to each history health state to obtain corresponding temporary feature quantity;
training the original random forest model according to all the temporary feature quantities to obtain corresponding temporary prediction fault types, wherein the method comprises the following steps:
obtaining corresponding health state evaluation indexes and temporary classification results according to all the temporary feature quantities and all the decision trees;
constructing a decision matrix according to all the health state evaluation indexes;
obtaining a weight matrix according to the decision matrix, including:
splitting the decision matrix to obtain a plurality of column vectors;
comparing the maximum value in each column vector with a preset threshold value respectively, and obtaining a weight matrix according to a comparison result, wherein the weight matrix comprises the following components:
when the maximum value is smaller than or equal to the preset threshold value, eliminating the corresponding decision tree;
when the maximum value is larger than the preset threshold value, obtaining an offset coefficient corresponding to each health state evaluation index according to the column vector;
obtaining weight coefficients corresponding to the health state evaluation indexes according to the offset coefficients, and constructing a weight matrix according to all the weight coefficients, wherein the weight matrix is used for representing the weight of each health state evaluation index, and constructing a voting matrix according to all the temporary classification results;
obtaining a voting result according to the weight matrix and the voting matrix, and obtaining a temporary fault type according to the voting result;
performing loss calculation according to the temporary prediction fault type and the historical health state until a loss function input meets a preset condition, and taking the original random forest model after parameter adjustment as the preset random forest model;
and judging the predicted fault type, and obtaining an early warning state according to a judging result.
2. The lightning arrester fault early warning method according to claim 1, wherein the obtaining, according to the column vector, the offset coefficient corresponding to each health state evaluation index includes:
obtaining an offset coefficient corresponding to each health state evaluation index according to the column vector through formula I; the formula one is:
；
wherein M is ik Is x ik Corresponding to the offset coefficient, x ik The evaluation index of the kth tree on the ith type of health state is that i belongs to (1, 2, … … N), N is the number of the health states, and x kmax Is the maximum value in the column vector.
3. The lightning arrester fault early warning method according to claim 1, wherein the obtaining the weight coefficient corresponding to the health state evaluation index according to the offset coefficient includes:
obtaining a weight coefficient corresponding to each health state evaluation index through a second method according to all the offset coefficients; the formula II is as follows:
；
wherein w is ik Is x ik Corresponding to the weight coefficient M ik Is x ik Corresponding to the offset coefficient, x ik The evaluation index of the kth tree on the ith type of health state is obtained.
4. A lightning arrester fault early warning device, comprising:
the acquisition unit is used for the current key characteristic information and the current weather information of the lightning arrester;
the fitting unit is used for fitting the current key characteristic information with the current weather information to obtain a plurality of fitting data;
the processing unit is used for respectively inputting the fitting data into a corresponding preset long-short-period memory model to obtain corresponding prediction key feature information, and obtaining target feature quantity according to all the prediction key feature information, wherein the preset long-short-period memory model is used for obtaining the key feature information corresponding to the state of the lightning arrester;
the processing unit is also used for obtaining target feature quantity according to all the predicted key feature information;
the processing unit is further configured to input the target feature value into a preset random forest model to obtain a predicted fault type of the lightning arrester, the preset random forest model is used for predicting the fault type of the lightning arrester, the preset random forest model is formed based on a plurality of decision trees, and a construction process of the preset random forest model includes: acquiring a historical health state of the lightning arrester, and acquiring corresponding historical key characteristic information according to the historical health state, wherein the health state comprises a normal state, an aging state, a damp state, surface pollution and other states; constructing according to all the history key feature information corresponding to each history health state to obtain corresponding temporary feature quantity; training the original random forest model according to all the temporary feature quantities to obtain corresponding temporary prediction fault types, wherein the method comprises the following steps: obtaining corresponding health state evaluation indexes and temporary classification results according to all the temporary feature quantities and all the decision trees; constructing a decision matrix according to all the health state evaluation indexes; obtaining a weight matrix according to the decision matrix, including: splitting the decision matrix to obtain a plurality of column vectors; comparing the maximum value in each column vector with a preset threshold value respectively, and obtaining a weight matrix according to a comparison result, wherein the weight matrix comprises the following components: when the maximum value is smaller than or equal to the preset threshold value, eliminating the corresponding decision tree; when the maximum value is larger than the preset threshold value, obtaining an offset coefficient corresponding to each health state evaluation index according to the column vector; obtaining weight coefficients corresponding to the health state evaluation indexes according to the offset coefficients, and constructing a weight matrix according to all the weight coefficients, wherein the weight matrix is used for representing the weight of each health state evaluation index, and constructing a voting matrix according to all the temporary classification results; obtaining a voting result according to the weight matrix and the voting matrix, and obtaining a temporary fault type according to the voting result; performing loss calculation according to the temporary prediction fault type and the historical health state until a loss function input meets a preset condition, and taking the original random forest model after parameter adjustment as the preset random forest model;
and the judging unit is used for judging the predicted fault type and obtaining an early warning state according to a judging result.
5. A computer device comprising a memory and a processor:
the memory is used for storing a computer program;
the processor is configured to implement the lightning arrester fault warning method according to any one of claims 1 to 3 when executing the computer program.
6. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program which, when read and run by a processor, implements the lightning arrester fault warning method according to any one of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311164420.9A CN116910668B (en) | 2023-09-11 | 2023-09-11 | Lightning arrester fault early warning method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311164420.9A CN116910668B (en) | 2023-09-11 | 2023-09-11 | Lightning arrester fault early warning method, device, equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116910668A CN116910668A (en) | 2023-10-20 |
| CN116910668B true CN116910668B (en) | 2024-04-02 |
Family
ID=88351444
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311164420.9A Active CN116910668B (en) | 2023-09-11 | 2023-09-11 | Lightning arrester fault early warning method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116910668B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020029327A1 (en) * | 2018-08-10 | 2020-02-13 | 江南大学 | Photovoltaic array fault diagnosis method based on improved random forest algorithm |
| CN111260249A (en) * | 2020-02-13 | 2020-06-09 | 武汉大学 | Electric power communication service reliability assessment and prediction method and device based on LSTM and random forest mixed model |
| CN111860359A (en) * | 2020-07-23 | 2020-10-30 | 江苏食品药品职业技术学院 | Point cloud classification method based on improved random forest algorithm |
| CN111985571A (en) * | 2020-08-26 | 2020-11-24 | 国网湖南省电力有限公司 | Low-voltage intelligent monitoring terminal fault prediction method, device, medium and equipment based on improved random forest algorithm |
| CN112951311A (en) * | 2021-04-16 | 2021-06-11 | 中国民航大学 | Hard disk fault prediction method and system based on variable weight random forest |
| CN112989976A (en) * | 2021-03-03 | 2021-06-18 | 南京航空航天大学 | Equipment failure mode prediction method based on double-depth learning model |
| CN113792495A (en) * | 2021-11-15 | 2021-12-14 | 广东电网有限责任公司揭阳供电局 | Lightning arrester operation state identification method and device based on online monitoring data |
| CN115561596A (en) * | 2022-10-17 | 2023-01-03 | 国网山西省电力公司物资分公司 | Lightning arrester insulation state assessment method based on random forest |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020159706A1 (en) * | 2019-01-28 | 2020-08-06 | Exxonmobil Research And Engineering Company | Methods and systems for fault detection and identification |
-
2023
- 2023-09-11 CN CN202311164420.9A patent/CN116910668B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020029327A1 (en) * | 2018-08-10 | 2020-02-13 | 江南大学 | Photovoltaic array fault diagnosis method based on improved random forest algorithm |
| CN111260249A (en) * | 2020-02-13 | 2020-06-09 | 武汉大学 | Electric power communication service reliability assessment and prediction method and device based on LSTM and random forest mixed model |
| CN111860359A (en) * | 2020-07-23 | 2020-10-30 | 江苏食品药品职业技术学院 | Point cloud classification method based on improved random forest algorithm |
| CN111985571A (en) * | 2020-08-26 | 2020-11-24 | 国网湖南省电力有限公司 | Low-voltage intelligent monitoring terminal fault prediction method, device, medium and equipment based on improved random forest algorithm |
| CN112989976A (en) * | 2021-03-03 | 2021-06-18 | 南京航空航天大学 | Equipment failure mode prediction method based on double-depth learning model |
| CN112951311A (en) * | 2021-04-16 | 2021-06-11 | 中国民航大学 | Hard disk fault prediction method and system based on variable weight random forest |
| CN113792495A (en) * | 2021-11-15 | 2021-12-14 | 广东电网有限责任公司揭阳供电局 | Lightning arrester operation state identification method and device based on online monitoring data |
| CN115561596A (en) * | 2022-10-17 | 2023-01-03 | 国网山西省电力公司物资分公司 | Lightning arrester insulation state assessment method based on random forest |
Non-Patent Citations (2)
| Title |
|---|
| Long Short-Term Memory Network for Predicting Wind-Induced Vibration Response of Lightning Rod Structures;Guifeng Zhao等;buildings;第13卷;1-22 * |
| 典型软件的故障仿真和预测方法;葛志等;计算机应用研究;第37卷(第S1期);230-236 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116910668A (en) | 2023-10-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106908674B (en) | Transformer state evaluation method based on multi-state quantity prediction | |
| WO2023083366A1 (en) | Method and apparatus for identifying running state of lightning arrester on basis of on-line monitoring data | |
| CN110912867A (en) | Intrusion detection method, device, equipment and storage medium for industrial control system | |
| CN109116444B (en) | PCA-kNN-based air quality model PM2.5Forecasting method | |
| CN106650037A (en) | State diagnosis method of lightning arrester based on support vector machine regression | |
| CN115936262B (en) | Yield prediction method, system and medium based on big data environment interference | |
| CN112598248A (en) | Load prediction method, load prediction device, computer equipment and storage medium | |
| CN112712209A (en) | Reservoir warehousing flow prediction method and device, computer equipment and storage medium | |
| Wang et al. | Traffic incident duration prediction based on partial least squares regression | |
| CN116128141B (en) | Storm surge prediction method and device, storage medium and electronic equipment | |
| CN117828413A (en) | Transformer oil temperature prediction method and system based on LSTM neural network | |
| Califano et al. | Novel risk assessment tools for the climate-induced mechanical decay of wooden structures: Empirical and machine learning approaches | |
| CN116910668B (en) | Lightning arrester fault early warning method, device, equipment and storage medium | |
| CN113986999A (en) | Thunder and lightning early warning method, early warning device, electronic equipment and computer storage medium | |
| CN111832653B (en) | XGboost algorithm-based arrester defect diagnosis method and system | |
| CN117251788A (en) | State evaluation method, device, terminal equipment and storage medium | |
| CN117114161A (en) | Method for predicting wind deflection flashover risk of power transmission line based on meta-learning | |
| Jones et al. | Closed loop statistical performance analysis of NK knock controllers | |
| Ahmed et al. | Extreme value statistics in semi-supervised models | |
| Lakhno et al. | Development of an intelligent subsystem for operating system incidents forecasting | |
| CN111754103A (en) | Enterprise risk image method, device, computer equipment and readable storage medium | |
| CN117478394B (en) | Network security analysis method, system, computer equipment and computer readable storage medium based on digital twin | |
| CN116633026B (en) | Clean energy perception monitoring method and device, computer equipment and storage medium | |
| CN118376865A (en) | Rapid detection method and system for direct current characteristics of lightning arrester | |
| CN116108974A (en) | Demand response baseline load prediction method and device considering meteorological factors |
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 |