CN112446171B - Power system transient stability monitoring method and device, terminal equipment and storage medium - Google Patents
Power system transient stability monitoring method and device, terminal equipment and storage medium Download PDFInfo
- Publication number
- CN112446171B CN112446171B CN202011241068.0A CN202011241068A CN112446171B CN 112446171 B CN112446171 B CN 112446171B CN 202011241068 A CN202011241068 A CN 202011241068A CN 112446171 B CN112446171 B CN 112446171B
- Authority
- CN
- China
- Prior art keywords
- node
- power system
- state diagram
- time
- sequence
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000001052 transient effect Effects 0.000 title claims abstract description 62
- 238000000034 method Methods 0.000 title claims abstract description 36
- 238000012544 monitoring process Methods 0.000 title claims abstract description 32
- 238000010586 diagram Methods 0.000 claims abstract description 165
- 239000010410 layer Substances 0.000 claims description 74
- 238000005070 sampling Methods 0.000 claims description 28
- 239000013598 vector Substances 0.000 claims description 24
- 238000004590 computer program Methods 0.000 claims description 19
- 238000012549 training Methods 0.000 claims description 15
- 230000006870 function Effects 0.000 claims description 12
- 230000004913 activation Effects 0.000 claims description 7
- 239000011159 matrix material Substances 0.000 claims description 7
- 239000002356 single layer Substances 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 6
- 238000012806 monitoring device Methods 0.000 claims description 5
- 238000001514 detection method Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 238000004088 simulation Methods 0.000 claims description 4
- 230000008859 change Effects 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000007418 data mining Methods 0.000 description 2
- 230000001687 destabilization Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005381 potential energy Methods 0.000 description 1
- 230000003449 preventive effect Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
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
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2474—Sequence data queries, e.g. querying versioned data
-
- 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
- G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
-
- 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
- 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
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/12—Timing analysis or timing optimisation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Business, Economics & Management (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Water Supply & Treatment (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Probability & Statistics with Applications (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Databases & Information Systems (AREA)
- Geometry (AREA)
- Fuzzy Systems (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a method, a device, a terminal device and a storage medium for monitoring transient stability of an electric power system, wherein the method carries out time and space correlation learning through state variables of bus nodes and node admittance matrixes of the system on time sequences before, during and after a circuit system fault is input into a space-time diagram convolution network, so as to acquire the instability probability of each generator node and realize monitoring of the transient stability state of each generator.
Description
Technical Field
The invention relates to the technical field of electronic power, in particular to a transient stability monitoring method of a power system.
Background
In recent years, with the increasing demand of human society for electric power resources, the complexity and load of the power grid are also increasing. These present a great safety hazard to the power system, especially in terms of transient stability of the power system. The steady state of the power system is monitored in real time, so that timely and effective measures can be taken, and the steady and continuous operation of the power grid can be guaranteed. The rapid on-line transient stability judgment is a basic tool for providing unstable early warning and system preventive control guidance.
Several existing analysis methods for transient stability of a power system: time domain method (stability is judged by solving the change of phase power angles of each generator by solving differential equation); energy method (solving a stability domain by using Lyapunov energy function); BCU method (combination of potential energy boundary method and dominant unstable equilibrium point method); neural network based methods (using the characteristic quantities in the power system as inputs, building an input-output map using the fitting capabilities of the neural network). With the continuous increase of data volume, data-driven technology (machine learning) provides a new idea for analyzing and solving problems, and recent deep learning has a non-negligible effect in various fields. In the analysis of the power system, the conventional Convolutional Neural Network (CNN) and the Recurrent Neural Network (RNN) respectively face the problems of neglecting the connection between local information and the long time spent on training and practical use, the models do not effectively utilize the topological characteristics of the power system, and the topology change of the power grid can have great influence on the transient stability characteristics of the power grid, so that the generalization performance of the models is poor, and the prediction accuracy of the models can be greatly reduced after the topology of the power grid is changed. In addition, in the current data-driven power angle transient stability analysis model, the transient power angle stability or the power angle transient margin of the whole power system is predicted, each generator set cannot be accurately detected, and the stable state of each generator set cannot be monitored.
Disclosure of Invention
The embodiment of the invention aims to provide a method for monitoring transient stability of a power system, which solves the problem that in the prior art, the transient stability of the power system is not accurately monitored due to the fact that the topological characteristics of the power system are not considered in the transient stability analysis of the power system, and the stable state of each generator set cannot be monitored.
To achieve the above objective, an embodiment of the present invention provides a method for monitoring transient stability of an electric power system, including the following steps:
acquiring historical characteristic data acquired by an electric power system, wherein the system sampling is performed from the time before the occurrence of the fault to the time after the occurrence of the fault;
constructing an input state diagram sequence according to the collected historical characteristic data, wherein the input state diagram sequence is formed according to a state diagram of the power system corresponding to each sampling point on a time sequence from the occurrence of a fault to the occurrence of the fault, and the state diagram is constructed by taking a bus of the power system as a node;
according to the power angle value of the generator after the fault occurs, obtaining a steady state result of each generator as a label corresponding to each generator;
inputting the input state diagram sequence and the steady state result of each generator into a model training based on a space-time diagram convolution network to obtain a transient stability prediction model of the power system;
after the fault occurs, collecting characteristic data of the power system in different fault modes, constructing an input state diagram sequence, and inputting the input state diagram sequence into a transient stability prediction model of the power system to obtain the instability probability of each node;
and monitoring the stable state of each generator according to the instability probability of each node.
Further, the inputting the input state diagram sequence and the steady state result of each generator into a model training based on a space-time diagram convolution network, wherein the model based on the space-time diagram convolution network specifically comprises:
the space-time diagram convolution network consists of 2 space-time convolution blocks and 1 output layer, wherein the space-time convolution blocks comprise two layers of time convolution layers and one layer of space convolution layer, the space convolution layer is positioned in the middle of the two layers of time convolution layers, and the space convolution layer adopts space-domain-based diagram convolution.
Further, in the construction of the state diagram by taking a bus of the power system as a node, the state diagram is specifically constructed by the following steps:
(1) Define state diagram G t The power system state is used for describing the power system state at the moment t;
(2) Node set V of definition graph, representing busbar set in power system, node V i Is a 6-dimensional vector V bus ,θ bus ,P G ,Q G ,P L ,Q L ]The node voltage amplitude, the node phase angle, the node active power output, the node reactive power output, the node active load and the node reactive load are respectively represented;
(3) Defining a set E of edges, and given an admittance matrix of the power system nodes at the moment t, defining the following time y ij E when not equal to 0 ij E is E; when y is ij When the value of the sum is =0,wherein y is ij Marking the transadmittance of the node i and the node j;
the input state diagram sequence is specifically:
wherein G is prefault A state diagram representing the power system before a fault occurs,a state diagram sequence representing the power system during the fault occurrence process, comprising a state diagram corresponding to each sampling point in the time sequence of the fault occurrence, +.>The state diagram sequence of the power system after the fault is generated comprises a state diagram corresponding to each sampling point in the time sequence after the fault is generated.
Further, according to the power angle value of the generator after the fault occurs, a steady state marking result of each generator is obtained and is used as a label corresponding to each generator, and the method specifically includes:
acquiring a generator i corresponding to a maximum power angle and a generator j corresponding to a minimum power angle in all generators of the power system obtained at a preset moment after long-time simulation;
respectively counting the number of generators with absolute values of differences of power angles of other generators and generators i and j smaller than 180, and selecting more generators with absolute values of differences of power angles of other generators smaller than 180 from the generators i and j as reference power angle generators;
setting the stable state of the reference generator as not unstable, marking the stable state as 0, calculating the absolute value of the difference between the power angles of other generators and the power angle of the reference generator, setting the generator mark with the absolute value of the difference between the power angles of the reference generator and the reference generator larger than 180 as unstable, marking the generator mark with the absolute value of the difference between the power angles of the reference generator and the reference generator smaller than 180 as not unstable, and marking the generator mark with the absolute value of the difference between the power angles of the reference generator and the reference generator as 0.
Further, after the fault occurs, feature data of the power system in different fault modes are collected, an input state diagram sequence is constructed and input into a transient stability prediction model of the power system, and the destabilization probability of each node is obtained, which specifically comprises:
inputting the input state diagram sequence into a transient stability prediction model of the power system, and obtaining an output state diagram sequence through a 2-layer space-time convolution block;
for each node, extracting a characteristic vector set corresponding to the node on a state diagram corresponding to each sampling point in the output state diagram sequence, and splicing to obtain a characteristic length vector of each node;
and outputting the characteristic length vector of each node through a single-layer full-connection layer of the output layer and a Sigmoid activation function to obtain the instability probability of each node.
The embodiment of the invention also provides a device for monitoring the transient stability of the power system, which comprises the following components:
the system comprises a historical characteristic data acquisition module, a fault detection module and a fault detection module, wherein the historical characteristic data acquisition module is used for acquiring historical characteristic data acquired by a power system, and the system sampling is carried out from the time before the fault occurs to the time after the fault occurs;
the input construction module is used for constructing an input state diagram sequence according to the collected historical characteristic data, wherein the input state diagram sequence is formed according to a state diagram of the power system corresponding to each sampling point on a time sequence from the occurrence of a fault to the occurrence of the fault, and the state diagram is constructed by taking a bus of the power system as a node;
the label acquisition module is used for acquiring a steady state result of each generator according to the power angle value of the generator after the fault occurs and taking the steady state result as a label corresponding to each generator;
the model training module is used for inputting the input state diagram sequence and the steady state result of each generator into model training based on a space-time diagram convolution network to obtain a transient stability prediction model of the power system;
the node instability probability acquisition module is used for acquiring characteristic data of the power system in different fault modes after the fault occurs, constructing an input state diagram sequence, and inputting the input state diagram sequence into the transient stability prediction model of the power system to obtain the instability probability of each node;
and the state monitoring module is used for monitoring the stable state of each generator according to the instability probability of each node.
Further, in the model training module, the model based on the space-time diagram convolutional network specifically includes:
the space-time diagram convolution network consists of 2 space-time convolution blocks and 1 output layer, wherein the space-time convolution blocks comprise two layers of time convolution layers and one layer of space convolution layer, the space convolution layer is positioned between the two layers of time convolution layers, and the space convolution layer adopts space-domain-based diagram convolution.
Further, the node instability probability obtaining module is specifically configured to:
inputting the input state diagram sequence into a transient stability prediction model of the power system, and processing the input state diagram sequence through a 2-layer space-time convolution block to obtain an output state diagram sequence;
for each node, extracting a characteristic vector set corresponding to the node on a state diagram corresponding to each sampling point in the output state diagram sequence, and splicing to obtain a characteristic length vector of each node;
and outputting the characteristic length vector of each node through a single-layer full-connection layer of an output layer and a Sigmoid activation function to obtain the instability probability of each node.
As a preferred embodiment of the present invention, the present invention further provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, where the processor executes the computer program to implement the method for monitoring transient stability of a power system according to the embodiment of the present invention.
Another embodiment of the present invention provides a storage medium, where the computer readable storage medium includes a stored computer program, and when the computer program runs, the device where the computer readable storage medium is located is controlled to execute the method for monitoring transient stability of an electric power system according to the embodiment of the present invention.
Compared with the prior art, the method has the following beneficial effects:
(1) Compared with other neural networks, the method for monitoring the transient stability of the power system provided by the embodiment of the invention has the advantages that the transient stability prediction model of the power system is built based on the space-time diagram convolution network, the topology information and the state quantity change information of the power system can be fully mined and utilized based on the space-time diagram convolution network modeling, the data mining of the time correlation and the space correlation of the input characteristics can be carried out, the prediction performance and the generalization performance of the model are improved, the model can adapt to the topology change of the power system, the output result of the model is more accurate, and the instability probability of each node obtained by output is more accurate;
(2) According to the invention, the state variables of bus nodes before, during and after the fault and the node admittance matrix of the system are input into the space-time diagram convolution network to perform time and space correlation learning, so that the instability probability of the generator nodes on the state diagram is obtained, and the transient stable state of each generator is monitored;
(3) The modeling is performed based on the space-time diagram convolution network, wherein a space convolution layer of the space-time diagram convolution network adopts space-domain-based diagram convolution, so that the time overhead is small, the influence of edges on a state diagram of the electric power system on each node can be better mined, the instability probability result of each node is more accurate, and the stable state of each generator set is more accurately detected.
Drawings
FIG. 1 is a schematic flow chart of an embodiment of a method for monitoring transient stability of a power system according to the present invention;
fig. 2 is a schematic structural diagram of an embodiment of a transient stability monitoring device for an electric power system according to the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, fig. 1 is a schematic flow chart of an embodiment of a method for monitoring transient stability of a power system according to the present invention; the embodiment of the invention provides a method for monitoring transient stability of a power system, which comprises the following steps of S1-S6:
s1, acquiring historical characteristic data acquired by a power system, and acquiring the system sample from the time before the occurrence of the fault to the time after the occurrence of the fault.
In this embodiment, the historical data acquired by the power system is used as a learning sample set by a time domain simulation method or by using a WAMS system, the acquisition interval can be acquired according to a preset time interval, and the system samples are acquired from before the occurrence of the fault until after the occurrence of the fault, so that the transient stability of the power system can be analyzed on a time scale.
S2, constructing an input state diagram sequence according to the collected historical characteristic data, wherein the input state diagram sequence is formed according to a state diagram of the power system corresponding to each sampling point on a time sequence from the occurrence of a fault to the occurrence of the fault, and the state diagram is constructed by taking a bus of the power system as a node.
In this embodiment, given characteristic data of the bus and the generator from before to after the occurrence of the fault, the bus may be taken as a node, and a state diagram of the power system may be constructed so as to perform relevant learning on the characteristic data of the node from the time series based on the bus node.
Specifically, the state diagram is constructed by the following steps:
(1) Define state diagram G t The power system state is used for describing the power system state at the moment t;
(2) Node set for defining a graphV is the sum of the buses in the power system, and the node V i Is a 6-dimensional vector V bus ,θ bus ,P G ,Q G ,P L ,Q L ]The node voltage amplitude, the node phase angle, the node active power output, the node reactive power output, the node active load and the node reactive load are respectively represented;
(3) Defining a set E of edges, and given an admittance matrix of the power system nodes at the moment t, defining the following time y ij E when not equal to 0 ij E is E; when y is ij When the value of the sum is =0,wherein y is ij Marking the transadmittance of the node i and the node j;
the input state diagram sequence is specifically:
wherein G is prefault A state diagram representing the power system before a fault occurs,a state diagram sequence representing the power system during the fault occurrence process, comprising a state diagram corresponding to each sampling point in the time sequence of the fault occurrence, +.>The state diagram sequence of the power system after the fault is generated comprises a state diagram corresponding to each sampling point in the time sequence after the fault is generated.
It should be noted that each sampling point on the time sequence of occurrence of the fault is f 1 、f 2 、...、f N Each sampling point in the time sequence after the fault is generated is p 1 、p 2 、...、p m 。
And S3, acquiring a steady state result of each generator according to the power angle value of the generator after the fault occurs, and taking the steady state result as a label corresponding to each generator.
It should be noted that when the power system is running in steady state, all synchronous generators in the system run synchronously, i.e. the power angle is a steady value. After the system is disturbed, if the generator rotor can still recover synchronous operation after a period of motion change, namely the power angle reaches a stable value, the system is stable in the power angle, otherwise, the power angle is unstable. Therefore, according to the power angle value of the generator set predicted by the tide, the generator stability after an accident can be obtained by the following steps:
acquiring a generator i corresponding to a maximum power angle and a generator j corresponding to a minimum power angle in all generators of the power system obtained at a preset moment after long-time simulation;
respectively counting the number of generators with absolute values of differences of power angles of other generators and generators i and j smaller than 180, and selecting more generators with absolute values of differences of power angles of other generators smaller than 180 from the generators i and j as reference power angle generators;
setting the stable state of the reference generator as not unstable, marking the stable state as 0, calculating the absolute value of the difference between the power angles of other generators and the power angle of the reference generator, setting the generator mark with the absolute value of the difference between the power angles of the reference generator and the reference generator larger than 180 as unstable, marking the generator mark with the absolute value of the difference between the power angles of the reference generator and the reference generator smaller than 180 as not unstable, and marking the generator mark with the absolute value of the difference between the power angles of the reference generator and the reference generator as 0.
In this embodiment, the steady state flag of each generator is not particularly limited, and the steady state generator may be marked with 1, the unstable generator with 0, or other symbol marks.
S4, inputting the input state diagram sequence and the steady state result of each generator into a model training based on a space-time diagram convolution network to obtain a transient stability prediction model of the power system.
In this embodiment, the space-time diagram convolutional network (STGCN, full name Spatial Temporal Graph Convolutional Networks) is modeled based on busbar nodes of the system, the space-time diagram convolutional network is composed of 2 space-time convolutional blocks (Spatial-Temporal convolution block) and 1 Output Layer (Output Layer), the space-time convolutional blocks are composed of two layers of time convolutional layers and one Layer of space convolutional Layer, the space convolutional Layer adopts space-domain based diagram convolutional, so that the space-domain based diagram convolutional system has the advantage of small time overhead, and the influence of edges on each node can be well mined, so that the destabilization probability of each node can be accurately predicted.
And S5, after the fault occurs, collecting characteristic data of the power system in different fault modes, constructing an input state diagram sequence, and inputting the input state diagram sequence into a transient stability monitoring model of the power system to obtain the instability probability of each node.
In this embodiment, step S5 specifically includes:
inputting the input state diagram sequence into a transient stability prediction model of the power system, and obtaining an output state diagram sequence through a 2-layer space-time convolution block;
for each node, extracting a characteristic vector set corresponding to the node on a state diagram corresponding to each sampling point in the output state diagram sequence, and splicing to obtain a characteristic length vector of each node;
and outputting the characteristic length vector of each node through a single-layer full-connection layer of the output layer and a Sigmoid activation function to obtain the instability probability of each node.
The following describes the calculation of the power system state stability prediction model in detail according to the embodiment of the present invention:
step a: the following operation is performed on the first space-time convolution block;
step a-1: inputting the input sequence state diagram sequence T G l-1 Firstly, carrying out time convolution on the time dimension of each node by using a 1-dimensional convolution kernel, obtaining new characteristics of each node of each state diagram in the sequence after convolution, and updating the original state diagram sequence to form a new state diagram sequence T G l-1 (1) ；
Step a-2: for state diagram sequence T G l-1 (1) By applying spatial domain based graph convolution (Spatial Convolution), the convolution kernel extracts the spatial correlation between nodes through the node impedance matrix of each state graph and calculates new node characteristics, updates the state graph sequence to form a new state graph sequence T G l-1 (2) ；
Step a-3: for the sequence output by b, similar to the operation in a, the time convolution is carried out on the time dimension of each node, and a state diagram sequence T is output G l ；
Step b: 2 space-time convolution blocks are stacked, and the total input state diagram sequence of the model isProcessing to obtain state diagram sequence
Step c: extracting the state diagram sequence for the generator node iCharacteristic vectors corresponding to the generator node i in the state diagram corresponding to each sampling point are spliced to form characteristic length vectors, and a single prediction value between 0 and 1 is output after passing through a single-layer full-connection layer and a Sigmoid activation function to represent the probability of instability of the generator. All generator nodes share the fully connected layer.
And S6, monitoring the stable state of each generator according to the instability probability of each node.
The embodiment provided by the invention has the following beneficial effects:
(1) According to the power system transient stability monitoring method provided by the embodiment of the invention, the power system transient stability prediction model is constructed based on the space-time diagram convolution network, so that the topology information and the time-varying information of the state quantity of the power system can be fully mined and utilized, and the data mining on the time correlation and the space correlation of the input characteristics is carried out, so that the prediction performance and generalization of the model are improved, and the power system transient stability prediction model can be adapted to the topology change of the power system;
(2) According to the invention, the state variables of bus nodes before, during and after the fault and the node admittance matrix of the system are input into the space-time diagram convolution network to perform time and space correlation learning, so that the instability probability of the generator nodes on the state diagram is obtained, and the transient stable state of each generator is monitored;
(3) The modeling is performed based on the space-time diagram convolution network, wherein a space convolution layer of the space-time diagram convolution network adopts space-domain-based diagram convolution, so that the time overhead is small, the influence of edges on a state diagram of the electric power system on each node can be better mined, the instability probability result of each node is more accurate, and the stable state of each generator set is more accurately detected.
As a preferred embodiment of the present invention, please refer to fig. 2, fig. 2 is a schematic structural diagram of an embodiment of a transient stability monitoring device for an electric power system, which includes:
the historical characteristic data acquisition module 1 is used for acquiring historical characteristic data acquired by a power system, and the system sampling is carried out from the time before the occurrence of the fault to the time after the occurrence of the fault;
the input construction module 2 is configured to construct an input state diagram sequence according to the collected historical characteristic data, wherein the input state diagram sequence is formed according to a state diagram of the power system corresponding to each sampling point on a time sequence from the occurrence of a fault to the occurrence of the fault, and the state diagram is constructed by taking a bus of the power system as a node;
the tag acquisition module 3 is used for acquiring a steady state result of each generator according to the power angle value of the generator after the fault occurs, and taking the steady state result as a tag corresponding to each generator;
the model training module 4 is used for inputting the input state diagram sequence and the steady state result of each generator into model training based on a space-time diagram convolution network to obtain a transient stability prediction model of the power system;
the node instability probability acquisition module 5 is used for acquiring characteristic data of the power system in different fault modes after the fault occurs, constructing an input state diagram sequence, and inputting the input state diagram sequence into a transient stability prediction model of the power system to obtain the instability probability of each node;
and the state monitoring module 6 is used for monitoring the stable state of each generator according to the instability probability of each node.
Preferably, in the model training module 4, the model based on the space-time diagram convolutional network specifically includes:
the space-time diagram convolution network consists of 2 space-time convolution blocks and 1 output layer, wherein the space-time convolution blocks comprise two layers of time convolution layers and one layer of space convolution layer, the space convolution layer is positioned between the two layers of time convolution layers, and the space convolution layer adopts space-domain-based diagram convolution.
Specifically, the node instability probability obtaining module 5 is specifically configured to:
inputting the input state diagram sequence into a transient stability prediction model of the power system, and processing the input state diagram sequence through a 2-layer space-time convolution block to obtain an output state diagram sequence;
for each node, extracting a characteristic vector set corresponding to the node on a state diagram corresponding to each sampling point in the output state diagram sequence, and splicing to obtain a characteristic length vector of each node;
and outputting the characteristic length vector of each node through a single-layer full-connection layer of an output layer and a Sigmoid activation function to obtain the instability probability of each node.
The embodiment of the invention also provides terminal equipment. The apparatus includes: a processor, a memory, and a computer program stored in the memory and executable on the processor. The processor, when executing the computer program, implements the steps in the embodiments of the method for monitoring transient stability of the electric power system, for example, steps S1 to S6 shown in fig. 1.
The processor may be a central processing unit (Central Processing Unit, CPU), other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), off-the-shelf programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. The general purpose processor may be a microprocessor or the processor may be any conventional processor or the like, which is a control center of the terminal device, and which connects various parts of the entire terminal device using various interfaces and lines.
The memory may be used to store the computer program and/or module, and the processor may implement various functions of the terminal device by running or executing the computer program and/or module stored in the memory and invoking data stored in the memory. The memory may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program (such as a sound playing function, an image playing function, etc.) required for at least one function, and the like; the storage data area may store data (such as audio data, phonebook, etc.) created according to the use of the handset, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as a hard disk, memory, plug-in hard disk, smart Media Card (SMC), secure Digital (SD) Card, flash Card (Flash Card), at least one disk storage device, flash memory device, or other volatile solid-state storage device.
Wherein the terminal device integrated modules/units may be stored in a computer readable storage medium if implemented in the form of software functional units and sold or used as stand alone products. Based on such understanding, the present invention may implement all or part of the flow of the method of the above embodiment, or may be implemented by a computer program to instruct related hardware, where the computer program may be stored in a computer readable storage medium, and when the computer program is executed by a processor, the computer program may implement the steps of each of the method embodiments described above. Wherein the computer program comprises computer program code which may be in source code form, object code form, executable file or some intermediate form etc. The computer readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a U disk, a removable hard disk, a magnetic disk, an optical disk, a computer Memory, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), an electrical carrier signal, a telecommunications signal, a software distribution medium, and so forth.
It should be noted that the above-described apparatus embodiments are merely illustrative, and the units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. In addition, in the drawings of the embodiment of the device provided by the invention, the connection relation between the modules represents that the modules have communication connection, and can be specifically implemented as one or more communication buses or signal lines. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
While the foregoing is directed to the preferred embodiments of the present invention, it will be appreciated by those skilled in the art that changes and modifications may be made without departing from the principles of the invention, such changes and modifications are also intended to be within the scope of the invention.
Claims (9)
1. The method for monitoring the transient stability of the power system is characterized by comprising the following steps of:
acquiring historical characteristic data acquired by an electric power system, wherein the system sampling is performed from the time before the occurrence of the fault to the time after the occurrence of the fault;
constructing an input state diagram sequence according to the collected historical characteristic data, wherein the input state diagram sequence is formed according to a state diagram of the power system corresponding to each sampling point on a time sequence from the occurrence of a fault to the occurrence of the fault, and the state diagram is constructed by taking a bus of the power system as a node;
according to the power angle value of the generator after the fault occurs, obtaining a steady state result of each generator as a label corresponding to each generator;
inputting the input state diagram sequence and the steady state result of each generator into a model training based on a space-time diagram convolution network to obtain a transient stability prediction model of the power system;
after the fault occurs, collecting characteristic data of the power system in different fault modes, constructing an input state diagram sequence, and inputting the input state diagram sequence into a transient stability prediction model of the power system to obtain the instability probability of each node;
monitoring the stable state of each generator according to the instability probability of each node;
in the construction of the state diagram by taking a bus of a power system as a node, the state diagram is specifically constructed by the following steps:
(1) Define state diagram G t The power system state is used for describing the power system state at the moment t;
(2) Node set V of definition graph, representing busbar set in power system, node V i Is a 6-dimensional vector V bus ,θ bus ,P G ,Q G ,P L ,Q L ]The node voltage amplitude, the node phase angle, the node active power output, the node reactive power output, the node active load and the node reactive load are respectively represented;
(3) Defining a set E of edges, and given an admittance matrix of the power system nodes at the moment t, defining the following time y ij E when not equal to 0 ij E is E; when y is ij When the value of the sum is =0,wherein y is ij Marking the transadmittance of the node i and the node j; e, e ij Is the edge connecting node i and node j;
the input state diagram sequence is specifically:
wherein G is prefault A state diagram representing the power system before a fault occurs,a state diagram sequence representing the power system in the fault occurrence process, including a state diagram corresponding to each sampling point in the time sequence of fault occurrence,the state diagram sequence of the power system after the fault is generated comprises a state diagram corresponding to each sampling point in the time sequence after the fault is generated.
2. The method for monitoring transient stability of a power system according to claim 1, wherein said inputting said input state diagram sequence and steady state results of each of said generators into a model training based on a space-time diagram convolutional network, wherein said model based on a space-time diagram convolutional network specifically comprises:
the space-time diagram convolution network consists of 2 space-time convolution blocks and 1 output layer, wherein the space-time convolution blocks comprise two layers of time convolution layers and one layer of space convolution layer, the space convolution layer is positioned in the middle of the two layers of time convolution layers, and the space convolution layer adopts space-domain-based diagram convolution.
3. The method for monitoring transient stability of a power system according to claim 1, wherein the step of obtaining a steady state result of each generator according to a power angle value of the generator after the occurrence of the fault as a label corresponding to each generator specifically comprises:
acquiring a generator i corresponding to a maximum power angle and a generator j corresponding to a minimum power angle in all generators of the power system obtained at a preset moment after long-time simulation;
respectively counting the number of generators with absolute values of differences of power angles of other generators and generators i and j smaller than 180, and selecting more generators with absolute values of differences of power angles of other generators smaller than 180 from the generators i and j as reference power angle generators;
setting the stable state of the reference power angle generator as not unstable, marking as 0, calculating the absolute value of the difference between the power angles of other generators and the reference power angle generator, setting the generator mark with the absolute value of the difference between the power angles of the reference power angle generator larger than 180 as unstable, marking as 1, and setting the generator mark with the absolute value of the difference between the power angles of the reference power angle generator smaller than 180 as not unstable, marking as 0.
4. The method for monitoring transient stability of a power system according to claim 2, wherein after the fault occurs, collecting characteristic data of the power system in different fault modes, constructing an input state diagram sequence, and inputting the input state diagram sequence into a transient stability prediction model of the power system to obtain a instability probability of each node, and specifically comprises:
inputting the input state diagram sequence into a transient stability prediction model of the power system, and obtaining an output state diagram sequence through a 2-layer space-time convolution block;
for each node, extracting a characteristic vector set corresponding to the node on a state diagram corresponding to each sampling point in the output state diagram sequence, and splicing to obtain a characteristic length vector of each node;
and outputting the characteristic length vector of each node through a single-layer full-connection layer of the output layer and a Sigmoid activation function to obtain the instability probability of each node.
5. A transient stability monitoring device for an electrical power system, comprising:
the system comprises a historical characteristic data acquisition module, a fault detection module and a fault detection module, wherein the historical characteristic data acquisition module is used for acquiring historical characteristic data acquired by a power system, and the system sampling is carried out from the time before the fault occurs to the time after the fault occurs;
the input construction module is used for constructing an input state diagram sequence according to the collected historical characteristic data, wherein the input state diagram sequence is formed according to a state diagram of the power system corresponding to each sampling point on a time sequence from the occurrence of a fault to the occurrence of the fault, and the state diagram is constructed by taking a bus of the power system as a node;
the label acquisition module is used for acquiring a steady state result of each generator according to the power angle value of the generator after the fault occurs and taking the steady state result as a label corresponding to each generator;
the model training module is used for inputting the input state diagram sequence and the steady state result of each generator into model training based on a space-time diagram convolution network to obtain a transient stability prediction model of the power system;
the node instability probability acquisition module is used for acquiring characteristic data of the power system in different fault modes after the fault occurs, constructing an input state diagram sequence, and inputting the input state diagram sequence into the transient stability prediction model of the power system to obtain the instability probability of each node;
the state monitoring module is used for monitoring the stable state of each generator according to the instability probability of each node;
in the construction of the state diagram by taking a bus of a power system as a node, the state diagram is specifically constructed by the following steps:
(1) Define state diagram G t The power system state is used for describing the power system state at the moment t;
(2) Node set V of definition graph, representing busbar set in power system, node V i Is a 6-dimensional vector V bus ,θ bus ,P G ,Q G ,P L ,Q L ]The node voltage amplitude, the node phase angle, the node active power output, the node reactive power output, the node active load and the node reactive load are respectively represented;
(3) Defining a set E of edges, and given an admittance matrix of the power system nodes at the moment t, defining the following time y ij E when not equal to 0 ij E is E; when y is ij When the value of the sum is =0,wherein y is ij Marking the transadmittance of the node i and the node j; e, e ij Is the edge connecting node i and node j;
the input state diagram sequence is specifically:
wherein G is prefault A state diagram representing the power system before a fault occurs,a state diagram sequence representing the power system in the fault occurrence process, including a state diagram corresponding to each sampling point in the time sequence of fault occurrence,the state diagram sequence of the power system after the fault is generated comprises a state diagram corresponding to each sampling point in the time sequence after the fault is generated.
6. The power system transient stability monitoring device of claim 5, wherein said model training module, said space-time diagram convolutional network-based model, comprises:
the space-time diagram convolution network consists of 2 space-time convolution blocks and 1 output layer, wherein the space-time convolution blocks comprise two layers of time convolution layers and one layer of space convolution layer, the space convolution layer is positioned between the two layers of time convolution layers, and the space convolution layer adopts space-domain-based diagram convolution.
7. The power system transient stability monitoring device of claim 6, wherein said node instability probability acquisition module is specifically configured to:
inputting the input state diagram sequence into a transient stability prediction model of the power system, and processing the input state diagram sequence through a 2-layer space-time convolution block to obtain an output state diagram sequence;
for each node, extracting a characteristic vector set corresponding to the node on a state diagram corresponding to each sampling point in the output state diagram sequence, and splicing to obtain a characteristic length vector of each node;
and outputting the characteristic length vector of each node through a single-layer full-connection layer of an output layer and a Sigmoid activation function to obtain the instability probability of each node.
8. A terminal device comprising a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, the processor implementing the power system transient stability monitoring method according to any one of claims 1 to 4 when the computer program is executed.
9. A computer readable storage medium, characterized in that the computer readable storage medium comprises a stored computer program, wherein the computer program, when run, controls a device in which the computer readable storage medium is located to perform the method for monitoring the transient stability of an electrical power system according to any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011241068.0A CN112446171B (en) | 2020-11-09 | 2020-11-09 | Power system transient stability monitoring method and device, terminal equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011241068.0A CN112446171B (en) | 2020-11-09 | 2020-11-09 | Power system transient stability monitoring method and device, terminal equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112446171A CN112446171A (en) | 2021-03-05 |
| CN112446171B true CN112446171B (en) | 2023-11-07 |
Family
ID=74736558
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011241068.0A Active CN112446171B (en) | 2020-11-09 | 2020-11-09 | Power system transient stability monitoring method and device, terminal equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112446171B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109033702A (en) * | 2018-08-23 | 2018-12-18 | 国网内蒙古东部电力有限公司电力科学研究院 | A kind of Transient Voltage Stability in Electric Power System appraisal procedure based on convolutional neural networks CNN |
| CN111461392A (en) * | 2020-01-23 | 2020-07-28 | 华中科技大学 | Power failure prediction method and system based on graph neural network |
| CN111756034A (en) * | 2020-06-12 | 2020-10-09 | 清华大学 | Transient voltage stability evaluation method for power system based on graph space-time network |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20200167666A1 (en) * | 2018-11-28 | 2020-05-28 | Citrix Systems, Inc. | Predictive model based on digital footprints of web applications |
-
2020
- 2020-11-09 CN CN202011241068.0A patent/CN112446171B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109033702A (en) * | 2018-08-23 | 2018-12-18 | 国网内蒙古东部电力有限公司电力科学研究院 | A kind of Transient Voltage Stability in Electric Power System appraisal procedure based on convolutional neural networks CNN |
| CN111461392A (en) * | 2020-01-23 | 2020-07-28 | 华中科技大学 | Power failure prediction method and system based on graph neural network |
| CN111756034A (en) * | 2020-06-12 | 2020-10-09 | 清华大学 | Transient voltage stability evaluation method for power system based on graph space-time network |
Non-Patent Citations (1)
| Title |
|---|
| 基于WAMS受扰电压轨迹簇特征的电力系统暂态稳定性预测;冀鲁豫 等;高电压技术;第41卷(第3期);第807-814页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112446171A (en) | 2021-03-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wilson et al. | Deep learning-aided cyber-attack detection in power transmission systems | |
| US11176446B2 (en) | Compositional prototypes for scalable neurosynaptic networks | |
| CN108121945A (en) | A kind of multi-target detection tracking, electronic equipment and storage medium | |
| Wang et al. | Online analysis of voltage security in a microgrid using convolutional neural networks | |
| CN115238980A (en) | Micro-grid transient power angle trajectory prediction method based on attention mechanism and space-time diagram convolutional neural network | |
| CN114021433A (en) | Construction method and application of dominant instability mode recognition model of power system | |
| CN114332104A (en) | Power grid power transmission scene RGB point cloud semantic segmentation multi-stage model joint optimization method | |
| CN109685805A (en) | A kind of image partition method and device | |
| CN110348109A (en) | The method and terminal device of three-dimensional artificial training data processing | |
| Hassani et al. | Regression models with graph-regularization learning algorithms for accurate fault location in smart grids | |
| CN112446171B (en) | Power system transient stability monitoring method and device, terminal equipment and storage medium | |
| Matejek et al. | Synapse-aware skeleton generation for neural circuits | |
| CN115795360A (en) | Cable fault detection method based on artificial neural network | |
| CN116166642A (en) | Spatio-temporal data filling method, system, equipment and medium based on guide information | |
| CN111832358A (en) | Point cloud semantic analysis method and device | |
| CN107102543A (en) | The forming method and device of a kind of energy router anti-interference controller | |
| CN113489011A (en) | Power system transient stability evaluation method and device, electronic equipment and storage medium | |
| Hu et al. | Graph transformer based dynamic multiple graph convolution networks for traffic flow forecasting | |
| Devie et al. | Transient stability prediction in multimachine system using data mining techniques | |
| Dabou et al. | Supervised learning of overcomplete dictionaries for rapid response-based dynamic stability prediction | |
| Amorim et al. | Faults location in transmission lines through neural networks | |
| CN115577573B (en) | Method, device, equipment and storage medium for predicting output current of synchronous generator | |
| CN111291464A (en) | Dynamic equivalence method and device for power system | |
| CN115661821B (en) | Loop detection method, loop detection device, electronic apparatus, storage medium, and program product | |
| CN117274637A (en) | Pedestrian track prediction method, device, computer equipment and storage medium |
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 |