CN113850017A - System-level fault analysis system and method based on power flow change map - Google Patents
System-level fault analysis system and method based on power flow change map Download PDFInfo
- Publication number
- CN113850017A CN113850017A CN202110994173.XA CN202110994173A CN113850017A CN 113850017 A CN113850017 A CN 113850017A CN 202110994173 A CN202110994173 A CN 202110994173A CN 113850017 A CN113850017 A CN 113850017A
- Authority
- CN
- China
- Prior art keywords
- fault
- power grid
- module
- flow change
- change map
- 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
- 230000008859 change Effects 0.000 title claims abstract description 94
- 238000004458 analytical method Methods 0.000 title claims abstract description 26
- 238000000034 method Methods 0.000 title claims abstract description 20
- 238000003745 diagnosis Methods 0.000 claims abstract description 45
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 25
- 238000013473 artificial intelligence Methods 0.000 claims abstract description 9
- 238000005259 measurement Methods 0.000 claims abstract description 4
- 230000007246 mechanism Effects 0.000 claims abstract description 3
- 238000012549 training Methods 0.000 claims description 17
- 238000013135 deep learning Methods 0.000 claims description 16
- 230000006870 function Effects 0.000 claims description 16
- 238000012545 processing Methods 0.000 claims description 16
- 238000010586 diagram Methods 0.000 claims description 14
- 238000005457 optimization Methods 0.000 claims description 13
- 238000013528 artificial neural network Methods 0.000 claims description 10
- 238000005516 engineering process Methods 0.000 claims description 9
- 230000001052 transient effect Effects 0.000 claims description 9
- 230000010354 integration Effects 0.000 claims description 7
- 238000012544 monitoring process Methods 0.000 claims description 7
- 238000004040 coloring Methods 0.000 claims description 6
- 238000010801 machine learning Methods 0.000 claims description 6
- 210000002569 neuron Anatomy 0.000 claims description 6
- 230000003044 adaptive effect Effects 0.000 claims description 4
- 238000010276 construction Methods 0.000 claims description 4
- 239000011159 matrix material Substances 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 238000012800 visualization Methods 0.000 claims description 4
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 3
- 230000003993 interaction Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 2
- 238000006243 chemical reaction Methods 0.000 claims description 2
- 238000004891 communication Methods 0.000 claims description 2
- 238000013527 convolutional neural network Methods 0.000 claims description 2
- 229910003460 diamond Inorganic materials 0.000 claims description 2
- 239000010432 diamond Substances 0.000 claims description 2
- 230000002452 interceptive effect Effects 0.000 claims description 2
- 238000011176 pooling Methods 0.000 claims description 2
- 238000004088 simulation Methods 0.000 claims description 2
- 230000011664 signaling Effects 0.000 abstract 1
- 238000004590 computer program Methods 0.000 description 7
- 230000005540 biological transmission Effects 0.000 description 4
- 230000008447 perception Effects 0.000 description 4
- 238000003860 storage Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000005520 cutting process Methods 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 238000007635 classification algorithm Methods 0.000 description 1
- 230000003749 cleanliness Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000013075 data extraction Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000010355 oscillation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- 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
- G06N3/08—Learning methods
-
- 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
- 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/04—Power grid distribution networks
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Evolutionary Computation (AREA)
- Entrepreneurship & Innovation (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Artificial Intelligence (AREA)
- Operations Research (AREA)
- Development Economics (AREA)
- Quality & Reliability (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Game Theory and Decision Science (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Educational Administration (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Cheminformatics (AREA)
Abstract
The invention discloses a system-level fault analysis system and method based on a power flow change map.A fault diagnosis model is constructed by using an artificial intelligence algorithm, the fault power flow change map is used as the input of the model, the power grid fault corresponding to the fault power flow change map is used as the output, and the model is trained through a large number of samples to generate a power grid fault diagnosis model; meanwhile, periodically acquiring power grid section data, comparing power grid section data of adjacent periods to generate a power flow change map, inputting the power flow change map into a power grid fault diagnosis model, judging whether a diagnosis result of equipment fault exists or not, and issuing a power grid system level alarm if the diagnosis result is that the power grid has the fault. Compared with the traditional alarm based on a power grid model and an operation mechanism, the method is faster mainly depending on remote signaling deflection and remote measurement change, does not depend on a single signal to cause leakage or error alarm, is organically combined with the traditional alarm result, and improves the alarm accuracy.
Description
Technical Field
The invention belongs to the field of power system automation, and particularly relates to a system-level fault analysis system and method based on a power flow change map.
Background
With the leading energy pattern towards cleanliness, the power grid will become the main platform for future energy configuration. The electric load of China is mainly concentrated in the middle east area, and the energy resource is mainly concentrated in the three-north area and the southwest area. The ultra-high voltage power grid has the characteristics of large transmission capacity, long transmission distance and the like, and the capacity of optimizing and configuring resources across the power grid can be fully exerted by developing the ultra-high voltage.
The formation of the extra-high voltage alternating current and direct current hybrid power grid supports the large-scale optimized configuration of electric power energy in China, and simultaneously, due to the significant change of the grid pattern and the multiple occurrence of extreme natural disasters, the safe operation situation of the power grid faces more severe challenges, the difficulty of dispatching and fault disposal is obviously increased, and the method mainly shows the following aspects: the whole operation of the power grid presents an integrated characteristic, the impact global characteristic of a fault on the system is obvious, and the fault perception difficulty is obviously increased; the loss of extra-high voltage direct current high power causes huge impact on each level of power grid, and the combined handling efficiency of each level of scheduling faults needs to be improved; the regulation and control personnel generate a large amount of valuable information in the process of processing the power grid accident, the information is not deeply mined, and the information is not interacted and issued in real time.
In conclusion, the speed and the accuracy of power grid fault diagnosis are improved, and the analysis of system-level faults is the key for supporting the safe operation of the extra-high voltage alternating current and direct current hybrid power grid.
Disclosure of Invention
The invention aims to provide a system-level fault analysis system and method based on a tidal current change map, which are used for realizing system-level fault perception and diagnosis of a power grid based on a support platform of an intelligent power grid dispatching technology and solving the problem of fault analysis of an extra-high voltage alternating current-direct current hybrid power grid.
In order to achieve the purpose, the system level fault analysis system based on the power flow change map comprises a deep learning module based on the power flow change map, a power grid system level fault sensing module and a power grid event warning information merging and sorting module; the deep learning module based on the power flow change map is used for generating a key equipment fault sample set, forming a fault power flow change map corresponding to the key equipment fault sample set, constructing a fault diagnosis model by using an artificial intelligence algorithm, taking the fault power flow change map as the input of the model, taking the power grid fault corresponding to the fault power flow change map as the output, and training the model through a large number of samples to generate a power grid fault diagnosis model; the power grid system level fault sensing module periodically acquires power grid section data based on a power grid fault diagnosis model, compares the power grid section data of adjacent periods to generate a power flow change map, inputs the power flow change map into the power grid fault diagnosis model, judges whether a diagnosis result of equipment fault exists or not, and issues a power grid system level alarm if the diagnosis result is that the power grid has the fault; the power grid event alarm information merging and sorting module is used for summarizing power grid steady-state transient operation information based on the power grid system level alarm issued by the power grid system level fault sensing module as a trigger and merging the information as the associated detailed information of the system level alarm.
A system-level fault analysis method based on a power flow change map comprises the following steps:
generating a key equipment fault sample set by adopting a deep learning module based on a power flow change map, and forming a fault power flow change map corresponding to the key equipment fault sample set;
the deep learning module based on the power flow change map constructs a fault diagnosis model by using an artificial intelligence algorithm, the fault power flow change map is used as the input of the model, the power grid fault corresponding to the fault power flow change map is used as the output, and the model is trained through a large number of samples to generate a power grid fault diagnosis model;
the power grid system level fault sensing module periodically acquires power grid section data based on the power grid fault diagnosis model, compares the power grid section data of adjacent periods to generate a power flow change map, inputs the power flow change map into the power grid fault diagnosis model, judges whether a diagnosis result of equipment fault exists or not, and issues a power grid system level alarm if the diagnosis result is that the power grid has the fault;
the power grid event alarm information merging and sorting module is used for summarizing power grid steady-state transient operation information based on the power grid system level alarm issued by the power grid system level fault sensing module as a trigger and merging the information as the associated detailed information of the system level alarm.
The invention has the beneficial effects that: based on the deep learning technology, the power grid system level fault sensing and analysis are carried out through the power grid tide change map, the power grid fault alarm studying and judging speed and accuracy are improved, and the power grid fault sensing capability is improved.
Drawings
FIG. 1 is a schematic diagram of the system of the present invention;
FIG. 2 is a schematic diagram of power flow change map generation of a power grid;
FIG. 3 is a schematic diagram of a method for adaptive fault event set modeling;
FIG. 4 is a network architecture diagram of VGG 11;
the system comprises a deep learning module based on a power flow change map, a 2-power grid system level fault perception module, a 3-power grid event alarm information merging and sorting module, an 11-adaptive fault event set modeling module, a 12-fault power flow change map drawing module and a 13-deep neural network training module.
Detailed Description
The invention is described in further detail below with reference to the following figures and specific examples:
a system level fault analysis system based on a power flow change map is shown in figure 1 and comprises a deep learning module 1 based on the power flow change map, a power grid system level fault perception module 2 and a power grid event alarm information merging and sorting module 3;
the deep learning module 1 based on the power flow change map is used for generating a key equipment fault sample set, forming a fault power flow change map corresponding to the key equipment fault sample set, constructing a fault diagnosis model by using an artificial intelligence algorithm, taking the fault power flow change map as the input of the model, taking the power grid fault corresponding to the fault power flow change map as the output, training the model through a large number of samples, and generating a power grid fault diagnosis model;
the power grid system level fault sensing module 2 periodically acquires power grid section data based on a power grid fault diagnosis model, compares the power grid section data of adjacent periods to generate a power flow change map, inputs the power flow change map into the power grid fault diagnosis model, judges whether a diagnosis result of equipment fault exists or not, and issues a power grid system level alarm if the diagnosis result is that the power grid has the fault;
the power grid event alarm information merging and sorting module 3 is used for summarizing power grid steady-state transient operation information based on the power grid system level alarm issued by the power grid system level fault sensing module 2 as a trigger and merging the information as the associated detailed information of the system level alarm.
In the above technical scheme, the power grid steady-state transient operation information includes power grid frequency information, section flow out-of-limit information, line trip information and unit trip information.
In the technical scheme, the deep learning module 1 based on the trend change map comprises a self-adaptive fault event set modeling module 11, a fault trend change map drawing module 12 and a deep neural network training module 13;
the self-adaptive fault event set modeling module 11 is used for calculating key equipment N-1 and N-2 by adopting an off-line simulation and on-line safety analysis calculation mode based on an on-line safety analysis module of a smart grid dispatching technology support platform to generate a key equipment fault sample set;
the fault current change map drawing module 12 is used for generating a power grid fault current change map, and the specific implementation method comprises model map construction and primitive coloring; the model graph construction is that different primitives represent different types of equipment to form a power grid model formed in a primitive form; the primitive coloring is used for simulating the fault of key equipment, coloring the primitive according to the size of the power flow change, and abstracting the power grid power flow change into a fault power flow change map easy for machine learning;
the deep neural network training module 13 constructs a fault prediction model by using an artificial intelligence algorithm, takes the fault power flow change map as the input of the fault prediction model, takes the power grid fault corresponding to the power flow change map as the output, trains the fault prediction model through a large number of samples, and generates a power grid fault diagnosis model.
In the above technical solution, the method for generating the key device fault sample set includes:
in order to collect a fault set to the maximum extent, simulating a plurality of power grid operating states, simulating a fault of a certain device, recording active power changes of all primary devices of the power grid before and after the fault, forming a characteristic value corresponding to the fault, and simulating N-1 and N-2 faults of all key devices in different power grid operating states; meanwhile, collecting experience in actual work to form a fault event set;
the power grid operation state comprises a maximum operation mode, a minimum operation mode, load changes at different peak-valley moments and a power grid state of grid topology changes;
the specific implementation method for constructing the model diagram comprises the following steps: different primitives represent different types of equipment, and a pentagon represents a line, a circle represents a load, a diamond represents a main transformer and a square represents a unit; performing matrix layout on all the devices subjected to graphical conversion, and arranging all the devices in a certain plant station in a line in a primitive form; the equipment types from left to right are a line, a load, a main transformer and a unit in sequence, and each type of equipment is sequenced from left to right according to the voltage level from high to low; counting the quantity of each type of equipment of each plant, and taking the maximum value of the quantity of the equipment of a single plant as the column number of the equipment; all the stations are arranged in multiple rows to form a power grid model formed in a primitive form. Therefore, the dimension reduction processing is carried out on the actual multidimensional variable, the processing efficiency is improved, and the interference factor is reduced.
In the above technical solution, the deep neural network training module 13 includes a deep network model selection module, a model loss function and optimization algorithm module, and a machine learning model training module; the deep network model selection module is used for more accurately and abstractly automatically extracting characteristics from the power grid flow change map; the model loss function and optimization algorithm module is used for reducing the overfitting risk in the power grid fault type classification process; the machine learning model training module trains based on the characteristics of the power grid flow change map and generates a model classifier.
In the above technical solution, the deep network model selection module adopts a VGG network model, the VGG network model is constructed by repeatedly iterating a convolutional layer and a pooling layer on the basis of CNN, the VGG network model comprises 11 layers in total, wherein, the 1 st layer to the 8 th layer are convolution layers, the 9 th layer to the 11 th layer are full connection layers, the deep network model selection module divides the power grid fault types into 10 categories, in order to enable the VGG network model to meet the classification requirements and prevent the network from overfitting, the full connection layer of 512 neurons is adopted to replace the 9 th full connection layer, the full connection layer containing 256 neurons is adopted to replace the 10 th full connection layer, the full connection layer containing 10 neurons is adopted to replace the 11 th full connection layer, the output of the SoftMax classifier in the deep neural network training module 13 corresponds to 10 labels, and the network structure of the SoftMax classifier is shown in fig. 4.
In the technical scheme, the model loss function and the model loss function adopted by the optimization algorithm module are cross entropy loss functions, and regularization terms are added to reduce the risk of overfitting;
the model loss function and optimization algorithm module adopts an optimization algorithm which is an adam (adaptive motion estimation) optimization algorithm, is an optimization algorithm capable of replacing the traditional random to replace the random gradient descent, and can iteratively update the weight of the neural network based on training data. Adam estimates and dynamically adjusts the learning rate of each parameter by utilizing a first-order matrix and a second-order matrix of the gradient, and has the main advantages that after bias correction, the learning rate of each iteration has a range, so that the parameters do not have large oscillation but relatively stable change;
the deep network model selection module adopts a Relief algorithm. The power grid has multidimensional variables, a large amount of data can be generated during power grid operation at each moment, the variable data contain a large amount of meaningless interference data, effective dimensionality reduction of the power grid data is necessary, and a Relief algorithm is selected in consideration of the fact that a variable feature extraction method is beneficial to implementation of a classification algorithm.
In the technical scheme, the power grid event warning information merging and sorting module 3 comprises a multi-channel data access module, a multi-thread data processing module, a fault information integration module and a multi-screen linkage and visualization large-screen pushing and displaying module;
the multi-channel data access module accesses data of various channels by adopting various data access modes; the multithreading data processing module uses a multithreading processing technology to improve the processing speed and efficiency of fault information; the fault information integration module integrates basic fault information, important section information and multi-form fault information of a user-defined monitoring object, which are acquired based on a smart grid dispatching technology support platform; aiming at the operating characteristics of an extra-high voltage direct current receiving end power grid, the multi-screen linkage and visualization large-screen pushing and displaying module integrates power grid operation key information such as frequency coordination control, dynamic ACE (adaptive communication index), key section monitoring and the like after a fault and pushes the key information to a scheduling mechanism within a fault influence range.
The power grid event warning information merging and sorting module 3 is mainly used for merging and sorting power grid information after an extra-high voltage alternating current/direct current single/bipolar locking fault. Under the condition of high-power transmission of extra-high voltage alternating current and direct current, if a blocking fault occurs, severe impact can be caused to a power grid at a transmitting end and a receiving end, a power grid stability control device at the transmitting end is subjected to a series of influences such as a power cutting machine, a frequency of the power grid at the receiving end is greatly dropped, a section is overloaded, voltage fluctuation and tide is greatly transferred, and a dispatcher is required to quickly and accurately deal with the fault. The merging and integrated display technology of the alarm key information is particularly important for improving the fault handling efficiency.
In the technical scheme, aiming at the current situation of fault information data access, in order to improve the data reliability and the access efficiency, the data access is carried out in a mode of file, message bus, real-time library reading and wide area event service.
The multiple data access modes are file interaction, message bus, real-time library reading and wide area event service; the data adopting the file interactive data access mode comprises thunder and lightning, mountain fire and meteorological information, the data adopting the message bus data access mode comprises protection data and WAMS data, and the data adopting the real-time library read data access mode comprises relevant measurement data of the fault equipment.
Aiming at thunder and lightning, mountain fire and meteorological information, because data are stored in a meteorological system and other comprehensive data networks, the data are acquired in a file interaction acquisition mode in consideration of small data volume and data cross-region transmission. And collecting the protection data and the WAMS data in a II area message bus mode. The warning information of the SCADA monitoring system is interacted in real time through an I area message bus, and the related measurement data of the fault equipment is acquired through a real-time library interface through multiple incidence relations among stations, graphs and equipment and ID information.
In the technical scheme, the fault information integration is to integrate the basic fault information, the important section information and the fault information of the customized monitoring object in multiple forms. For example, a user can customize information of a sending-end power grid unit safety control device cutting unit, receiving-end power grid system frequency, key sections before and after a fault, key voltage, key power plant tide situation and the like. After actual failure, the failure data processing service performs automatic data extraction and information integration according to failure equipment and user-defined configuration conditions and pushes a human-computer interface.
A system-level fault analysis method based on a power flow change map comprises the following steps:
firstly, a key equipment fault sample set is generated by adopting a deep learning module based on a power flow change map, and a fault power flow change map corresponding to the key equipment fault sample set is formed, as shown in figure 2,
secondly, the deep learning module based on the power flow change map constructs a fault diagnosis model by using an artificial intelligence algorithm, the fault power flow change map is used as the input of the model, the power grid fault corresponding to the fault power flow change map is used as the output, and the model is trained through a large number of samples to generate a power grid fault diagnosis model, as shown in fig. 3;
thirdly, the power grid system level fault sensing module periodically acquires power grid section data based on the power grid fault diagnosis model, compares the power grid section data of adjacent periods to generate a power flow change map, inputs the power flow change map into the power grid fault diagnosis model, judges whether a diagnosis result of equipment fault exists or not, and issues a power grid system level alarm if the diagnosis result is that the power grid has the fault;
and finally, the power grid event alarm information merging and sorting module is used for summarizing the steady-state transient operation information of the power grid based on the power grid system level alarm issued by the power grid system level fault sensing module as a trigger and merging the steady-state transient operation information as the associated detailed information of the system level alarm.
Details not described in this specification are within the skill of the art that are well known to those skilled in the art. The computer environment that this patent software realized recommends CPU3.2GHz, memory 16GB, hard disk 250GB or higher configuration.
As will be appreciated by one skilled in the art, 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 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 flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams 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.
Claims (10)
1. A system-level fault analysis system based on a power flow change map is characterized in that: the power grid event warning information merging and sorting system comprises a deep learning module (1) based on a trend change map, a power grid system level fault sensing module (2) and a power grid event warning information merging and sorting module (3);
the deep learning module (1) based on the power flow change map is used for generating a key equipment fault sample set, forming a fault power flow change map corresponding to the key equipment fault sample set, constructing a fault diagnosis model by using an artificial intelligence algorithm, taking the fault power flow change map as the input of the model, taking the power grid fault corresponding to the fault power flow change map as the output, training the model through a large number of samples, and generating a power grid fault diagnosis model;
the power grid system level fault sensing module (2) periodically acquires power grid section data based on a power grid fault diagnosis model, compares the power grid section data of adjacent periods to generate a power flow change map, inputs the power flow change map into the power grid fault diagnosis model, judges whether a diagnosis result of equipment fault exists or not, and issues a power grid system level alarm if the diagnosis result is that the power grid has the fault;
the power grid event alarm information merging and sorting module (3) is used for summarizing power grid steady-state transient operation information based on the power grid system level alarm issued by the power grid system level fault sensing module (2) as a trigger and merging the information as the associated detailed information of the system level alarm.
2. The system-level fault analysis system based on the power flow change map as claimed in claim 1, wherein: the deep learning module (1) based on the power flow change map comprises a self-adaptive fault event set modeling module (11), a fault power flow change map drawing module (12) and a deep neural network training module (13);
the self-adaptive fault event set modeling module (11) is based on an online safety analysis module of a smart grid dispatching technology support platform, and calculates key equipment N-1 and N-2 by adopting an offline simulation and online safety analysis calculation mode to generate a key equipment fault sample set;
the fault current change map drawing module (12) is used for generating a power grid fault current change map, and the specific implementation method comprises model map construction and primitive coloring; the model graph construction is that different primitives represent different types of equipment to form a power grid model formed in a primitive form; the primitive coloring is used for simulating the fault of key equipment, coloring the primitive according to the size of the power flow change, and abstracting the power grid power flow change into a fault power flow change map easy for machine learning;
the deep neural network training module (13) utilizes an artificial intelligence algorithm to construct a fault prediction model, the fault power flow change map is used as the input of the fault prediction model, the power grid fault corresponding to the power flow change map is used as the output, and the fault prediction model is trained through a large number of samples to generate a power grid fault diagnosis model.
3. The system-level fault analysis system based on the power flow change map as claimed in claim 2, wherein: the method for generating the key equipment fault sample set comprises the following steps:
simulating various power grid operating states, simulating a fault of a certain device, recording active power changes of all primary devices of the power grid before and after the fault, forming a characteristic value corresponding to the fault, and simulating N-1 and N-2 faults of all key devices in different power grid operating states; meanwhile, collecting experience in actual work to form a fault event set;
the power grid operation state comprises a maximum operation mode, a minimum operation mode, load changes at different peak-valley moments and a power grid state of grid topology changes;
the specific implementation method for constructing the model diagram comprises the following steps: different primitives represent different types of equipment, and a pentagon represents a line, a circle represents a load, a diamond represents a main transformer and a square represents a unit; performing matrix layout on all the devices subjected to graphical conversion, and arranging all the devices in a certain plant station in a line in a primitive form; the equipment types from left to right are a line, a load, a main transformer and a unit in sequence, and each type of equipment is sequenced from left to right according to the voltage level from high to low; counting the quantity of each type of equipment of each plant, and taking the maximum value of the quantity of the equipment of a single plant as the column number of the equipment; all the stations are arranged in multiple rows to form a power grid model formed in a primitive form.
4. The system-level fault analysis system based on the power flow change map as claimed in claim 2, wherein: the deep neural network training module (13) comprises a deep network model selection module, a model loss function and optimization algorithm module and a machine learning model training module; the deep network model selection module is used for automatically extracting features from the power grid flow change map; the model loss function and optimization algorithm module is used for reducing the overfitting risk in the power grid fault type classification process; the machine learning model training module trains based on the characteristics of the power grid flow change map and generates a model classifier.
5. The system-level fault analysis system based on the power flow change map as claimed in claim 4, wherein:
the deep network model selection module adopts a VGG network model, the VGG network model is constructed through iteration convolution layer and pooling layer on the basis of CNN, the VGG network model contains 11 layers totally, wherein, the 1 st layer to the 8 th layer are convolution layer, the 9 th layer to the 11 th layer are full-connection layer, the deep network model selection module divides the grid fault type into 10 categories, adopts 512 full-connection layer of neurons to replace the 9 th layer full-connection layer, adopts full-connection layer containing 256 neurons to replace the 10 th layer full-connection layer, adopts full-connection layer containing 10 neurons to replace the 11 th layer full-connection layer, the output of SoftMax classifier in the deep neural network training module (13) corresponds to 10 labels.
6. The system-level fault analysis system based on the power flow change map as claimed in claim 4, wherein:
the model loss function and the model loss function adopted by the optimization algorithm module are cross entropy loss functions, and regularization items are added to reduce the risk of overfitting;
the model loss function and optimization algorithm module adopts an optimization algorithm which is an Adam optimization algorithm, and the deep network model selection module adopts a Relief algorithm.
7. The system-level fault analysis system based on the power flow change map as claimed in claim 1, wherein:
the power grid steady-state transient state operation information comprises power grid frequency information, section tide out-of-limit information, line trip information and unit trip information;
the power grid event alarm information merging and arranging module (3) comprises a multi-channel data access module, a multi-thread data processing module, a fault information integration module and a multi-screen linkage and visualization large-screen pushing and displaying module;
the multi-channel data access module accesses data of various channels by adopting various data access modes; the multithreading data processing module uses a multithreading processing technology to improve the processing speed and efficiency of fault information; the fault information integration module integrates basic fault information, important section information and multi-form fault information of a user-defined monitoring object, which are acquired based on a smart grid dispatching technology support platform; aiming at the operating characteristics of an extra-high voltage direct current receiving end power grid, the multi-screen linkage and visualization large-screen pushing and displaying module integrates power grid operation key information such as frequency coordination control, dynamic ACE (adaptive communication index), key section monitoring and the like after a fault and pushes the key information to a scheduling mechanism within a fault influence range.
8. The system-level fault analysis system based on the power flow change map as claimed in claim 7, wherein:
the multiple data access modes are file interaction, message bus, real-time library reading and wide area event service; the data adopting the file interactive data access mode comprises thunder and lightning, mountain fire and meteorological information, the data adopting the message bus data access mode comprises protection data and WAMS data, and the data adopting the real-time library read data access mode comprises relevant measurement data of the fault equipment.
9. The system-level fault analysis system based on the power flow change map as claimed in claim 7, wherein:
the fault information integration is to integrate basic fault information, important section information and fault information of multiple forms of the user-defined monitoring object.
10. A system-level fault analysis method based on a power flow change map is characterized by comprising the following steps: it comprises the following steps:
generating a key equipment fault sample set by adopting a deep learning module based on a power flow change map, and forming a fault power flow change map corresponding to the key equipment fault sample set;
the deep learning module based on the power flow change map constructs a fault diagnosis model by using an artificial intelligence algorithm, the fault power flow change map is used as the input of the model, the power grid fault corresponding to the fault power flow change map is used as the output, and the model is trained through a large number of samples to generate a power grid fault diagnosis model;
the power grid system level fault sensing module periodically acquires power grid section data based on the power grid fault diagnosis model, compares the power grid section data of adjacent periods to generate a power flow change map, inputs the power flow change map into the power grid fault diagnosis model, judges whether a diagnosis result of equipment fault exists or not, and issues a power grid system level alarm if the diagnosis result is that the power grid has the fault;
the power grid event alarm information merging and sorting module is used for summarizing power grid steady-state transient operation information based on the power grid system level alarm issued by the power grid system level fault sensing module as a trigger and merging the information as the associated detailed information of the system level alarm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110994173.XA CN113850017B (en) | 2021-08-27 | 2021-08-27 | System-level fault analysis system and method based on tide change map |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110994173.XA CN113850017B (en) | 2021-08-27 | 2021-08-27 | System-level fault analysis system and method based on tide change map |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113850017A true CN113850017A (en) | 2021-12-28 |
| CN113850017B CN113850017B (en) | 2024-04-02 |
Family
ID=78976316
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110994173.XA Active CN113850017B (en) | 2021-08-27 | 2021-08-27 | System-level fault analysis system and method based on tide change map |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113850017B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117560267A (en) * | 2024-01-11 | 2024-02-13 | 南京邮电大学 | Fault diagnosis method and rapid recovery method for multimode field network |
| CN117872038A (en) * | 2024-03-11 | 2024-04-12 | 浙江大学 | DC micro-grid instability fault source positioning method and device based on graph theory |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019233047A1 (en) * | 2018-06-07 | 2019-12-12 | 国电南瑞科技股份有限公司 | Power grid dispatching-based operation and maintenance method |
| CN111650921A (en) * | 2020-05-20 | 2020-09-11 | 国网江苏省电力有限公司泰州供电分公司 | Intelligent power grid regulation and control system equipment fault diagnosis method and system |
| CN111999614A (en) * | 2020-08-21 | 2020-11-27 | 国网河南省电力公司周口供电公司 | Local discharge multi-source fusion detection early warning method for high-voltage switch cabinet |
| WO2021108680A1 (en) * | 2019-11-25 | 2021-06-03 | Strong Force Iot Portfolio 2016, Llc | Intelligent vibration digital twin systems and methods for industrial environments |
| CN113268604A (en) * | 2021-05-19 | 2021-08-17 | 国网辽宁省电力有限公司 | Self-adaptive expansion method and system for knowledge base |
| CN113283602A (en) * | 2021-05-19 | 2021-08-20 | 国网辽宁省电力有限公司 | Power grid system fault diagnosis method based on offline knowledge learning and online dynamic correction |
-
2021
- 2021-08-27 CN CN202110994173.XA patent/CN113850017B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019233047A1 (en) * | 2018-06-07 | 2019-12-12 | 国电南瑞科技股份有限公司 | Power grid dispatching-based operation and maintenance method |
| WO2021108680A1 (en) * | 2019-11-25 | 2021-06-03 | Strong Force Iot Portfolio 2016, Llc | Intelligent vibration digital twin systems and methods for industrial environments |
| CN111650921A (en) * | 2020-05-20 | 2020-09-11 | 国网江苏省电力有限公司泰州供电分公司 | Intelligent power grid regulation and control system equipment fault diagnosis method and system |
| CN111999614A (en) * | 2020-08-21 | 2020-11-27 | 国网河南省电力公司周口供电公司 | Local discharge multi-source fusion detection early warning method for high-voltage switch cabinet |
| CN113268604A (en) * | 2021-05-19 | 2021-08-17 | 国网辽宁省电力有限公司 | Self-adaptive expansion method and system for knowledge base |
| CN113283602A (en) * | 2021-05-19 | 2021-08-20 | 国网辽宁省电力有限公司 | Power grid system fault diagnosis method based on offline knowledge learning and online dynamic correction |
Non-Patent Citations (1)
| Title |
|---|
| 白浩;李鹏;袁智勇;于力;姜臻;: "人工智能在配电网高阻接地故障检测中的应用及展望", 南方电网技术, no. 02, 20 February 2019 (2019-02-20) * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117560267A (en) * | 2024-01-11 | 2024-02-13 | 南京邮电大学 | Fault diagnosis method and rapid recovery method for multimode field network |
| CN117560267B (en) * | 2024-01-11 | 2024-03-26 | 南京邮电大学 | Fault diagnosis method and rapid recovery method for multimode field network |
| CN117872038A (en) * | 2024-03-11 | 2024-04-12 | 浙江大学 | DC micro-grid instability fault source positioning method and device based on graph theory |
| CN117872038B (en) * | 2024-03-11 | 2024-05-17 | 浙江大学 | DC micro-grid instability fault source positioning method and device based on graph theory |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113850017B (en) | 2024-04-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Xie et al. | A review of machine learning applications in power system resilience | |
| CN107330056B (en) | Wind power plant SCADA system based on big data cloud computing platform and operation method thereof | |
| CN107358366B (en) | Distribution transformer fault risk monitoring method and system | |
| CN108375715A (en) | A kind of distribution network line fault risk day prediction technique and system | |
| CN113850017B (en) | System-level fault analysis system and method based on tide change map | |
| CN107918830A (en) | A kind of distribution Running State assessment system and method based on big data technology | |
| CN114021848A (en) | Generating capacity demand prediction method based on LSTM deep learning | |
| CN118487276B (en) | Power grid safety dynamic management and control method and system for power guarantee object | |
| CN113283602A (en) | Power grid system fault diagnosis method based on offline knowledge learning and online dynamic correction | |
| CN115730749B (en) | Power dispatching risk early warning method and device based on fusion power data | |
| CN105574604A (en) | Power network operation event-oriented monitoring, pre-judging and analyzing system | |
| CN112612819A (en) | Big data analysis and mining method and system for pumped storage power station | |
| CN110361609A (en) | Extra-high voltage equipment monitors system and method | |
| CN113408132B (en) | Reliability evaluation method for regional autonomous power grid considering stochastic fault scene | |
| Wang et al. | A Comprehensive Review on the Development of Data-Driven Methods for Wind Power Prediction and AGC Performance Evaluation in Wind-Thermal Bundled Power Systems | |
| CN201666923U (en) | Device for processing DSP failure diagnosis of electric power system by using immune genetic algorithm | |
| Ma et al. | Research Progress of Distribution Network Digitization Based on Digital Twin Technology | |
| CN112000708A (en) | Abnormal data processing method and system based on regulation and control adapted data fusion | |
| Zhang et al. | Distributed Architecture of Power Grid Asset Management and Future Research Directions | |
| CN106651113A (en) | Dispatcher operation risk evaluation method based on advanced N-1 analysis | |
| CN111046633A (en) | LSTM-based power grid primary frequency modulation data prediction method and device | |
| Guan et al. | Enhancing security and resilience of bulk power systems via multisource big data learning | |
| CN114548762A (en) | Real-time power system cascading failure risk assessment method and system based on space-time diagram neural network | |
| CN113689064A (en) | Transparent transformer substation management system architecture | |
| Jingyu et al. | Statistical analysis of distribution network fault information based on multi-source heterogeneous data mining |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20220704 Address after: 100031 No. 86 West Chang'an Avenue, Beijing, Xicheng District Applicant after: STATE GRID CORPORATION OF CHINA Applicant after: STATE GRID ZHEJIANG ELECTRIC POWER Co.,Ltd. Applicant after: BEIJING KEDONG POWER CONTROL SYSTEM Co.,Ltd. Applicant after: NARI Group Corp. Address before: 100031 No. 86 West Chang'an Avenue, Beijing, Xicheng District Applicant before: STATE GRID CORPORATION OF CHINA Applicant before: STATE GRID ZHEJIANG ELECTRIC POWER Co.,Ltd. Applicant before: BEIJING KEDONG POWER CONTROL SYSTEM Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |