CN116227538B - Clustering and deep learning-based low-current ground fault line selection method and equipment - Google Patents
Clustering and deep learning-based low-current ground fault line selection method and equipment Download PDFInfo
- Publication number
- CN116227538B CN116227538B CN202310460342.0A CN202310460342A CN116227538B CN 116227538 B CN116227538 B CN 116227538B CN 202310460342 A CN202310460342 A CN 202310460342A CN 116227538 B CN116227538 B CN 116227538B
- Authority
- CN
- China
- Prior art keywords
- probability
- fault
- zero sequence
- confidence
- sequence current
- 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
- 238000010187 selection method Methods 0.000 title claims abstract description 23
- 238000013135 deep learning Methods 0.000 title claims abstract description 20
- 238000000034 method Methods 0.000 claims abstract description 44
- 238000012549 training Methods 0.000 claims abstract description 40
- 238000012360 testing method Methods 0.000 claims abstract description 8
- 238000011056 performance test Methods 0.000 claims abstract description 4
- 238000004590 computer program Methods 0.000 claims description 19
- 238000004422 calculation algorithm Methods 0.000 claims description 15
- 238000010586 diagram Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 8
- 230000001629 suppression Effects 0.000 claims description 8
- 238000009826 distribution Methods 0.000 claims description 6
- 238000003860 storage Methods 0.000 claims description 6
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 230000008859 change Effects 0.000 description 12
- 241000764238 Isis Species 0.000 description 8
- 230000006870 function Effects 0.000 description 8
- 230000001052 transient effect Effects 0.000 description 8
- 210000002569 neuron Anatomy 0.000 description 7
- 102100028952 Drebrin Human genes 0.000 description 6
- 101000838600 Homo sapiens Drebrin Proteins 0.000 description 6
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 230000004927 fusion Effects 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 230000007935 neutral effect Effects 0.000 description 3
- 238000007500 overflow downdraw method Methods 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 2
- 239000003990 capacitor Substances 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000005315 distribution function Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 210000002364 input neuron Anatomy 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 210000004205 output neuron Anatomy 0.000 description 2
- 238000003786 synthesis reaction Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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
- G01R31/50—Testing of electric apparatus, lines, cables or components for short-circuits, continuity, leakage current or incorrect line connections
- G01R31/52—Testing for short-circuits, leakage current or ground faults
-
- 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
- G01R31/50—Testing of electric apparatus, lines, cables or components for short-circuits, continuity, leakage current or incorrect line connections
- G01R31/58—Testing of lines, cables or conductors
-
- 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)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Supply And Distribution Of Alternating Current (AREA)
- Emergency Protection Circuit Devices (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the application provides a small current ground fault line selection method and equipment based on clustering and deep learning, wherein the method comprises the following steps: s10, acquiring historical data of fault electrical characteristic values of all outgoing lines of the low-current grounding system when single-phase grounding faults occur; constructing a training set and a testing set according to the historical data; s20, establishing a fault prediction model based on a deep belief network according to the fault electrical characteristic quantity; s30, training the fault prediction model through a training set to obtain a trained fault prediction model; s40, establishing a fuzzy probability interval; s50, performing performance test on the trained fault prediction model by adopting a test set, performing confidence coefficient judgment on an output result of the fault prediction model through a fuzzy probability interval, outputting an identification result of a fault line, analyzing a small-current grounding fault pulling sequence, improving the pulling accuracy of the fault line, and being suitable for the technical field of power systems.
Description
Technical Field
The application relates to the technical field of power systems, in particular to a small-current ground fault line selection method and equipment based on clustering and deep learning.
Background
The problem of single-phase grounding fault line selection of the small-current grounding system is always an important research topic in the power system, and the research pace of the single-phase grounding fault line selection method of the small-current grounding system never stops from the most original method for judging a fault line by selecting line and switching off according to manual experience to the selection and pulling strategy for judging the fault line by adopting multi-element information fusion.
Currently, the common line selection methods are mainly divided into three types: selecting lines based on fault steady-state information, selecting lines based on fault transient state information and selecting lines based on multi-information fusion;
the line selection method based on fault steady-state or transient state information generally adopts steady-state electric quantity or transient state electric quantity change when a single-phase grounding fault occurs in a low-current grounding system to identify a single-phase grounding fault line, so that the fault line is effectively judged, an auxiliary decision is provided for power grid accident handling, but a plurality of defects exist in the actual application process of the fault line only by means of single electric quantity change; when a single-phase grounding fault occurs in a small-current grounding system, the electrical quantity change after the fault is affected by the small-current grounding type, system elements, fault positions, transition resistance and the like, and accurate judgment of a fault line is difficult to realize only depending on a single electrical quantity change, so that the method for judging the fault line by adopting multi-element information fusion has wider application in practical application.
The method is based on multi-information fusion line selection, which generally adopts the electrical quantity change after various faults occur to effectively identify the fault type, and meanwhile, the information intelligent fusion method is utilized to judge the multi-information so as to accurately judge the fault line; the commonly adopted intelligent information fusion method comprises the following steps: D-S evidence theory, genetic algorithm, fuzzy control theory and the like, and comprehensively deciding and judging the multi-element fault judgment information through the intelligent information fusion method; the D-S evidence theory is most widely applied, and the probability calculation is mainly carried out by carrying out evidence synthesis through multivariate information to generate a basic trust distribution function, but the method is poor in robustness, is excessively sensitive to the evidence synthesis variation in the probability calculation and can influence the accuracy of a judgment result, and for the change of the electric quantity of the electric power system, the data has abrupt uncertainty, so that a fault line selection strategy of the small-current grounding system has higher requirements on the accuracy and the robustness.
Disclosure of Invention
In order to solve one of the technical defects, the embodiment of the application provides a small-current ground fault line selection method and equipment based on clustering and deep learning, which can analyze small-current ground fault line selection sequence bits and improve the selection accuracy of fault lines.
According to a first aspect of embodiments of the present application, there is provided a small current ground fault line selection method based on clustering and deep learning, including:
s10, acquiring historical data of fault electrical characteristic values of all outgoing lines of the low-current grounding system when single-phase grounding faults occur; constructing a training set and a testing set according to the historical data;
wherein the fault electrical characteristic quantity includes: zero sequence current amplitude, active power, reactive power and zero sequence current;
s20, establishing a fault prediction model based on a deep belief network according to the fault electrical characteristic quantity;
s30, training the fault prediction model through a training set to obtain a trained fault prediction model;
s40, establishing a fuzzy probability interval;
s50, performing performance test on the trained fault prediction model by adopting a test set, performing confidence judgment on the output result of the fault prediction model through a fuzzy probability interval, and outputting the identification result of the fault line.
According to a second aspect of embodiments of the present application, there is provided an electronic device, including: a memory; a processor; a computer program;
wherein the computer program is stored in the memory and configured to be executed by the processor to implement the method as described above.
According to a third aspect of embodiments of the present application, there is provided a computer-readable storage device, characterized in that a computer program is stored thereon; the computer program is executed by a processor to implement the method as described above.
The small current ground fault line selection method and the small current ground fault line selection equipment based on clustering and deep learning provided by the embodiment of the application have the technical effects that:
1. according to the method, the zero sequence current amplitude, reactive power and fault phase current are trained through the deep confidence network respectively, a fault line prediction model is built, the fault line is effectively judged according to the total four electrical characteristic quantities of the zero sequence current amplitude, the reactive power, the fault phase current and the active power when the low-current grounding system generates single-phase grounding faults, the confidence rate judgment is carried out on the output result of the fault prediction model by combining the fuzzy probability interval, the fault probability of the selected and pulled line is calculated, the accurate analysis and the reliable judgment of the selected and pulled sequence position of the low-current grounding faults are realized, and the practicability is extremely strong.
2. In the application, aiming at the condition that the neutral point of the low-current grounding system is grounded through the arc suppression coil, the active power is established to be used as a fault electrical characteristic quantity for prediction, so that the applicability of the application is improved, and the safe and stable operation of the power distribution network system is ensured.
3. In the method, the threshold value of the fuzzy probability interval is determined based on an ADPC clustering algorithm, the optimal clustering center is selected through a self-adaptive mechanism, the purpose of accurately judging the fault route selection prediction result according to fault characteristic data is achieved, the fuzzy probability interval is determined based on the clustering algorithm, the problem of confidence degree selection of the fault route prediction result can be effectively solved, and a reliable and reasonable decision basis is provided for fault route selection auxiliary decision; and the prediction result of the fault prediction model is subjected to line selection recognition based on the probability interval of fuzzy control, so that the auxiliary decision efficiency of small-current ground fault line selection is effectively improved, and the safe and stable operation of the power grid is ensured.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute an undue limitation to the application. In the drawings:
fig. 1 is a schematic flow chart of a small current ground fault line selection method based on clustering and deep learning according to an embodiment of the present application;
fig. 2 is a schematic flow chart of step S30 provided in the embodiment of the present application;
fig. 3 is a schematic flow chart of step S40 provided in the embodiment of the present application;
fig. 4 is a schematic flow chart of step S4011 provided in the embodiment of the present application;
fig. 5 is a schematic flow chart of a small current ground fault line selection method based on clustering and deep learning according to a second embodiment of the present application.
Description of the embodiments
In order to make the technical solutions and advantages of the embodiments of the present application more apparent, the following detailed description of exemplary embodiments of the present application is given with reference to the accompanying drawings, and it is apparent that the described embodiments are only some of the embodiments of the present application and not exhaustive of all the embodiments. It should be noted that, in the case of no conflict, the embodiments and features in the embodiments may be combined with each other.
In carrying out the present application, the inventors found that: in the research of single-phase earth fault line selection of a small-current grounding system, the extracted fault electrical characteristic quantity has complex correlation with the result of line selection, and the neural network has good performance in the aspects of representing and explaining the complex nonlinear relation. The deep confidence network (Deep Belief Network, DBN) has wide application and excellent effect in the aspect of merging multiple information to perform result recognition, probability generation is realized mainly by using a multi-RBM mode, the relation between input data and output labels is established by constructing a joint distribution function, and the neural network model has quite expansibility and strong flexibility.
Based on the method, the invention provides a small-current ground fault line selection method and equipment based on clustering and deep learning.
Examples
As shown in fig. 1, in an embodiment of the present application, a small current ground fault line selection method based on clustering and deep learning is provided, including:
s10, acquiring historical data of fault electrical characteristic values of all outgoing lines of the low-current grounding system when single-phase grounding faults occur; constructing a training set and a testing set according to the historical data;
wherein the fault electrical characteristic quantity includes: zero sequence current amplitude, active power, reactive power and zero sequence current;
s20, establishing a fault prediction model based on a deep belief network according to the fault electrical characteristic quantity;
s30, training the fault prediction model through a training set to obtain a trained fault prediction model;
s40, establishing a fuzzy probability interval;
s50, performing performance test on the trained fault prediction model by adopting a test set, performing confidence judgment on the output result of the fault prediction model through a fuzzy probability interval, and outputting the identification result of the fault line.
In this embodiment:
the zero sequence current amplitude value is that when the voltage is close to the peak value, the grounding fault in the small current grounding system mainly occurs, the discharge of the fault phase capacitor and the charging of the non-fault phase capacitor can cause an obvious transient process, and the amplitude value of the transient current can be obviously increased; and selecting the transient amplitude change of the zero sequence current after the fault occurs to realize the selection of a fault line. The zero sequence current amplitude in half cycle to one cycle after fault occurrence is generally taken as the characteristic electric quantity, and the expression is as follows:
wherein:set of zero sequence current amplitudes for each outgoing line when single-phase earth fault occurs to systemClose and/or fill>Zero sequence current amplitude for the ith line.
Reactive power and zero sequence current (IQ change) are suddenly changed when a single-phase grounding fault occurs in the low-current grounding system, and the reactive power and the zero sequence current of a fault phase are obviously larger compared with a non-fault phase value in a steady-state process after the fault occurs; therefore, the steady state value after the transient change of the reactive power and the zero sequence current can be used as the characteristic electric quantity. The steady state values of the third to fourth periods after the fault are taken as characteristic electric quantity, and the expression is as follows:
wherein:and->Respectively representing a zero sequence current and a reactive power steady state value set of each outgoing line when a single-phase earth fault occurs in the system,/->And->And the zero sequence current and the reactive power value of the ith outgoing line are respectively.
The zero sequence current amplitude and the IQ change are adopted as the basis for judging the small current grounding fault line by using the main characteristic electrical quantity; however, when the low-current grounding system is grounded through the arc suppression coil, zero sequence current can be compensated by the arc suppression coil to cause smaller zero sequence current change, which is likely to cause line selection failure, and when the system neutral point is connected with the arc suppression coil to cause the system to generate an active loop, as the characteristic signal is amplified, the active power amplitude of the fault line in the transient period can be found to be far greater than that of the non-fault line, therefore, the application also increases the active power as another important judgment basis for the condition that the neutral point is grounded through the arc suppression coil, and the method specifically comprises the following steps:
wherein:active power set of each outgoing line when single-phase earth fault occurs to system, < >>And (5) an active power value of the ith outgoing line.
In this embodiment, the step S20 of establishing a fault prediction model based on the deep belief network according to the fault electrical feature value includes: respectively predicting and establishing three prediction models aiming at fault characteristic electric quantity;
the deep belief network DBN1 is: fault line predictor model based on zero sequence current amplitude;
the deep belief network DBN2 is: a fault line predictor model based on the IQ change feature quantity;
the deep belief network DBN3 is: active power based fault line predictor models.
In this embodiment, as shown in fig. 2, S30, the training set trains the fault prediction model, and the steps for obtaining the trained fault prediction model are as follows:
s301, setting parameters of a Deep Belief Network (DBN), where the parameters include: inputting the number of neurons of a layer, the number of RBM layers, momentum parameters and learning rate;
in this embodiment, N input layer neurons of the deep belief network DBN1, 2N input layer neurons of the deep belief network DBN2, and N input layer neurons of the deep belief network DBN31 (N is the number of lines); the momentum parameter β may be 0.05; the learning rate α may be 0.001;
s302, input sample data in a training set is used as an input vector and is input into a first layer RBM of a fault prediction model, so that unsupervised training is completed;
in this embodiment, the input sample data in the training set is a D group, specifically:
the input data of the deep belief network DBN2 are:
s303, training a next RBM by taking the feature vector extracted after the first RBM training as an input vector;
s304, circularly executing the steps S203-S204, and sequentially completing training of RBM of each layer to obtain output characteristics of RBM of the top layer;
in the training process of each layer of RBM, obtaining a local optimal parameter of each layer of RBM, wherein the calculation formula is as follows:
in the above formula, beta is a momentum parameter, alpha is a learning rate, v is a gradient,the weight gradient is w is a DBN network weight parameter;
s305, a Softmax classifier is arranged on the top layer of the deep confidence network, the features extracted by the RBM are input into the Softmax classifier, and classification training is carried out by combining with fault line labels.
In this embodiment, the output of the fault prediction model is:
wherein,,numbering a fault line corresponding to the i-th group of fault electrical characteristic quantity data;
in this embodiment, probability that the feature belongs to each category may be calculated by using a probability calculation function, so as to complete the classification task; the calculation expression of the probability value function is as follows:
in the above formula, q represents a label class, which is a fault line name, q=n;
represents the i-th input sample data sequence, +.>A classification result label value corresponding to the ith input sample data is represented;
and the probability value of the ith sample data corresponding to various output results is represented, and theta is a built-in parameter of the softmax network.
In this embodiment, the computation expression of the cross entropy loss function in the Softmax classifier training is:
wherein m is the number of training samples, q represents the number of categories,representing input samplesThe root is->Network output of time,/->Representing a target output class of the ith sample;
s306, performing supervised fine tuning on the whole network parameters of the fault prediction model by using a labeled sample and using an error back propagation algorithm to enable the network performance to approach global optimum, so as to obtain a trained fault prediction model;
the output of the trained fault prediction model is as follows: LD1, LD2 and LD3, LD1 is the fault output result corresponding to deep belief network DBN1, LD2 is the fault output result corresponding to deep belief network DBN2, LD3 is the fault output result corresponding to deep belief network DBN 3.
wherein,,representing the true value of the jth neuron, for example>Predictive value representing the j-th neuron output,/->Representing the number of output neurons;
in the method, in the process of the invention,an adjustment value representing the connection weight between the ith input neuron and the jth output neuron in the kth network layer,/for the connection weight>To adjust step size +.>Representing the value of the ith input neuron.
The adjustment value of the final connection weight is as follows:
in the embodiment, the zero sequence current amplitude, reactive power and fault phase current are trained through the deep confidence network respectively, a fault line prediction model is built, the fault line is effectively judged according to four electrical characteristic quantities of the zero sequence current amplitude, the reactive power, the fault phase current and the active power when a single-phase earth fault occurs in the low-current grounding system, the confidence rate judgment is carried out on the output result of the fault prediction model by combining the fuzzy probability interval, the fault probability of the selected and pulled line is calculated, the accurate analysis and the reliable judgment of the selected and pulled sequence position of the low-current grounding fault are realized, and the practicability is extremely strong.
Examples
As shown in fig. 3, on the basis of the first embodiment, a fuzzy probability interval is established based on step S40 of the clustering and deep learning-based low-current ground fault line selection method; comprising the following steps:
s401, establishing fuzzy interval probability thresholds corresponding to the electrical characteristic quantities of all faults according to the training set;
s402, establishing a fuzzy probability interval based on the zero sequence current amplitude, the active power, the reactive power and the zero sequence current based on the fuzzy interval probability threshold.
S401, establishing fuzzy interval probability thresholds corresponding to the electrical feature quantities of all faults according to the training set; comprising the following steps:
s4011, clustering a sample database in a training set by adopting a self-adaptive density peak clustering algorithm, and solving a clustering center corresponding to each fault electrical characteristic quantity;
s4012, setting a fuzzy interval probability threshold corresponding to the fault electrical characteristic quantity according to the optimal clustering center;
the fuzzy interval probability threshold value corresponding to each fault electrical characteristic quantity comprises the following steps: a fuzzy interval probability threshold value based on the amplitude of the zero sequence current, a fuzzy interval probability threshold value based on the active power and a fuzzy interval probability threshold value based on the reactive power and the zero sequence current.
In this embodiment, as shown in fig. 4, in step S4011, a self-adaptive density peak clustering algorithm is adopted to cluster sample databases in a training set, and a cluster center corresponding to each fault electrical feature is obtained; comprising the following steps:
s4011-1, dividing the training set into: a sample data set S1 based on zero sequence current amplitude, a sample data set S2 based on reactive power and a sample data set S3 based on active power;
s4011-2, setting the number of clustering centers, and clustering the sample data set S1, the sample data set S2 and the sample data set S3 by adopting a self-adaptive density peak clustering algorithm to obtain clustering centers respectively corresponding to the sample data set S1, the sample data set S2 and the sample data set S3Cluster center->And clustering center->;
s4011-3, the calculation process of the clustering center corresponding to each sample data set comprises the following steps:
s4011-3-1 calculates local density corresponding to training set of sample data setAnd relative distance->;
in the above equation, i and j are two samples in the training set,for the density of the ith dot, +.>The Euclidean distance representing the ith and jth points; />Representing a cutoff distance; />The calculation formula of (2) is as follows:
wherein:and->The k element value in samples i and j; k is the number of elements in a single sample.
wherein: s is the training set, and the S is the training set,the method comprises the steps of carrying out a first treatment on the surface of the D is the number of samples in the training set;
Further, by changing the value of the cutoff distance, and calculating eachThe value corresponds to the coefficient of foundation, taking the +.>The value is taken as a final cut-off distance value; wherein the coefficient of kunning G is calculated as follows:
in the above formula: d is the number of samples in the training set S,for sample i +.>The proportion of the sample is as follows:
further, the final cut-off distance value is utilizedCalculating local Density +.>And relative distance->。
S4011-3-2 according to local densityAnd relative distance->Drawing a rho-delta distribution diagram;
s4011-3-3, taking the point with the largest descending trend and the previous points as clustering centers according to the distribution diagram of rho-delta; the method specifically comprises the following steps:
wherein:is the current u value; />And->The u values representing the previous and the subsequent instants, respectively.
Thereafter, according to the descending trend sumThe point with the largest descending trend and the former points are taken as clustering centers.
In this embodiment, step S4012 sets a fuzzy interval probability threshold corresponding to the fault electrical feature according to the optimal clustering center; comprising the following steps:
s4012-1, setting the number of clustering centers to be 3; then the first time period of the first time period,
S4012-2, determining a fuzzy interval probability threshold according to the clustering center; wherein,,
fuzzy interval probability threshold values based on zero sequence current amplitude values are respectively low threshold valuesMiddle threshold->High threshold->And critical threshold->The method comprises the steps of carrying out a first treatment on the surface of the And->、/>、/>;
Fuzzy interval probability threshold values based on reactive power are respectively low threshold valuesMiddle threshold->High threshold->And critical threshold->The method comprises the steps of carrying out a first treatment on the surface of the And->、/>、/>;
Fuzzy interval probability threshold values based on active power are respectively low threshold valuesMiddle threshold->High threshold->And critical threshold->: and->、/>、/>.;
The low threshold valueLow threshold->Low threshold->The zero sequence current amplitude value and the reactive power value and the active power value in the normal running state are respectively set.
In summary, the fuzzy interval probability threshold is as follows:
further, step S402 establishes a fuzzy probability interval based on the zero sequence current amplitude, the active power, the reactive power and the zero sequence current based on the fuzzy interval probability threshold; comprising the following steps:
s4021, judging zero sequence current amplitude of a line i to be judged based on a fuzzy interval probability threshold valueActive powerReactive power->And zero sequence current->The corresponding confidence probabilities;
in this embodiment, step S4021 is described, in which the zero sequence current amplitude of the line i to be determined is determined based on the probability threshold of the fuzzy intervalActive power->Reactive power->The corresponding confidence probabilities; comprising the following steps:
when the zero sequence current amplitude of the line i to be judgedThe method comprises the following steps:
S4022, setting a corresponding confidence coefficient judgment rule according to the type of the ground fault; comprising the following steps:
s4022-1, if the single-phase earth fault exists, the confidence rate judging rule is as follows:
when the zero sequence current amplitude isIf the confidence probability of the fault line is high, the confidence coefficient of the prediction result of the fault line is high;
when the zero sequence current amplitude isIs the medium probability, and reactive power +.>If the confidence probability of the fault line is low, the confidence coefficient of the prediction result of the fault line is medium;
when the zero sequence current amplitude isIs the medium probability, and reactive power +.>If the confidence probability of the fault line is medium probability or high probability, the confidence coefficient of the prediction result of the fault line is high;
when the zero sequence current amplitude isIs low and reactive power +.>The confidence probability of the prediction result of the fault line is low or medium probability;
when the zero sequence current amplitude isIs low and reactive power +.>If the confidence probability of the fault line is high, the confidence coefficient of the prediction result of the fault line is high;
to sum up, in this embodiment, in the zero sequence current amplitudeActive power->Reactive power->And zero sequence currentAmong the three types of criteria, the zero sequence current amplitude is preferentially considered, namely when the fault zero sequence current amplitude criterion is in a high probability interval, the prediction reliability of a fault line is high probability; secondly, considering fault characteristic IQ criteria, describing the accuracy of line selection when the low-current ground fault is as follows:
s4022-2, if the arc suppression coil grounding fault exists, the confidence rate judging rule is as follows:
when the zero sequence current amplitude isIf the confidence probability of the fault line is high, the confidence coefficient of the prediction result of the fault line is high;
when the zero sequence current amplitude isIs the medium probability, and reactive power +.>The confidence probability of (2) is low probability, active power +.>If the confidence probability of the fault line is low, the confidence coefficient of the prediction result of the fault line is medium;
when the zero sequence current amplitude isIs the medium probability, and reactive power +.>Confidence probability of (2) is medium probability or high probability, active power +.>If the confidence probability of the fault line is medium probability or high probability, the confidence coefficient of the prediction result of the fault line is high;
when the zero sequence current amplitude isIs low and reactive power +.>Confidence probability of (2)Is low probability, active power +.>The confidence probability of the prediction result of the fault line is low;
when the zero sequence current amplitude isIs low and reactive power +.>Confidence probability of (2) is medium probability or high probability, active power +.>If the confidence probability of the fault line is medium probability or high probability, the confidence coefficient of the prediction result of the fault line is medium.
In summary, when the low-current grounding type is the grounding mode of the arc suppression coil, the fault line selection accuracy is as shown in the table:
in the embodiment, in S50, confidence coefficient judgment is performed on the output result of the fault prediction model through the fuzzy probability interval, and the identification result of the fault line is output; comprising the following steps:
when the confidence coefficient of the prediction result is high, outputting a prediction fault line with the zero sequence current amplitude as a criterion;
outputting a predicted fault line with reactive power and zero sequence current as criteria when the confidence coefficient of the predicted result is middle;
when the confidence of the prediction result is low, the failure prediction line is not output.
As shown in fig. 5, in this embodiment, when the confidence coefficient of the prediction result is high, a fault output result LD1 corresponding to the deep confidence network DBN1 is output; and when the confidence coefficient of the predicted result is middle, outputting fault output results LD1 and LD2 corresponding to the deep confidence networks DBN1 and DBN 2.
In the embodiment, the threshold value of the fuzzy probability interval is determined based on the ADPC clustering algorithm, the optimal clustering center is selected through the self-adaptive mechanism, the purpose of accurately judging the fault route selection prediction result according to the fault characteristic data is achieved, the fuzzy probability interval is determined based on the clustering algorithm, the problem of confidence degree selection of the fault route prediction result can be effectively solved, and a reliable and reasonable decision basis is provided for fault route selection auxiliary decision; and the prediction result of the fault prediction model is subjected to line selection recognition based on the probability interval of fuzzy control, so that the auxiliary decision efficiency of small-current ground fault line selection is effectively improved, and the safe and stable operation of the power grid is ensured.
In addition, the application also provides electronic equipment, which comprises:
a memory;
a processor; and
a computer program;
wherein the computer program is stored in the memory and configured to be executed by the processor to implement the method as described above.
Furthermore, the present application also provides a computer-readable storage device having a computer program stored thereon; the computer program is executed by a processor to implement the method as described above.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein. The schemes in the embodiments of the present application may be implemented in various computer languages, for example, C language, VHDL language, verilog language, object-oriented programming language Java, and transliteration scripting language JavaScript, etc.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In the description of the present application, it should be understood that the terms "center," "longitudinal," "transverse," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," and the like indicate an orientation or positional relationship based on that shown in the drawings, merely for convenience of description and to simplify the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be configured and operated in a particular orientation, and thus should not be construed as limiting the present application.
Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include one or more such feature. In the description of the present application, the meaning of "plurality" is at least two, such as two, three, etc., unless explicitly defined otherwise.
In this application, unless specifically stated and limited otherwise, the terms "mounted," "connected," "secured," and the like are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally formed; may be mechanically connected, may be electrically connected or may communicate with each other; can be directly connected or indirectly connected through an intermediate medium, and can be communicated with the inside of two elements or the interaction relationship of the two elements. The specific meaning of the terms in this application will be understood by those of ordinary skill in the art as the case may be.
While preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the application.
It will be apparent to those skilled in the art that various modifications and variations can be made in the present application without departing from the spirit or scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims and the equivalents thereof, the present application is intended to cover such modifications and variations.
Claims (10)
1. The small current ground fault line selection method based on clustering and deep learning is characterized by comprising the following steps of:
s10, acquiring historical data of fault electrical characteristic values of all outgoing lines of the low-current grounding system when single-phase grounding faults occur; constructing a training set and a testing set according to the historical data;
wherein the fault electrical characteristic quantity includes: zero sequence current amplitude, active power, reactive power and zero sequence current;
s20, establishing a fault prediction model based on a deep belief network according to the fault electrical characteristic quantity;
s30, training the fault prediction model through a training set to obtain a trained fault prediction model;
s40, establishing a fuzzy probability interval based on an ADPC clustering algorithm;
s50, performing performance test on the trained fault prediction model by adopting a test set, performing confidence judgment on the output result of the fault prediction model through a fuzzy probability interval, and outputting the identification result of the fault line.
2. The small current ground fault line selection method based on clustering and deep learning according to claim 1, wherein the step S40 is based on an ADPC clustering algorithm to establish a fuzzy probability interval; comprising the following steps:
s401, establishing fuzzy interval probability thresholds corresponding to the electrical characteristic quantities of all faults according to the training set;
s402, establishing a fuzzy probability interval based on the zero sequence current amplitude, the active power, the reactive power and the zero sequence current based on the fuzzy interval probability threshold.
3. The small current ground fault line selection method based on clustering and deep learning according to claim 2, wherein the step S401 is to establish fuzzy interval probability thresholds corresponding to the electrical feature quantities of each fault according to a training set; comprising the following steps:
s4011, clustering a sample database in a training set by adopting a self-adaptive density peak clustering algorithm, and solving a clustering center corresponding to each fault electrical characteristic quantity;
s4012, setting a fuzzy interval probability threshold corresponding to the fault electrical characteristic quantity according to the optimal clustering center;
the fuzzy interval probability threshold value corresponding to each fault electrical characteristic quantity comprises the following steps: a fuzzy interval probability threshold value based on the amplitude of the zero sequence current, a fuzzy interval probability threshold value based on the active power and a fuzzy interval probability threshold value based on the reactive power and the zero sequence current.
4. The small current grounding fault line selection method based on clustering and deep learning according to claim 3, wherein the step S4011 adopts an adaptive density peak clustering algorithm to cluster sample databases in a training set, and obtains a clustering center corresponding to each fault electrical characteristic quantity; comprising the following steps:
s4011-1, dividing the training set into: a sample data set S1 based on zero sequence current amplitude, a sample data set S2 based on reactive power and a sample data set S3 based on active power;
s4011-2, setting the number of clustering centers, and clustering the sample data set S1, the sample data set S2 and the sample data set S3 by adopting an adaptive density peak clustering algorithm to obtain clustering centers C respectively corresponding to the sample data set S1, the sample data set S2 and the sample data set S3 1 Cluster center C 2 And a cluster center C 3 ;
S4011-3, the calculation process of the clustering center corresponding to each sample data set comprises the following steps:
s4011-3-1 calculates local density rho corresponding to sample data set training set i And relative distance delta i ;
S4011-3-2 according to local density ρ i And relative distance delta i Drawing a rho-delta distribution diagram;
s4011-3-3, taking the point with the largest descending trend and the previous points as clustering centers according to the distribution diagram of rho-delta.
5. The small current ground fault line selection method based on clustering and deep learning according to claim 4, wherein the step S4012 sets a fuzzy interval probability threshold corresponding to the fault electrical feature according to the optimal clustering center; comprising the following steps:
s4012-1, setting the number of clustering centers to be 3; then the first time period of the first time period,
clustering center C corresponding to sample data set S1 1 The method comprises the following steps: c (C) 1 ={c M.1 ,c M.2 ,c M.3 };
Clustering center C corresponding to sample data set S2 2 The method comprises the following steps: c (C) 2 ={c A.1 ,c A.2 ,c A.3 };
Clustering center C corresponding to sample data set S3 3 The method comprises the following steps: c (C) 3 ={c P.1 ,c P.2 ,c P.3 };
S4012-2, determining a fuzzy interval probability threshold according to the clustering center; wherein,,
fuzzy interval probability threshold values based on zero sequence current amplitude are respectively low threshold value I M.l Intermediate threshold I M.m High threshold I M.h And critical threshold I M.n The method comprises the steps of carrying out a first treatment on the surface of the And I M.m =C M.3 、I M.h =C M.2 、I M.n =C M.1 ;
Fuzzy interval probability threshold values based on reactive power are respectively low threshold value I A.l Intermediate threshold I A.m High threshold I A.h And critical threshold I A.n The method comprises the steps of carrying out a first treatment on the surface of the And I A.m =C A.3 、I A.h =C A.2 、I A.n =C A.1 ;
Fuzzy interval probability threshold values based on active power are respectively low threshold value I P.l Intermediate threshold I P.m High threshold I P.h And critical threshold I P.n : and I P.m =C P.3 、I P.h =C P.2 、I P.n =C P.1 .;
The low threshold I M.l Low threshold I A.l Low threshold I A.l The zero sequence current amplitude value and the reactive power value and the active power value in the normal running state are respectively set.
6. The small current ground fault line selection method based on clustering and deep learning according to claim 5, wherein the step S402 establishes a fuzzy probability interval based on zero sequence current amplitude, active power, reactive power and zero sequence current based on a fuzzy interval probability threshold; comprising the following steps:
s4021, judging zero sequence current amplitude I of line I to be judged based on fuzzy interval probability threshold M.i Active power I P.i Reactive power I A.i And zero sequence current I Q.i The corresponding confidence probabilities;
s4022, setting a corresponding confidence coefficient judgment rule according to the type of the ground fault; comprising the following steps:
s4022-1, if the single-phase earth fault exists, the confidence rate judging rule is as follows:
when the zero sequence current amplitude I M.i If the confidence probability of the fault line is high, the confidence coefficient of the prediction result of the fault line is high;
when the zero sequence current amplitude I M.i Is the medium probability, and reactive power I A.i If the confidence probability of the fault line is low, the confidence coefficient of the prediction result of the fault line is medium;
when the zero sequence current amplitude I M.i Is the medium probability, and reactive power I A.i If the confidence probability of the fault line is medium probability or high probability, the confidence coefficient of the prediction result of the fault line is high;
when the zero sequence current amplitude I M.i Is of low probability and reactive power I A.i The confidence probability of the prediction result of the fault line is low or medium probability;
when the zero sequence current amplitude I M.i Is of low probability and reactive power I A.i If the confidence probability of the fault line is high, the confidence coefficient of the prediction result of the fault line is high;
s4022-2, if the arc suppression coil grounding fault exists, the confidence rate judging rule is as follows:
when the zero sequence current amplitude I M.i If the confidence probability of the fault line is high, the confidence coefficient of the prediction result of the fault line is high;
when the zero sequence current amplitude I M.i Is the medium probability, and reactive power I A.i Confidence probability of (2) is low probability, active power I P.i Confidence probability of (2)If the probability is low, the confidence coefficient of the prediction result of the fault line is medium;
when the zero sequence current amplitude I M.i Is the medium probability, and reactive power I A.i The confidence probability of (1) is medium probability or high probability, active power I P.i If the confidence probability of the fault line is medium probability or high probability, the confidence coefficient of the prediction result of the fault line is high;
when the zero sequence current amplitude I M.i Is of low probability and reactive power I A.i Confidence probability of (2) is low probability, active power I P.i The confidence probability of the prediction result of the fault line is low;
when the zero sequence current amplitude I M.i Is of low probability and reactive power I A.i The confidence probability of (1) is medium probability or high probability, active power I P.i If the confidence probability of the fault line is medium probability or high probability, the confidence coefficient of the prediction result of the fault line is medium.
7. The small current ground fault line selection method based on clustering and deep learning as claimed in claim 6, wherein the step S4021 is to judge the zero sequence current amplitude I of the line I to be judged based on the fuzzy interval probability threshold M.i Active power I P.i Reactive power I A.i The corresponding confidence probabilities; comprising the following steps:
when the zero sequence current amplitude I of the line I to be judged M.i The method comprises the following steps:
I M.h ≤I M.i <I M.n zero sequence current amplitude I M.i The confidence probability of (2) is high probability;
I M.m ≤I M.i <I M.h zero sequence current amplitude I M.i The confidence probability of (2) is a medium probability;
I M.l ≤I M.i <I M.m zero sequence current amplitude I M.i The confidence probability of (2) is low probability;
when the active power I of the line I to be judged P.i The method comprises the following steps:
I P.h ≤I P.i <I P.n zero sequence current amplitude I M.i The confidence probability of (2) is high probability;
I P.m ≤I P.i <I P.h zero sequence current amplitude I M.i The confidence probability of (2) is a medium probability;
I P.l ≤I P.i <I P.m zero sequence current amplitude I M.i The confidence probability of (2) is low probability;
when the reactive power I of the line I to be judged A.i The method comprises the following steps:
I A.h ≤I A.i <I A.n zero sequence current amplitude I M.i The confidence probability of (2) is high probability;
I A.m ≤I A.i <I A.h zero sequence current amplitude I M.i The confidence probability of (2) is a medium probability;
I A.l ≤I A.i <I A.m zero sequence current amplitude I M.i Is a low probability.
8. The small current grounding fault line selection method based on clustering and deep learning as claimed in claim 6, wherein in S50, confidence ratio judgment is performed on the output result of the fault prediction model through a fuzzy probability interval, and the identification result of the fault line is output; comprising the following steps:
when the confidence coefficient of the prediction result is high, outputting a prediction fault line with the zero sequence current amplitude as a criterion;
outputting a predicted fault line with reactive power and zero sequence current as criteria when the confidence coefficient of the predicted result is middle;
when the confidence of the prediction result is low, the failure prediction line is not output.
9. An electronic device, comprising:
a memory;
a processor; and
a computer program;
wherein the computer program is stored in the memory and configured to be executed by the processor to implement the method of any one of claims 1 to 8.
10. A computer readable storage device having a computer program stored thereon; the computer program being executed by a processor to implement the method of any one of claims 1 to 8.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310460342.0A CN116227538B (en) | 2023-04-26 | 2023-04-26 | Clustering and deep learning-based low-current ground fault line selection method and equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310460342.0A CN116227538B (en) | 2023-04-26 | 2023-04-26 | Clustering and deep learning-based low-current ground fault line selection method and equipment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116227538A CN116227538A (en) | 2023-06-06 |
CN116227538B true CN116227538B (en) | 2023-07-11 |
Family
ID=86585776
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310460342.0A Active CN116227538B (en) | 2023-04-26 | 2023-04-26 | Clustering and deep learning-based low-current ground fault line selection method and equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116227538B (en) |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117347788B (en) * | 2023-10-17 | 2024-06-11 | 国网四川省电力公司电力科学研究院 | Power distribution network single-phase earth fault class probability prediction method |
CN117092445A (en) * | 2023-10-19 | 2023-11-21 | 盛隆电气集团有限公司 | Fault detection method and system of power distribution system based on big data |
CN117992896B (en) * | 2024-04-03 | 2024-06-21 | 北京天阳睿博科技有限公司 | High-voltage power supply fault positioning line selection system for coal mining area |
Family Cites Families (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
AT10073U9 (en) * | 2008-01-14 | 2009-02-15 | Avl List Gmbh | METHOD AND DEVICE FOR ANALYZING AND EVALUATING MEASUREMENT DATA OF A MEASURING SYSTEM |
CN101227086B (en) * | 2008-01-22 | 2010-06-02 | 昆明理工大学 | Power distribution network cable-line commingle line fault route selection anastomosing method using extension theory |
CN202676854U (en) * | 2012-06-13 | 2013-01-16 | 山东电力集团公司德州供电公司 | Fuzzy comprehensive decision power distribution network single-phase earth fault line selection system |
CN106370981B (en) * | 2016-09-30 | 2019-03-26 | 中国石油化工股份有限公司 | A kind of distribution network failure route selection method based on fuzzy cluster analysis |
CN108562821B (en) * | 2018-05-08 | 2021-09-28 | 中国电力科学研究院有限公司 | Method and system for determining single-phase earth fault line selection of power distribution network based on Softmax |
CN110163255B (en) * | 2019-04-22 | 2021-11-16 | 济南大学 | Data stream clustering method and device based on density peak value |
CN110941918B (en) * | 2019-12-30 | 2024-03-19 | 国网安徽省电力有限公司 | Intelligent substation fault analysis system |
CN114611596A (en) * | 2022-03-07 | 2022-06-10 | 哈尔滨理工大学 | Self-adaptive density peak value clustering method |
CN115112992A (en) * | 2022-06-07 | 2022-09-27 | 国网山西省电力公司大同供电公司 | Fuzzy entropy weight-based low-current ground fault comprehensive line selection method for Vague set |
-
2023
- 2023-04-26 CN CN202310460342.0A patent/CN116227538B/en active Active
Also Published As
Publication number | Publication date |
---|---|
CN116227538A (en) | 2023-06-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN116227538B (en) | Clustering and deep learning-based low-current ground fault line selection method and equipment | |
CN109818349B (en) | Power grid robust state prediction method based on multidimensional state matrix sliding matching | |
CN108053128B (en) | Electric network transient stability rapid evaluation method based on ELM and TF | |
CN108832619A (en) | Transient stability evaluation in power system method based on convolutional neural networks | |
CN112101684B (en) | Plug-in hybrid electric vehicle real-time energy management method and system | |
CN111523785A (en) | Power system dynamic security assessment method based on generation countermeasure network | |
CN110994604B (en) | Power system transient stability assessment method based on LSTM-DNN model | |
CN112508442B (en) | Transient stability assessment method and system based on automatic and interpretable machine learning | |
CN103177188A (en) | Electric system load dynamic characteristic classifying method based on characteristic mapping | |
CN104899135A (en) | Software defect prediction method and system | |
CN110837915B (en) | Low-voltage load point prediction and probability prediction method for power system based on hybrid integrated deep learning | |
CN115469260B (en) | Hausdorff-based current transformer anomaly identification method and system | |
CN110007235A (en) | A kind of accumulator of electric car SOC on-line prediction method | |
CN109787821B (en) | Intelligent prediction method for large-scale mobile client traffic consumption | |
CN114154700A (en) | User power consumption prediction method based on transformer model | |
CN113459867A (en) | Electric vehicle charging process fault early warning method based on adaptive deep confidence network | |
CN112632840A (en) | Power grid transient stability evaluation method based on adaptive differential evolution algorithm and ELM | |
CN112327189A (en) | KNN algorithm-based energy storage battery health state comprehensive judgment method | |
CN113988215A (en) | Power distribution network metering cabinet state detection method and system | |
CN112651519A (en) | Secondary equipment fault positioning method and system based on deep learning theory | |
CN112821424A (en) | Power system frequency response analysis method based on data-model fusion drive | |
CN115526258A (en) | Power system transient stability evaluation method based on Spearman correlation coefficient feature extraction | |
CN113988558B (en) | Power grid dynamic security assessment method based on blind area identification and electric coordinate system expansion | |
CN112327190B (en) | Method for identifying health state of energy storage battery | |
Dhurandhar et al. | Enhancing simple models by exploiting what they already know |
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 |