CN113837477A - Data dual-drive power grid fault prediction method, device and equipment under typhoon disaster - Google Patents
Data dual-drive power grid fault prediction method, device and equipment under typhoon disaster Download PDFInfo
- Publication number
- CN113837477A CN113837477A CN202111139000.6A CN202111139000A CN113837477A CN 113837477 A CN113837477 A CN 113837477A CN 202111139000 A CN202111139000 A CN 202111139000A CN 113837477 A CN113837477 A CN 113837477A
- Authority
- CN
- China
- Prior art keywords
- disaster
- data
- causing
- typhoon
- power grid
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 59
- 230000003068 static effect Effects 0.000 claims abstract description 44
- 230000015654 memory Effects 0.000 claims abstract description 20
- 230000007246 mechanism Effects 0.000 claims abstract description 17
- 238000013528 artificial neural network Methods 0.000 claims abstract description 15
- 238000013507 mapping Methods 0.000 claims abstract description 10
- 238000012549 training Methods 0.000 claims description 49
- 238000012360 testing method Methods 0.000 claims description 44
- 238000004422 calculation algorithm Methods 0.000 claims description 36
- 230000006870 function Effects 0.000 claims description 18
- 238000011160 research Methods 0.000 claims description 13
- 230000008569 process Effects 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 11
- 239000011159 matrix material Substances 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 9
- 238000005457 optimization Methods 0.000 claims description 7
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 5
- 210000002569 neuron Anatomy 0.000 claims description 4
- 238000012423 maintenance Methods 0.000 claims description 3
- 238000002156 mixing Methods 0.000 claims description 3
- 238000001556 precipitation Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 4
- 239000010410 layer Substances 0.000 description 22
- 238000010586 diagram Methods 0.000 description 8
- 230000008859 change Effects 0.000 description 6
- 238000004590 computer program Methods 0.000 description 5
- 230000006403 short-term memory Effects 0.000 description 5
- 230000009471 action Effects 0.000 description 4
- 230000009977 dual effect Effects 0.000 description 4
- 238000000605 extraction Methods 0.000 description 4
- 201000004569 Blindness Diseases 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 3
- 239000000284 extract Substances 0.000 description 3
- 230000007787 long-term memory Effects 0.000 description 3
- 230000007774 longterm Effects 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 239000011229 interlayer Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- JTJMJGYZQZDUJJ-UHFFFAOYSA-N phencyclidine Chemical class C1CCCCN1C1(C=2C=CC=CC=2)CCCCC1 JTJMJGYZQZDUJJ-UHFFFAOYSA-N 0.000 description 1
- 238000012958 reprocessing Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
-
- 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
- G06N3/045—Combinations of networks
-
- 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
- G06N3/048—Activation functions
-
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4007—Scaling of whole images or parts thereof, e.g. expanding or contracting based on interpolation, e.g. bilinear interpolation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Economics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
- Quality & Reliability (AREA)
Abstract
The invention discloses a method, a device and equipment for predicting power grid faults under a data dual-drive typhoon disaster, wherein the method comprises the following steps: the method comprises the steps of constructing a disaster-causing data set, balancing the disaster-causing data set, constructing a double-channel prediction model and predicting by using the double-channel prediction model, classifying multiple influence factors of the disaster-suffering situation of the power distribution network under typhoon disasters into static data and dynamic data, extracting the characteristics of the static data by using a feedforward neural network, extracting the characteristics of the dynamic data by using a long-short term memory network after the multi-head self-attention mechanism is strengthened, finally fusing all the extracted characteristics by using a linear layer, and establishing the mapping relation between the multiple influence factors and the disaster-suffering situation of the power distribution network. The stability of the effect of static data on the power distribution network disaster situation and the time-varying property and the accumulative property of the effect of dynamic data on the power distribution network disaster situation are fully considered, and a power distribution network fault prediction model with higher accuracy and stronger interpretability under the typhoon disaster is constructed.
Description
Technical Field
The invention belongs to the technical field of power grid fault prediction, and particularly relates to a method, a device and equipment for predicting power grid faults under a data dual-drive typhoon disaster.
Background
Typhoon disasters have large influence range and long duration, and along with the change of global climate, the proportion of tropical cyclones with typhoons and above strength is continuously increased in recent years, thus posing great threat to the normal operation of power transmission and distribution networks in coastal areas. Compared with a power transmission network, the power distribution network has more equipment, the ageing problem of the equipment is serious, and the power distribution network is more easily influenced by natural disasters such as typhoon and the like. Therefore, it is necessary to research a power distribution network fault prediction method under a typhoon disaster aiming at the destructiveness of the typhoon and the vulnerability of the power distribution network, and provide reliable prior information for an elasticity enhancement strategy of the power distribution network.
The research on the power distribution network fault prediction method under the typhoon disaster mainly comprises a physical model based on a disaster-causing mechanism and a data driving model based on historical data. The research idea of the physical model is to establish a wind load model of the power distribution network line and the tower according to the probability distribution of the actual wind speed and the design wind speed of the equipment, and correct the model by combining the geographical environment of the equipment, the service life of the equipment and other factors, so that the fault probability of the line and the tower under the typhoon disaster is obtained. The research idea of the data-driven model is to construct a data set containing disaster-causing factors and network faults based on historical meteorological information, geographic information and power grid information, learn the data set through a machine learning model and establish a corresponding mapping relation. Meanwhile, in consideration of the fact that fault data of the power distribution network under the typhoon disaster contains a large number of samples with zero fault number, a machine learning model generates large prediction deviation in the samples with the non-zero fault number, the existing research generally adopts a synthetic minority over-sampling technology (SMOTE) to generate a minority sample equilibrium data set, or adopts a cost-sensitive learning method to endow different types with different punishment coefficients, so that the learning side weight of the model on the minority samples is improved.
Under the limitation of modeling complexity, the physical model is difficult to comprehensively and finely model the influence factors of the power distribution network equipment faults, so that certain prediction accuracy is lost. With the increasing perfection of data collection and management systems of power departments and meteorological departments, current research focuses on predicting the fault conditions of power distribution networks under typhoon disasters through data driving models. However, the existing data driving model only considers the relationship between each influence factor in each time section and the power distribution network fault, and does not consider the accumulative performance of partial factors on the power distribution network fault. Meanwhile, the SMOTE algorithm adopted in the current research has certain blindness and randomness in the selection process of a sample synthesis object, the quality of a few types of generated samples is poor, when a cost sensitive learning method determines punishment coefficients of various types, parameters need to be adjusted repeatedly according to model performance, and the adjustment direction is subjective.
Disclosure of Invention
The invention provides a method, a device and equipment for predicting the power grid fault under the typhoon disaster by data dual drive, which improve the accuracy and interpretability of the method for predicting the power distribution network fault under the typhoon disaster and enhance the resistance of the power distribution network to the typhoon disaster.
In order to achieve the purpose, the invention provides a method for predicting the fault of a power distribution network under a typhoon disaster by using data dual drive, which comprises the following steps of:
step1, collecting multivariate influence data of power grid faults under typhoon disasters and the sum of permanent tripping times of a power grid in a predicted area, dividing the data into static data and dynamic data according to time domain variation attributes of the data, and constructing a disaster-causing data set by utilizing the static data, the dynamic data and the sum of the permanent tripping times of the power grid in the predicted area;
step2, carrying out equalization processing on the disaster-causing data set;
step3, extracting the characteristics of static data in the disaster-causing data set by using a feedforward neural network, extracting the sequence characteristics of dynamic data in the disaster-causing data set by using a long-short term memory network and a multi-head self-attention mechanism, establishing a dual-channel prediction model of the power grid fault under the typhoon disaster, and solving and adjusting the model parameters based on the disaster-causing data set after sample equalization processing to finally obtain an optimized dual-channel prediction model; and evaluating the performance of the device; if the performance meets the requirements, performing the step4, otherwise, continuing to perform optimization;
and 4, collecting corresponding multivariate influence data of the prediction region under the future typhoon disaster, constructing a disaster causing data set, inputting the disaster causing data set into the two-channel prediction model optimized in the step3, and obtaining a prediction value of the power grid fault condition of the research region under the future typhoon disaster.
Further, in step1, the static data includes forest coverage, land type, maintenance degree of the power grid and population density, and the dynamic data includes distance between a typhoon center and an area center, center minimum air pressure of the typhoon, near-center maximum wind speed of the typhoon, moving direction angle of the typhoon, seven-level wind circle radius, average wind speed of a prediction area and precipitation of the prediction area.
Further, the process of step2 is: dividing a minority sample set according to distribution of disaster-causing data sets in a high-dimensional space by using a Borderline-SMOTE1 algorithm, and generating samples according to the minority samples at a decision boundary after division; and then, checking the difference of the data distribution of the training set and the test set through a discriminant model, performing parameter tuning on the Borderline-SMOTE1 algorithm according to the difference, and finally, applying the Borderline-SMOTE1 algorithm after parameter optimization to balance the disaster-causing data set.
Further, step2 comprises the following steps:
step 2.1, calculating m nearest neighbor samples of each light fault sample by using a K nearest neighbor algorithm;
2.2, according to the proportion of the light fault samples in the m nearest neighbor samples of the light fault samples, classifying the light fault samples into safety samples, dangerous samples and noise samples;
step 2.3, aiming at each danger class sample xiSelecting a desired number of light fault samples from the K nearest neighbor samplesThen, the process is carried out;
step 2.4, for each selected neighbor sample xj' Generation of light fault class New samples x using linear interpolationi,j；
Step 2.5, adding the generated mild fault type new sample into the original disaster-causing training set to obtain an updated disaster-causing data set;
and 2.6, checking the updated disaster causing data set, performing the step3 if the updated disaster causing data set meets the requirement, and performing parameter adjustment on the Borderline-SMOTE1 algorithm if the updated disaster causing data set does not meet the requirement until the updated disaster causing data set meets the requirement.
Further, step 2.6 comprises the steps of:
step 2.6.1, randomly sampling the disaster-causing training set to ensure that the number of samples of the sampled training set is equal to that of the samples of the disaster-causing test set; setting labels of the training set samples and the test set samples as 0 and 1 respectively, mixing to form a discrimination data set, and dividing the discrimination data set into a new training set and a new test set according to a proportion;
step 2.6.2, on the basis of the new training set and the new testing set, taking a cross entropy function as a loss function, obtaining the gradient of each parameter value of the discrimination model through an error back propagation method, and further updating all parameters of the discrimination model through an Adam gradient descent algorithm to obtain the discrimination accuracy of the disaster-causing training set and the disaster-causing testing set;
step 2.6.3, measuring the sample distribution difference of the disaster-causing training set and the disaster-causing testing set by using the ability of distinguishing the disaster-causing training set and the disaster-causing testing set by using a distinguishing model, and adjusting parameters such as the nearest neighbor sample number of the Borderline-SMOTE1 algorithm when the distinguishing accuracy is higher than the accuracy threshold; and when the judgment accuracy is lower than the accuracy threshold, executing the step 3.
Further, step3 comprises the following steps:
3.1, extracting static characteristics from the static data based on a feedforward neural network; extracting dynamic features from the dynamic data based on a long-short term memory network and a multi-head attention mechanism;
step 3.2, splicing the static characteristics and the dynamic characteristics, mapping the static characteristics and the dynamic characteristics into prediction probabilities of all fault condition types of the power grid through a linear layer, and obtaining a prediction model by taking a maximum probability value corresponding to a disaster-suffered type as a prediction fault condition type of a sample; measuring the difference degree between the predicted value and the actual value by using a cross entropy function as a loss function; then obtaining the gradient value of each parameter in the cross entropy function pair model through an error back propagation algorithm; finally, updating the parameters of the prediction model by using a small-batch Adam algorithm according to the learning rate, the batch size and the number of neurons in each layer;
and 3.3, taking precision ratio and recall ratio as a basic index system, introducing a macro-average mechanism, comprehensively considering the performance of the prediction model in different types of sample sets in the disaster-causing test set, and evaluating the prediction model.
Further, step 3.3 comprises the steps of:
step 3.3.1, according to the predicted value obtained after the disaster causing test set is input into the prediction model, counting whether each sample in the disaster causing test set belongs to the actual value and the predicted value of the disaster type, and forming three two-class confusion moments;
step 3.3.2, obtaining a group of true positive TPs corresponding to each confusion matrix according to the matrix elementsiFalse positive FPiTrue negative TNiAnd false negative FNiFurther obtain the corresponding precision ratio PiAnd recall ratio Ri；
Step 3.3.3, according to the precision ratio PiRecall ratio RiAnd F1 to obtain macro precision macro-P, macro recall macro-R and macro F1 value macro-F1;
and 3.3.4, evaluating the performance of the power grid fault condition prediction model under the typhoon disaster according to the four indexes of the macro precision ratio, the macro F1 and the accuracy.
A power grid fault prediction device under a typhoon disaster comprises:
the acquisition module is used for acquiring data and transmitting the acquired data to the calculation output module; the data comprises multivariate influence data of power grid faults under typhoon disasters, the sum of permanent tripping times of the power grid in a predicted area and real-time typhoon data;
and the calculation output module is used for training a prediction model according to the acquired data set and outputting a power grid fault prediction value according to the prediction model and the real-time typhoon data.
A computer device comprising an electrically connected memory and a processor, the memory having stored thereon a computing program operable on the processor, when executing the computing program, performing the steps of the method of any of claims 1-8.
Compared with the prior art, the invention has at least the following beneficial technical effects:
the method classifies the multiple influence factors of the power distribution network fault condition under the typhoon disaster into static data and dynamic data, extracts the characteristics of the static data by using a feedforward neural network, extracts the characteristics of the dynamic data by using a long-short term memory network after the multi-head self-attention mechanism is strengthened, and finally fuses all the extracted characteristics by using a linear layer to establish the mapping relation between the multiple influence factors and the power distribution network fault condition. The two-channel prediction model constructed by the method fully considers the stability of the static data on the power distribution network disaster situation and the time-varying property and the accumulative property of the dynamic data on the power distribution network fault situation, and constructs the power distribution network fault prediction model with higher accuracy and stronger interpretability under the typhoon disaster.
The Borderline-SMOTE1 algorithm used by the invention identifies the samples at the decision boundary based on the K nearest neighbor algorithm, and synthesizes new samples by using random linear interpolation, so that the defects of blindness and randomness of the sample generation process, subjectivity and complexity of a penalty coefficient determination mode and the like in the existing sample imbalance processing mode are overcome, the imbalance degree of a disaster-causing data set is effectively reduced, a better data base is laid for the training of a power grid fault condition prediction model, the accuracy of a power distribution network fault prediction method under a typhoon disaster is improved, and the capability of the power distribution network in resisting the typhoon disaster is further enhanced.
Drawings
FIG. 1 is a schematic diagram of a disaster-causing data set;
FIG. 2 is a schematic diagram of the classification of light fault class samples of the Borderline-SMOTE1 algorithm;
FIG. 3 is a schematic diagram of a discriminant model test sample distribution;
FIG. 4 is a block diagram of an LSTM;
FIG. 5 is a network architecture diagram of a two-channel prediction model;
fig. 6 is a schematic block diagram of a power grid fault prediction apparatus provided in the present invention;
fig. 7 is a schematic structural diagram of a computer device provided in the present invention.
Detailed Description
In order to make the objects and technical solutions of the present invention clearer and easier to understand. The present invention will be described in further detail with reference to the following drawings and examples, wherein the specific examples are provided for illustrative purposes only and are not intended to limit the present invention.
In the description of the present invention, it is to be understood that the terms "center", "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", and the like, indicate orientations or positional relationships based on those shown in the drawings, and are used only for convenience in describing the present invention and for simplicity in description, and do not indicate or imply that the referenced devices or elements must have a particular orientation, be constructed and operated in a particular orientation, and thus, are not to be construed as limiting the present invention. Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of the present invention, "a plurality" means two or more unless otherwise specified. In the description of the present invention, it should be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly, e.g., as meaning either a fixed connection, a removable connection, or an integral connection; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
Example 1
Referring to fig. 1, the method for predicting the power distribution network fault under the typhoon disaster based on static and dynamic data dual drive comprises four parts, namely, building a disaster-causing data set, balancing the disaster-causing data set, building a dual-channel prediction model and predicting the fault condition of the power distribution network in the area under the future typhoon disaster.
Step1, selecting multiple influence factors of a power distribution network fault under a typhoon disaster from the four angles of meteorological information, geographic information, power grid information and population information, dividing the multiple influence factors into static data and dynamic data according to the time domain change attribute of the data (the change amplitude of the data during the typhoon crossing), and constructing a disaster-causing data set;
step2, aiming at the sample unbalance phenomenon in the disaster-causing data set, using a Borderline-SMOTE1 algorithm to divide the minority sample set according to the distribution of the disaster-causing data set in a high-dimensional space, and generating samples aiming at the minority sample at the decision boundary after division; then, the difference of data distribution of the training set and the test set is detected through a discriminant model, parameters of the Borderline-SMOTE1 algorithm are adjusted and optimized according to the difference, and finally the Borderline-SMOTE1 algorithm after parameter optimization is applied to balance the disaster-causing data set;
and 3, extracting the characteristics of static data in the disaster-causing data set by using a feedforward neural network, extracting the sequence characteristics of dynamic data in the disaster-causing data set by using a long-short term memory network (LSTM) and a multi-head self-attention mechanism, establishing a dual-channel prediction model of the power distribution network fault under the typhoon disaster, solving and optimizing model parameters based on the disaster-causing data set after sample equalization processing by combining a cross entropy loss function, an error back propagation method and the like, finally obtaining the optimized dual-channel prediction model, and evaluating the performance of the optimized dual-channel prediction model. And 4, if the performance meets the requirements, performing the step4, otherwise, continuing to perform optimization.
And 4, collecting data corresponding to a certain research area under the future typhoon disaster, constructing a disaster causing data set, inputting the disaster causing data set into the two-channel prediction model optimized in the step3, and obtaining a prediction value of the fault condition of the power distribution network in the research area under the future typhoon disaster.
The specific process of each step is as follows:
1. constructing disaster-causing data sets
The method selects disaster-causing data from the four angles of meteorological information, geographic information, power grid information and population information, divides the disaster-causing data into static data and dynamic data according to the time domain change condition of the data during the typhoon crossing period, forms a sample of a disaster-causing data set together, and finally forms the disaster-causing data set by combining with the fault condition type of the power distribution network under the typhoon disaster.
In the time scale of typhoon passing, part of disaster causing data basically does not change, and the influence on the fault condition of the power distribution network is stable. Therefore, the disaster-causing data is classified into static data, including four kinds of data, namely forest coverage, land type, maintenance degree of a power grid and population density. And part of disaster-causing data has larger change along with time, and has time-varying property and accumulation on the influence of the fault condition of the power distribution network. Therefore, the disaster-causing data is classified into dynamic data which comprises eight kinds of data including the distance between the center of the typhoon and the center of the area, the central lowest air pressure of the typhoon, the near-central maximum wind speed of the typhoon, the moving direction angle of the typhoon, the radius of a seven-level wind ring, the average wind speed of the area and the precipitation of the area. It is noted that static data consists of data for a single time slice, and dynamic data consists of sequential data for 48 hours.
Considering that typhoon disasters are accompanied by strong wind and strong rainfall, and electric power elements such as overhead lines, underground cables, towers and the like of a power distribution network can be damaged to a certain extent, the method sums permanent tripping times of the power distribution network for 24 continuous hours under the typhoon disasters, considers that the regional power distribution network normally operates when the sum of the tripping times is 0, considers that the regional power distribution network has slight faults when the sum of the tripping times is 1-9, and considers that the regional power distribution network has severe faults when the sum of the tripping times is more than 9, and takes the three disaster-suffering situation types of the power distribution network fault situation as the labels of the disaster-causing data sets.
2. In summary, the invention combines static and dynamic data to form a disaster-causing data set sample, and uses the type of the power distribution network fault condition as a data set sample label to jointly form a final disaster-causing data set, and divides the final disaster-causing data set into a disaster-causing training set and a disaster-causing testing set according to the proportion of eight to two. A schematic representation of the specimen and specimen label is shown in FIG. 1, where f1、f2、f3And f4For static data, f,i，jDynamic data, i-5, 6, … … 12; j ═ 1,2, … … 48; item i the j hour of dynamic data, nLO,kThe number of trips in the kth hour, k is 1,2, … … 24; and (5) equalizing the disaster-causing data sets.
Typhoon, one of extreme natural disasters, has low probability of occurrence and limited coverage area. Therefore, the disaster-causing data set has the largest number of samples of the normal operation class, the second number of samples of the light fault class and the smallest number of samples of the heavy fault class, that is, the light fault class samples and the heavy fault class samples are both the few class samples.
The sample imbalance phenomenon of the disaster-causing data set causes the power distribution network fault condition prediction model to lack learning aiming at the minority samples in the training process, and finally the prediction accuracy of the power distribution network fault condition prediction model on the minority samples is low. Considering that the cost-sensitive learning method has certain subjectivity and the parameter adjustment process is complicated, the method reduces the imbalance degree of the disaster-causing data set based on the Borderline-SMOTE1 algorithm and verifies the quality of the generated few samples through the discriminant model.
1) Borderline-SMOTE1 sample generation algorithm
The common SMOTE algorithm in the oversampling technology has larger blindness and randomness when selecting a target sample for sample generation, and is easy to generate a new sample which has no meaning or interference to a defined decision boundary. Therefore, based on the Borderline-SMOTE1 algorithm, the method divides the minority samples according to the type distribution characteristics around the minority samples, selects the minority samples close to the decision boundary to generate the samples, and reduces the imbalance degree of the disaster-causing data set. The algorithm steps of Borderline-SMOTE1 are explained by taking the generation of light fault class samples as an example, and the generation process of heavy fault class samples is the same. It should be noted that the sample generation algorithm is only applied to the disaster-causing training set.
Step 1: calculating m nearest neighbor samples of each light fault type sample by using a K nearest neighbor algorithm;
step 2: according to the proportion of the light failure samples in the m nearest neighbor samples of the light failure class samples, the light failure samples are classified into the following three classes, and the classification schematic diagram is shown in fig. 2.
(1) Safety sample: more than half of the nearest neighbor samples are light failure samples, such as the sample A in FIG. 2;
(2) dangerous type samples: less than half of the nearest neighbor samples are light failure samples, such as the B samples in fig. 2;
(3) noise-like samples: the nearest neighbor sample has no light fault sample, such as the C sample in fig. 2;
step 3: for each hazard class sample xiSelecting a required number of light fault samples from K nearest neighbor samples;
step 4: for each selected neighbor sample x'jGenerating a light fault class new sample x using linear interpolationi,jThe calculation formula is as follows:
xi,j＝xi+γ(x′j-xi) (1)
wherein γ is a random number between 0 and 1.
Step 5: and adding the generated mild fault new sample into the original disaster-causing training set.
2) Inspecting disaster-causing data sets
Considering that the data distribution of the disaster-causing training set is artificially changed by adding the few samples, when the quality of the generated samples is low, the sample distribution difference of the disaster-causing training set and the disaster-causing testing set is increased, and further the generalization capability of the prediction model on the disaster-causing testing set is reduced. Therefore, the present invention designs a discriminant model, performs sample distribution inspection on the disaster-causing training set and the disaster-causing testing set after the generated samples are added, and adjusts the parameter setting of the sample generation method according to the inspection result, and the specific principle is shown in fig. 3. The specific process of the discriminant model for detecting the sample distribution difference is described below.
(1) And (3) construction of a discrimination data set: the discrimination data set is based on the thought of self-supervision learning, and the sample division conditions of the training set and the test set are used as the label source of the discrimination data set. Considering that the number of samples in the disaster-causing training set is generally multiple times of the number of samples in the disaster-causing test set, random sampling is performed on the disaster-causing training set, so that the number of samples in the training set after sampling is equal to that in the disaster-causing test set. Then, labels of the training set samples and the test set samples are respectively set to be 0 and 1, a discrimination data set is formed by mixing, and the discrimination data set is divided into a new training set and a new test set according to the ratio of 8: 2.
(2) And (3) judging the training process of the model: and on the basis of the new training set and the new testing set, the cross entropy function is used as a loss function, the gradient of each parameter value of the discrimination model is obtained by an error back propagation method, and then all parameters of the discrimination model are updated by an Adam gradient descent algorithm to obtain the discrimination accuracy of the disaster-causing training set and the disaster-causing testing set.
(3) Analyzing the test result of the discrimination model: and measuring the sample distribution difference of the disaster-causing training set and the disaster-causing testing set by using the capability of distinguishing the disaster-causing training set and the disaster-causing testing set by using the discrimination model. When the judgment accuracy is higher than the accuracy threshold, the difference of the sample distribution of the disaster-causing training set and the disaster-causing test set is large, and the disaster-causing training set needs to be reconstructed, namely parameters such as the nearest neighbor sample number of the Borderline-SMOTE1 algorithm are adjusted; when the judgment accuracy is lower than the accuracy threshold, the sample distribution difference of the two is small, and the method can be directly used for training and testing a prediction model. Wherein, the accuracy threshold of the discriminant model is generally set to 70%.
3. Constructing a two-channel prediction model
In order to consider the stability of the action of static data and the time-varying property and the accumulative property of the action of dynamic data, the invention provides an interpretable neural network architecture which is used for respectively extracting the characteristics of the static data and the dynamic data and further establishing the mapping relation between the static data and the fault condition type of a power distribution network under the typhoon disaster. The following describes the feature extraction process of static and dynamic data and the training method of the two-channel prediction model in detail.
3.1 feed-forward neural network-based static feature extraction
The feedforward neural network is composed of an input layer, a hidden layer and an output layer, all layers of neurons are connected, and an in-layer connection structure and a cross-layer connection structure do not exist, so that the information transmission process of the feedforward neural network is unidirectional. In consideration of the stability of the static data of the typhoon disaster on the fault condition of the power distribution network, the static characteristics of the static data which is kept unchanged within 48 hours are extracted layer by adopting a multi-layer feedforward neural network.
3.2 dynamic feature extraction based on Long-short term memory network and Multi-headed attention mechanism
Unlike feed-forward neural networks, long-short term memory (LSTM) networks not only transfer information from layer to layer, but also within the same layer. With the addition of such an intra-layer connection structure, the LSTM has "memory" and "transitivity" in the processing of data. Meanwhile, the unit structure of the LSTM comprises a plurality of gate structures, so that the problems of gradient disappearance and gradient explosion caused by an interlayer connection structure can be effectively solved.
Each unit of the LSTM comprises three gate structures, namely a forgetting gate, an input gate and an output gate, and long-term memory and short-term memory processed by the gate structures are simultaneously transferred when information is transferred between the same layers, wherein the unit structure of the LSTM is shown in FIG. 4. LSTM inputs information x according to current timetAnd short-term memory of the last moment ht-1Respectively calculating the forgetting gate control signal ftGate control signal i is inputtAnd output gate control signal ot：
ft＝σ(Ufxt+Wfht-1+bf) (2)
it＝σ(Uixt+Wiht-1+bi) (3)
ot＝σ(Uoxt+Woht-1+bo) (4)
Where σ denotes Sigmoid activation function, UfFor the current input xtWeight of connection with forgetting gate structure, UiFor the current input xtConnection weights to input gate structure, UOFor the current input xtConnection weight to output gate structure, WfShort term memory h for the last momentt-1Connection weight with forget-to-open structure, WiShort term memory h for the last momentt-1Connection weights to input gate structure, WOShort term memory h for the last momentt-1Connection weights to output gate structures, bfFor biasing of forgetting gate structures, biFor biasing of the input gate structure, boIs the biasing of the output gate structure.
LSTM inputs information x to current moment based on three gating signalstAnd short-term memory of the last moment ht-1Performing reprocessing to update the long-term memory ctAnd short term memory htThe specific calculation formula is as follows:
in the formula (I), the compound is shown in the specification,
to be candidate for long-term memory, UcFor inputting information and candidate long-term memory
Connection weight of WcFor short-term memory and candidate long-term memory
Connection weight of bcTo be candidate for long-term memory
Is input offset.
In order to further enhance the feature extraction capability of the network to the dynamic data, the invention adopts a multi-head attention mechanism, utilizes a plurality of mapping subspaces to extract key components in the known data in an omnibearing and multi-angle manner, and maximizes the utilization of the known data information. The multi-head attention mechanism firstly maps data Q to multiple subspaces, and calculates the relevance and dependency between the data by using the self-attention formula attention (Q). Head corresponding to the ith headiThe specific calculation formula of (Q) is as follows:
headi(Q)＝Attention(QWi Q,QWi K,QWi V) (9)
in the formula (d)QFor the dimension of the input data Q, i is 1,2i Q、Wi K、Wi VRespectively, the subspace transformation matrix corresponding to the ith head.
The outputs of all headers are then stitched and mapped by linear layers to a final attention-weighted value, i.e., multihead (q):
MultiHead(Q)＝Concat(head1,...,headh)WO (10)
in which Concat is the splicing operation, WoThe mapping matrix is output.
The method firstly utilizes the LSTM network to extract the dynamic data characteristics, then adds the multi-head attention mechanism layer to the LSTM network, and further extracts the deep dynamic data characteristics in the dynamic data, thereby laying a foundation for establishing the final mapping relationship.
3.3 network Structure and training method of two-channel prediction model
The method comprises the steps of processing static data by adopting a feedforward neural network, processing dynamic data by adopting an LSTM network enhanced by a multi-head self-attention mechanism, finally splicing deep features extracted by the two, mapping the deep features into prediction probabilities of various fault condition types of the power distribution network through a linear layer, taking the maximum probability value corresponding to a disaster-suffered type as a prediction fault condition type of a sample, and obtaining a network structure of a prediction model as shown in figure 5. Corresponding batch normalization layers and nonlinear activation functions need to be added behind the first linear layer of the feedforward neural network, and convergence of the prediction model is improved.
Because the prediction of the fault condition types of the power distribution network under the typhoon disaster belongs to the classification problem, the invention uses the cross entropy function as the loss function on the basis of the disaster-causing training set to measure the difference degree between the predicted value and the actual value. And then obtaining the gradient value of each parameter in the cross entropy function pair model through an error back propagation algorithm. And finally, updating the parameters of the prediction model by using a small-batch Adam algorithm according to the learning rate, the batch size, the number of neurons in each layer and other hyper-parameters.
3.4 evaluation method of two-channel prediction model
The prediction of the fault condition types of the power distribution network is a three-classification problem, and the sample quantity of each class of the disaster-causing test set is unequal. In order to relieve the dominant effect of most types of sample evaluation results on the prediction accuracy and comprehensively consider the performance of the prediction model in each type, the invention takes precision ratio, recall ratio and F1 measurement as a basic index system, and introduces a macro-average mechanism to comprehensively consider the performance of the prediction model in different types of sample sets in a disaster-causing test set, and the specific process is described as follows.
Firstly, according to a predicted value obtained after a disaster-causing test set is input into a prediction model, counting whether each sample in the disaster-causing test set belongs to an actual value and a predicted value of the disaster-causing type, and forming three binary confusion moments in total. After the matrix is formed, a group of true positive TPs corresponding to each confusion matrix is obtained according to the matrix elementsiFalse positive FPiTrue negative TNiAnd false negative FNiFurther obtain the corresponding precision ratio PiAnd recall ratio Ri. And finally, obtaining three indexes of macro precision macro-P, macro recall macro-R and macro F1 value macro-F1 according to a macro average mechanism, and comprehensively measuring the performance of the prediction model, wherein a specific calculation formula is as follows.
In consideration of the fact that the prediction accuracy can visually highlight the performance of the model, the performance of the power distribution network fault condition prediction model under the typhoon disaster is evaluated by selecting four indexes of macro precision, macro recall, macro F1 and accuracy.
4. Prediction with dual channel prediction model
The method comprises the steps of collecting meteorological prediction data issued by meteorological departments before typhoon passing, geographic data, population data and power grid data of each research area, constructing a corresponding disaster-causing data set, inputting the disaster-causing data set into a two-channel prediction model after parameter optimization, and obtaining a prediction value of the fault condition type of each research area power distribution network under the future typhoon disaster.
Example 2
As shown in fig. 6, the device for predicting the grid fault in the typhoon disaster provided by the invention comprises an acquisition module and a calculation output module;
the acquisition module is used for acquiring data and transmitting the acquired data to the calculation output module; the data comprises historical dynamic data, static data and real-time typhoon data, and the real-time typhoon data comprises dynamic data and static data.
And the calculation output module is used for training a prediction model according to the historical dynamic data, the static data and the sum of the permanent trip times of the power grid in the predicted area, and then outputting a power grid fault prediction value according to the prediction model, the real-time dynamic data and the static data.
Example 3
As shown in fig. 7, the computer device provided by the present invention includes a memory and a processor electrically connected to each other, wherein the memory stores a computing program executable on the processor, and the processor executes the computing program to implement the steps of the prediction method.
Example 4
The prediction means, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, all or part of the flow of the method according to the embodiments of the present invention may also be implemented by a computer program, which may be stored in a computer-readable storage medium, and when the computer program is executed by a processor, the steps of the method embodiments may be implemented. Wherein the computer program comprises computer program code, which may be in the form of source code, object code, an executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, usb disk, removable hard disk, magnetic disk, optical disk, computer memory, Read-only memory (ROM), Random Access Memory (RAM), electrical carrier wave signals, telecommunications signals, software distribution medium, etc. It should be noted that the computer readable medium may contain content that is subject to appropriate increase or decrease as required by legislation and patent practice in jurisdictions, for example, in some jurisdictions, computer readable media does not include electrical carrier signals and telecommunications signals as is required by legislation and patent practice.
The invention discloses a universal power distribution network fault condition prediction model under typhoon disasters, and the prediction is carried out by using the method disclosed by the invention, so that the inherent unbalanced problem of a data set can be effectively reduced, and the quality of a generated sample is improved. Meanwhile, the prediction method considers the stability of the static data action and the accumulative property of the dynamic data action, further improves the accuracy and the interpretability of the prediction model, and provides more accurate prediction information for the power distribution network to deal with the typhoon disasters.
The above-mentioned contents are only for illustrating the technical idea of the present invention, and the protection scope of the present invention is not limited thereby, and any modification made on the basis of the technical idea of the present invention falls within the protection scope of the claims of the present invention.
Claims (9)
1. The method for predicting the power grid fault under the typhoon disaster with double drives of data is characterized by comprising the following steps of:
step1, collecting multivariate influence data of power grid faults under typhoon disasters and the sum of permanent tripping times of a power grid in a predicted area, dividing the data into static data and dynamic data according to time domain variation attributes of the data, and constructing a disaster-causing data set by utilizing the static data, the dynamic data and the sum of the permanent tripping times of the power grid in the predicted area;
step2, carrying out equalization processing on the disaster-causing data set;
step3, extracting the characteristics of static data in the disaster-causing data set by using a feedforward neural network, extracting the sequence characteristics of dynamic data in the disaster-causing data set by using a long-short term memory network and a multi-head self-attention mechanism, establishing a dual-channel prediction model of the power grid fault under the typhoon disaster, and solving and adjusting the model parameters based on the disaster-causing data set after sample equalization processing to finally obtain an optimized dual-channel prediction model; and evaluating the performance of the device; if the performance meets the requirements, performing the step4, otherwise, continuing to perform optimization;
and 4, collecting corresponding multivariate influence data of the prediction region under the future typhoon disaster, constructing a disaster causing data set, inputting the disaster causing data set into the two-channel prediction model optimized in the step3, and obtaining a prediction value of the power grid fault condition of the research region under the future typhoon disaster.
2. The method for predicting the power grid fault under the data dual-drive typhoon disaster according to claim 1, wherein in the step1, static data comprise forest coverage, land types, maintenance degree and population density of the power grid, and dynamic data comprise the distance between the typhoon center and the regional center, the central minimum air pressure of the typhoon, the near-central maximum wind speed of the typhoon, the moving direction angle of the typhoon, the seven-level wind circle radius, the average wind speed of the predicted region and the precipitation of the predicted region.
3. The method for predicting the power grid fault under the data dual-drive typhoon disaster according to claim 1, wherein the process of the step2 is as follows: dividing a minority sample set according to distribution of disaster-causing data sets in a high-dimensional space by using a Borderline-SMOTE1 algorithm, and generating samples according to the minority samples at a decision boundary after division; and then, checking the difference of the data distribution of the training set and the test set through a discriminant model, performing parameter tuning on the Borderline-SMOTE1 algorithm according to the difference, and finally, applying the Borderline-SMOTE1 algorithm after parameter optimization to balance the disaster-causing data set.
4. The method for predicting the power grid fault under the data dual-drive typhoon disaster according to claim 1, wherein the step2 comprises the following steps:
step 2.1, calculating m nearest neighbor samples of each light fault sample by using a K nearest neighbor algorithm;
2.2, according to the proportion of the light fault samples in the m nearest neighbor samples of the light fault samples, classifying the light fault samples into safety samples, dangerous samples and noise samples;
step 2.3, aiming at each danger class sample xiSelecting a required number of light fault samples from K nearest neighbor samples;
step 2.4, for each selected neighbor sample x'jGeneration of mild fault class novels using linear interpolationSample xi,j；
Step 2.5, adding the generated mild fault type new sample into the original disaster-causing training set to obtain an updated disaster-causing data set;
and 2.6, checking the updated disaster causing data set, performing the step3 if the updated disaster causing data set meets the requirement, and performing parameter adjustment on the Borderline-SMOTE1 algorithm if the updated disaster causing data set does not meet the requirement until the updated disaster causing data set meets the requirement.
5. The method for predicting the power grid fault under the data dual-drive typhoon disaster according to claim 4, wherein the step 2.6 comprises the following steps:
step 2.6.1, randomly sampling the disaster-causing training set to ensure that the number of samples of the sampled training set is equal to that of the samples of the disaster-causing test set; setting labels of the training set samples and the test set samples as 0 and 1 respectively, mixing to form a discrimination data set, and dividing the discrimination data set into a new training set and a new test set according to a proportion;
step 2.6.2, on the basis of the new training set and the new testing set, taking a cross entropy function as a loss function, obtaining the gradient of each parameter value of the discrimination model through an error back propagation method, and further updating all parameters of the discrimination model through an Adam gradient descent algorithm to obtain the discrimination accuracy of the disaster-causing training set and the disaster-causing testing set;
step 2.6.3, measuring the sample distribution difference of the disaster-causing training set and the disaster-causing testing set by using the ability of distinguishing the disaster-causing training set and the disaster-causing testing set by using a distinguishing model, and adjusting parameters such as the nearest neighbor sample number of the Borderline-SMOTE1 algorithm when the distinguishing accuracy is higher than the accuracy threshold; and when the judgment accuracy is lower than the accuracy threshold, executing the step 3.
6. The method for predicting the power grid fault under the data dual-drive typhoon disaster according to claim 1, wherein the step3 comprises the following steps:
3.1, extracting static characteristics from the static data based on a feedforward neural network; extracting dynamic features from the dynamic data based on a long-short term memory network and a multi-head attention mechanism;
step 3.2, splicing the static characteristics and the dynamic characteristics, mapping the static characteristics and the dynamic characteristics into prediction probabilities of all fault condition types of the power grid through a linear layer, and obtaining a prediction model by taking a maximum probability value corresponding to a disaster-suffered type as a prediction fault condition type of a sample; measuring the difference degree between the predicted value and the actual value by using a cross entropy function as a loss function; then obtaining the gradient value of each parameter in the cross entropy function pair model through an error back propagation algorithm; finally, updating the parameters of the prediction model by using a small-batch Adam algorithm according to the learning rate, the batch size and the number of neurons in each layer;
and 3.3, taking precision ratio and recall ratio as a basic index system, introducing a macro-average mechanism, comprehensively considering the performance of the prediction model in different types of sample sets in the disaster-causing test set, and evaluating the prediction model.
7. The method for predicting the power grid fault under the data dual-drive typhoon disaster according to claim 1, wherein the step 3.3 comprises the following steps:
step 3.3.1, according to the predicted value obtained after the disaster causing test set is input into the prediction model, counting whether each sample in the disaster causing test set belongs to the actual value and the predicted value of the disaster type, and forming three two-class confusion moments;
step 3.3.2, obtaining a group of true positive TPs corresponding to each confusion matrix according to the matrix elementsiFalse positive FPiTrue negative TNiAnd false negative FNiFurther obtain the corresponding precision ratio PiAnd recall ratio Ri；
Step 3.3.3, according to the precision ratio PiRecall ratio RiAnd F1 to obtain macro precision macro-P, macro recall macro-R and macro F1 value macro-F1;
and 3.3.4, evaluating the performance of the power grid fault condition prediction model under the typhoon disaster according to the four indexes of the macro precision ratio, the macro F1 and the accuracy.
8. A power grid fault prediction device under typhoon disasters is characterized by comprising:
the acquisition module is used for acquiring data and transmitting the acquired data to the calculation output module; the data comprises multivariate influence data of power grid faults under typhoon disasters, the sum of permanent tripping times of the power grid in a predicted area and real-time typhoon data;
and the calculation output module is used for training a prediction model according to the acquired data set and outputting a power grid fault prediction value according to the prediction model and the real-time typhoon data.
9. A computer device, comprising: electrically connected memory and a processor, the memory having stored thereon a computing program operable on the processor, when executing the computing program, implementing the steps of the method of any of claims 1-8.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111139000.6A CN113837477B (en) | 2021-09-27 | 2021-09-27 | Method, device and equipment for predicting power grid faults under typhoon disasters driven by data |
| PCT/CN2022/081314 WO2023045278A1 (en) | 2021-09-27 | 2022-03-17 | Data dual-drive method, apparatus, and device for predicting power grid failure during typhoon |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111139000.6A CN113837477B (en) | 2021-09-27 | 2021-09-27 | Method, device and equipment for predicting power grid faults under typhoon disasters driven by data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113837477A true CN113837477A (en) | 2021-12-24 |
| CN113837477B CN113837477B (en) | 2023-06-27 |
Family
ID=78970684
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111139000.6A Active CN113837477B (en) | 2021-09-27 | 2021-09-27 | Method, device and equipment for predicting power grid faults under typhoon disasters driven by data |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN113837477B (en) |
| WO (1) | WO2023045278A1 (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023045278A1 (en) * | 2021-09-27 | 2023-03-30 | 西安交通大学 | Data dual-drive method, apparatus, and device for predicting power grid failure during typhoon |
| CN116016219A (en) * | 2022-12-20 | 2023-04-25 | 缀初网络技术(上海)有限公司 | Edge cloud server loss prediction method and device |
| CN117034755A (en) * | 2023-08-07 | 2023-11-10 | 兰州理工大学 | Cold-rolled steel mechanical property prediction method integrating multi-head attention mechanism |
| CN117216485A (en) * | 2023-11-09 | 2023-12-12 | 国网山东省电力公司电力科学研究院 | Objective weighting-based power transmission wave-recording bird damage fault judging method and system |
Families Citing this family (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113537552A (en) * | 2021-04-28 | 2021-10-22 | 桂林电子科技大学 | Student performance prediction method based on mixed deep learning and attention mechanism |
| CN116522764B (en) * | 2023-04-17 | 2023-12-19 | 华中科技大学 | Hot wave-flood composite disaster assessment method considering influence of climate change |
| CN116629465B (en) * | 2023-07-26 | 2024-01-12 | 李波 | Smart power grids video monitoring and risk prediction response system |
| CN117390948A (en) * | 2023-08-10 | 2024-01-12 | 苏州黑盾环境股份有限公司 | Multi-head attention long-short-term memory neural network based water chilling unit monitoring method |
| CN116910493B (en) * | 2023-09-12 | 2023-12-08 | 山东能源数智云科技有限公司 | Construction method and device of equipment fault diagnosis model based on multi-source feature extraction |
| CN116956197B (en) * | 2023-09-14 | 2024-01-19 | 山东理工昊明新能源有限公司 | Deep learning-based energy facility fault prediction method and device and electronic equipment |
| CN116933981B (en) * | 2023-09-15 | 2023-12-08 | 安徽方能电气技术有限公司 | Regional power stability analysis and evaluation method based on power outage and restoration event data monitoring |
| CN117422438B (en) * | 2023-10-07 | 2024-03-29 | 国家电网有限公司华东分部 | Method and device for determining reinforcement scheme of power transmission line |
| CN117056402B (en) * | 2023-10-12 | 2024-04-02 | 国网浙江省电力有限公司余姚市供电公司 | Motor diagnosis method and device based on multi-source signals and storage medium |
| CN117056734B (en) * | 2023-10-12 | 2024-02-06 | 山东能源数智云科技有限公司 | Method and device for constructing equipment fault diagnosis model based on data driving |
| CN117347803A (en) * | 2023-10-25 | 2024-01-05 | 爱科特科技(海南)有限公司 | Partial discharge detection method, system, equipment and medium |
| CN117150243B (en) * | 2023-10-27 | 2024-01-30 | 湘江实验室 | Fault isolation and estimation method based on fault influence decoupling network |
| CN117495147B (en) * | 2023-12-22 | 2024-03-22 | 中国石油大学(华东) | Crack network expansion intelligent prediction method considering intersegmental fracturing process difference |
| CN117742240A (en) * | 2023-12-28 | 2024-03-22 | 广州和兴机电科技有限公司 | Remote monitoring method and system of numerical control machine tool |
| CN117522177B (en) * | 2024-01-08 | 2024-03-12 | 国网江苏省电力有限公司信息通信分公司 | Smart power grid stability prediction method |
| CN117828490B (en) * | 2024-03-06 | 2024-05-17 | 南京信息工程大学 | Typhoon disaster forecasting method and system based on ensemble learning |
| CN117909719B (en) * | 2024-03-19 | 2024-05-24 | 勤源(江苏)科技有限公司 | Terminal internet of things equipment monitoring and early warning method and system based on autonomous learning algorithm |
| CN117973440B (en) * | 2024-04-02 | 2024-06-14 | 长江三峡集团实业发展(北京)有限公司 | Regional ionosphere delay prediction method based on LSTM-transducer model |
| CN117992861B (en) * | 2024-04-04 | 2024-06-21 | 国网湖北省电力有限公司 | Electric power data accuracy checking method and system |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101354757A (en) * | 2008-09-08 | 2009-01-28 | 中国科学院地理科学与资源研究所 | Method for predicting dynamic risk and vulnerability under fine dimension |
| US20170093905A1 (en) * | 2014-12-29 | 2017-03-30 | Cyence Inc. | Disaster Scenario Based Inferential Analysis Using Feedback for Extracting and Combining Cyber Risk Information |
| CN109814527A (en) * | 2019-01-11 | 2019-05-28 | 清华大学 | Based on LSTM Recognition with Recurrent Neural Network industrial equipment failure prediction method and device |
| JP2020042705A (en) * | 2018-09-13 | 2020-03-19 | いすゞ自動車株式会社 | Fault prediction apparatus, fault prediction method, and program |
| CN111191832A (en) * | 2019-12-25 | 2020-05-22 | 国电南瑞科技股份有限公司 | Typhoon disaster power distribution network tower fault prediction method and system |
| US20200285900A1 (en) * | 2019-03-06 | 2020-09-10 | Wuhan University | Power electronic circuit fault diagnosis method based on optimizing deep belief network |
| CN113191585A (en) * | 2021-03-23 | 2021-07-30 | 广东电网有限责任公司东莞供电局 | Typhoon disaster risk assessment method for power transmission line |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9262124B2 (en) * | 2011-11-21 | 2016-02-16 | International Business Machines Corporation | Natural disaster forecasting |
| CN113837477B (en) * | 2021-09-27 | 2023-06-27 | 西安交通大学 | Method, device and equipment for predicting power grid faults under typhoon disasters driven by data |
-
2021
- 2021-09-27 CN CN202111139000.6A patent/CN113837477B/en active Active
-
2022
- 2022-03-17 WO PCT/CN2022/081314 patent/WO2023045278A1/en unknown
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101354757A (en) * | 2008-09-08 | 2009-01-28 | 中国科学院地理科学与资源研究所 | Method for predicting dynamic risk and vulnerability under fine dimension |
| US20170093905A1 (en) * | 2014-12-29 | 2017-03-30 | Cyence Inc. | Disaster Scenario Based Inferential Analysis Using Feedback for Extracting and Combining Cyber Risk Information |
| JP2020042705A (en) * | 2018-09-13 | 2020-03-19 | いすゞ自動車株式会社 | Fault prediction apparatus, fault prediction method, and program |
| CN109814527A (en) * | 2019-01-11 | 2019-05-28 | 清华大学 | Based on LSTM Recognition with Recurrent Neural Network industrial equipment failure prediction method and device |
| US20200285900A1 (en) * | 2019-03-06 | 2020-09-10 | Wuhan University | Power electronic circuit fault diagnosis method based on optimizing deep belief network |
| CN111191832A (en) * | 2019-12-25 | 2020-05-22 | 国电南瑞科技股份有限公司 | Typhoon disaster power distribution network tower fault prediction method and system |
| CN113191585A (en) * | 2021-03-23 | 2021-07-30 | 广东电网有限责任公司东莞供电局 | Typhoon disaster risk assessment method for power transmission line |
Non-Patent Citations (5)
| Title |
|---|
| HUILIN ZHENG等: "A Stacking Ensemble Prediction Model for the Occurrences of Major Adverse Cardiovascular Events in Patients With Acute Coronary Syndrome on Imbalanced Data", IEEE ACCESS * |
| 张广平;张晨晓;谢忠;: "基于T-S模糊神经网络的模型在台风灾情预测中的应用――以海南为例", 灾害学, no. 02 * |
| 张远汀;龚伟伟;叶钰;徐希源;徐勋建;蔡泽林;陆佳政;韩俊浩;叶飞;许婧;: "应用机器学习技术预测强雨雪天气过程中的积雪", 科学技术与工程, no. 15 * |
| 李重桂;李录平;刘瑞;杨波;陈茜;邓子豪;: "风电机组智能状态评估与故障预测研究进展", 电站系统工程, no. 04 * |
| 郑凌铭;舒胜文;陈彬;吴涵;黄建业;钱健;: "强台风环境下基于格点化和支持向量机的10 kV杆塔受损量预测方法", 高电压技术, no. 01 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023045278A1 (en) * | 2021-09-27 | 2023-03-30 | 西安交通大学 | Data dual-drive method, apparatus, and device for predicting power grid failure during typhoon |
| CN116016219A (en) * | 2022-12-20 | 2023-04-25 | 缀初网络技术(上海)有限公司 | Edge cloud server loss prediction method and device |
| CN117034755A (en) * | 2023-08-07 | 2023-11-10 | 兰州理工大学 | Cold-rolled steel mechanical property prediction method integrating multi-head attention mechanism |
| CN117216485A (en) * | 2023-11-09 | 2023-12-12 | 国网山东省电力公司电力科学研究院 | Objective weighting-based power transmission wave-recording bird damage fault judging method and system |
| CN117216485B (en) * | 2023-11-09 | 2024-01-30 | 国网山东省电力公司电力科学研究院 | Objective weighting-based power transmission wave-recording bird damage fault judging method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113837477B (en) | 2023-06-27 |
| WO2023045278A1 (en) | 2023-03-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113837477B (en) | Method, device and equipment for predicting power grid faults under typhoon disasters driven by data | |
| Moradi Kordmahalleh et al. | A sparse recurrent neural network for trajectory prediction of atlantic hurricanes | |
| CN102496069B (en) | Cable multimode safe operation evaluation method based on fuzzy analytic hierarchy process (FAHP) | |
| CN115270965B (en) | Power distribution network line fault prediction method and device | |
| CN111680875B (en) | Unmanned aerial vehicle state risk fuzzy comprehensive evaluation method based on probability baseline model | |
| CN110414412B (en) | Wide-area power grid multiple disturbance accurate identification method and device based on big data analysis | |
| CN114548637A (en) | AHP-RST-based power communication backbone data network security comprehensive evaluation method | |
| CN106707099A (en) | Monitoring and locating method based on abnormal electricity consumption detection module | |
| CN113807570B (en) | XGBoost-based reservoir dam risk level assessment method and system | |
| CN108133225A (en) | A kind of icing flashover fault early warning method based on support vector machines | |
| CN112070128A (en) | Transformer fault diagnosis method based on deep learning | |
| CN110636066B (en) | Network security threat situation assessment method based on unsupervised generative reasoning | |
| CN114186641B (en) | Landslide susceptibility evaluation method based on deep learning | |
| CN115220133B (en) | Rainfall prediction method, device and equipment for multiple meteorological elements and storage medium | |
| CN111581880A (en) | Rolling bearing residual life prediction method based on APSO-DBN | |
| CN115267945A (en) | Thunder and lightning early warning method and system based on graph neural network | |
| CN110737976A (en) | mechanical equipment health assessment method based on multi-dimensional information fusion | |
| CN115130375A (en) | Rock burst intensity prediction method | |
| CN111723523B (en) | Estuary surplus water level prediction method based on cascade neural network | |
| CN112070103B (en) | Method for inverting atmospheric visibility through microwave link network gridding self-adaptive variable scale | |
| CN116109039A (en) | Data-driven anomaly detection and early warning system | |
| CN112541636A (en) | Power transmission line icing risk early warning method and device, medium and electronic equipment | |
| CN116738192A (en) | Digital twinning-based security data evaluation method and system | |
| CN112257914A (en) | Aviation safety cause and effect prediction method based on random forest | |
| CN115049026A (en) | Regression analysis method of space non-stationarity relation based on GSNNR |
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 |