CN117669103A - Novel intelligent calculation method for broadband oscillation damping coefficient of power system - Google Patents
Novel intelligent calculation method for broadband oscillation damping coefficient of power system Download PDFInfo
- Publication number
- CN117669103A CN117669103A CN202311631530.1A CN202311631530A CN117669103A CN 117669103 A CN117669103 A CN 117669103A CN 202311631530 A CN202311631530 A CN 202311631530A CN 117669103 A CN117669103 A CN 117669103A
- Authority
- CN
- China
- Prior art keywords
- node
- damping coefficient
- pagerank
- power system
- oscillation
- 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.)
- Pending
Links
- 230000010355 oscillation Effects 0.000 title claims abstract description 60
- 238000013016 damping Methods 0.000 title claims abstract description 59
- 238000004364 calculation method Methods 0.000 title claims abstract description 30
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 22
- 230000002708 enhancing effect Effects 0.000 claims abstract description 9
- 239000011159 matrix material Substances 0.000 claims description 27
- 239000003795 chemical substances by application Substances 0.000 claims description 15
- 230000006870 function Effects 0.000 claims description 12
- 238000000034 method Methods 0.000 claims description 12
- 238000005070 sampling Methods 0.000 claims description 9
- 238000013135 deep learning Methods 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 7
- 238000004088 simulation Methods 0.000 claims description 6
- 230000005540 biological transmission Effects 0.000 claims description 5
- 239000012634 fragment Substances 0.000 claims description 3
- 230000009191 jumping Effects 0.000 claims description 3
- 230000001629 suppression Effects 0.000 claims description 3
- 238000012502 risk assessment Methods 0.000 abstract description 7
- 238000005259 measurement Methods 0.000 abstract description 6
- 238000013528 artificial neural network Methods 0.000 abstract description 3
- 230000008859 change Effects 0.000 description 3
- 230000004044 response Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Abstract
The invention discloses a novel intelligent calculation method for broadband oscillation damping coefficient of a power system, which comprises the following steps: s1, generating a sample; s2, sorting importance of the measuring points based on a PageRank algorithm; s3, calculating damping coefficients of the measuring points; s4, enhancing the damping coefficient of the system. The invention has high precision: the system oscillation risk assessment precision can be effectively improved to 98.5%; the measurement requirement is low: the system oscillation risk assessment can be realized by collecting measurement signals of about 40% of the lines of the system; the calculation speed is high: compared with the traditional oscillation risk assessment based on the CPU, the calculation speed is improved by about 16 times by calling the GPU through the embedded calculation of the neural network.
Description
Technical Field
The invention relates to the field of electricity, in particular to a novel intelligent calculation method for broadband oscillation damping coefficient of a power system.
Background
The broadband oscillation risk assessment of the power system is a process of identifying potential problems through data collection, system modeling and oscillation characteristic analysis, comprehensively considering possible risks, and then taking corresponding risk management measures to ensure stable operation of the power system. The oscillation damping coefficient is an important parameter reflecting the oscillation change trend, and can be used for carrying out oscillation damping identification based on measured voltage/current phasor data from the data response perspective so as to evaluate the broadband oscillation risk.
The method for evaluating the broadband oscillation risk in the existing system is mainly based on an oscillation damping coefficient and an impedance network-based evaluation method, but the method is poor in accuracy of calculating the oscillation damping coefficient, high in measurement requirement and low in calculation speed.
Disclosure of Invention
In order to solve the problems of poor precision of calculating damping coefficients, high measurement requirement and low calculation speed in the prior art, the application provides a novel intelligent calculation method for broadband oscillation damping coefficients of a power system, and the problems are solved.
The application discloses a novel intelligent calculation method for broadband oscillation damping coefficient of a power system, which comprises the following steps:
s1, generating a sample, acquiring current data of a novel power system measuring point under renewable energy access through time domain simulation, dividing time sequence data into a plurality of data fragments according to oscillation discrimination time requirements, and further acquiring a data sample;
s2, sorting importance of measuring points based on PageRank algorithm, and sorting importance of topological nodes in the power system;
s3, measuring point damping coefficient calculation, and oscillation damping identification based on measured current data, so that the oscillation risk is estimated, and important information is provided for the subsequent deployment of the broadband oscillation suppression device;
s4, enhancing the system damping coefficient, and providing a broadband oscillation damping coefficient calculation method integrating PageRank and deep learning, selecting a key line, and combining the weight of the deep learning line to realize enhancing the system damping coefficient.
Preferably, the step S2 specifically includes the following steps:
s21, topology modeling: adopting PageRank algorithm, regarding the power system as an undirected graph, and representing the undirected graph by a binary matrix A;
s22, establishing a random Agent model: introducing a random Agent, traversing the power system by the Agent according to a topological connection relation, considering that energy is consumed at an original node when the Agent stays at the node in the traversing process, using a random browsing matrix P to represent a probability model, and using an element P of P ij Representing a probability of jumping from node i to node j;
s23, initializing PageRank parameters: initializing PageRank score v, setting hysteresis factor d, iteration times upper limit Imax, convergence threshold epsilon and counter n;
s24, increasing the value of a counter n, and starting to update the PageRank score when n does not reach the maximum iteration number;
s25, initializing PageRank score: the PageRank scores are regarded as node distribution over the whole power system, each node representing its relative importance by a PageRank score, these scores being regarded as a column vector v, v i A PageRank score representing node i;
s26, pageRank score updating: the importance of the nodes is converted into other node transmission weights through a PageRank algorithm, the PageRank score of one node depends on the number and the score of the nodes connected with the PageRank score, and the node transmission weight with the high PageRank score is more;
s27, detecting whether the PageRank value is converged, and if the convergence condition is met, ending the algorithm updating.
Preferably, the node A in the binary matrix A in the S21 i The following constraints are satisfied:
wherein q is the maximum column number of the matrix, A ij Representing the line coefficients from node i to node j, A i For the ith row of the binary matrix, representing node A i Connection with the rest nodes, A ii The binary matrix diagonal, i.e., the node itself, is represented, and the coefficient is set to 0 regardless of the node's self-loop.
Preferably, the method of S26 is as follows:
assuming that the probability that the Agent does not stay in the original node is d epsilon [0,1], generating a random number x when updating PageRank distribution each time, and the updating process is as follows:
wherein: v k ,v k+1 The node PageRank score vector is obtained after the k and k+1 iterations, I is a unit matrix, P is a random browsing matrix, and A is a topological binary matrix.
Preferably, the calculation formula of the damping coefficient in S3 is as follows:
wherein x (t) represents an oscillation signal and S is a modal amplitudeF is the oscillation frequency and θ is the modal phase; t is t 0 Is the initial time of signal acquisition, t k Is the current sampling time; m is the length of the data window; ts is the sampling period; σ represents the damping coefficient and k is the number of sampling points.
Preferably, the step of enhancing the damping coefficient of the system in S4 includes:
an enhancement model of the self-encoder structure is designed, and the encoder and decoder formulas of the self-encoder are as follows:
x′＝g φ (ω)；
in the formula, the output omega of the encoder is the line weight, the sum of the satisfaction is 1, l is the number of key lines, sigma is the damping coefficient of the key lines, v is the importance of the key lines, and x= [ sigma, v]Inputting a model; x' represents the decoder output; f (f) θ (x) Representing an encoder function; g φ (ω) represents a decoder function;
after obtaining the line weight omega, the damping coefficient of the obtained system is as follows
The enhancement model loss function includes a self-encoder loss and a damping coefficient loss, expressed as follows:
wherein: lambda is a constant value, and,as a loss function.
The invention has the beneficial effects that:
1. the precision is high: the intelligent calculation method of the broadband oscillation damping coefficient fused with Pagerank and deep learning can effectively improve the system oscillation risk assessment accuracy to 98.5%.
2. The measurement requirement is low: the risk assessment of system oscillation can be realized by collecting only about 40% of the measurement signals of the lines of the system.
3. The calculation speed is high: compared with the traditional oscillation risk assessment based on the CPU, the calculation speed is improved by about 16 times by calling the GPU through the embedded calculation of the neural network.
Drawings
FIG. 1 is a flow chart of a novel intelligent calculation method for broadband oscillation damping coefficient of a power system according to an embodiment of the invention;
FIG. 2 is a block diagram of a PageRank algorithm in accordance with an embodiment of the present invention;
FIG. 3 is a block diagram of a system damping coefficient intelligent enhancement algorithm according to an embodiment of the present invention;
FIG. 4 is a training curve of an intelligent enhancement model according to an embodiment of the present invention.
Detailed Description
For the purposes of making the objects, technical solutions and advantages of the present application more apparent, the present application will be described in further detail below with reference to the accompanying drawings and examples.
The application discloses a novel intelligent calculation method for broadband oscillation damping coefficient of a power system, as shown in fig. 1, comprising the following steps:
s1, generating a sample, acquiring current data of a novel power system measuring point under renewable energy access through time domain simulation, dividing time sequence data into a plurality of data fragments according to oscillation discrimination time requirements, and further acquiring a data sample;
s2, sorting importance of measuring points based on PageRank algorithm, and sorting importance of topological nodes in the power system;
the PageRank algorithm is an algorithm for evaluating the importance of nodes in an undirected graph structure to measure the relative importance of network systems such as the Internet. The principle of the PageRank algorithm is based on graph theory and probability theory, which aims to quantify the weights of nodes in a network, which can be represented as various entities. The power system is a typical undirected graph network, and in this embodiment, topology nodes in the power system are used as network nodes to order importance.
As shown in fig. 2, S2 includes the steps of:
s21, topology modeling: using PageRank algorithm, the power system is regarded as an undirected graph, and the undirected graph is represented by a binary matrix A, wherein nodes A in the binary matrix A are represented by the undirected graph i The following constraints are satisfied:
wherein q is the maximum column number of the matrix, A ij Representing the line coefficients from node i to node j, A i For the ith row of the binary matrix, representing node A i Connection with the rest nodes, A ii The binary matrix diagonal, i.e., the node itself, is represented, and the coefficient is set to 0 regardless of the node's self-loop.
S22, establishing a random Agent model: introducing a random Agent, traversing the power system by the Agent according to a topological connection relation, considering that energy is consumed at an original node when the Agent stays at the node in the traversing process, using a random browsing matrix P to represent a probability model, and using an element P of P ij Representing a probability of jumping from node i to node j;
s23, initializing PageRank parameters: initializing PageRank score v, setting hysteresis factor d, iteration times upper limit Imax, convergence threshold epsilon and counter n;
s24, increasing the value of a counter n, and starting to update the PageRank score when n does not reach the maximum iteration number;
s25, initializing PageRank score: the PageRank scores are regarded as node distribution over the whole power system, each node representing its relative importance by a PageRank score, these scores being regarded as a column vector v, v i A PageRank score representing node i;
s26, pageRank score updating: the importance of the nodes is converted into other node transmission weights through the PageRank algorithm, the PageRank score of one node depends on the number and the score of the nodes connected with the PageRank score, and the node with the high PageRank score transmits more weight.
Assuming that the probability that the Agent does not stay in the original node is d epsilon [0,1], generating a random number x when updating PageRank distribution each time, and the updating process is as follows:
wherein: v k ,v k+1 The node PageRank score vector is obtained after the k and k+1 iterations, I is a unit matrix, P is a random browsing matrix, and A is a topological binary matrix.
S27, detecting whether the PageRank value is converged, and if the convergence condition is met, ending the algorithm updating.
S3, calculating a damping coefficient of the measuring point, wherein the damping coefficient is an important parameter reflecting the oscillation change trend, and can be evaluated from the angle of data response. The damping coefficient can be used for oscillation discrimination, the oscillation discrimination based on data response does not depend on phasor data, and the oscillation damping identification can be performed based on measured current data, so that the oscillation risk is evaluated, and important information is provided for the subsequent deployment of the broadband oscillation suppression device.
In the method based on the oscillation damping coefficient, the oscillation signal is assumed to change in an exponential form, and the system damping coefficient can be estimated at different moments through a time sequence formed by oscillation assignment.
The damping coefficient is calculated by the following formula:
wherein x (t) represents an oscillation signal, S is a modal amplitude, f is an oscillation frequency, and θ is a modal phase; t is t 0 Is the initial time of signal acquisition, t k Is the current sampling time; m is the length of the data window; ts is the sampling period; σ represents the damping coefficient and k is the number of sampling points.
S4, enhancing the system damping coefficient, as shown in FIG. 3, providing a broadband oscillation damping coefficient calculation method integrating PageRank and deep learning, selecting a key line, and combining the weight of the deep learning line to realize enhancing the system damping coefficient.
In one embodiment, an enhancement model of the self-encoder structure is designed, the encoder and decoder formulas of the self-encoder are as follows:
x′＝g φ (ω)；
in the formula, the output omega of the encoder is the line weight, the sum of the satisfaction is 1, l is the number of key lines, sigma is the damping coefficient of the key lines, v is the importance of the key lines, and x= [ sigma, v]Inputting a model; x' represents the decoder output; f (f) θ (x) Representing an encoder function; g φ (ω) represents a decoder function;
after obtaining the line weight omega, the damping coefficient of the obtained system is as follows
The enhancement model loss function includes a self-encoder loss and a damping coefficient loss, expressed as follows:
wherein: lambda is a constant value, and,is a loss function such as root mean square value (Mean Squared Error, MSE).
In one specific embodiment, λ=0.1 is set,MSE is selected, and whether oscillation discrimination performance is achieved by using PageRank algorithm is tested.
The model training loss and accuracy curve is shown in fig. 4, an IEEE-118 full electromagnetic example simulation system is adopted in the test, subsynchronous oscillation frequency ranges 21.32Hz and 45.67Hz are generated, simulation step sizes 1e-8 are recorded together, simulation duration is 20s, the neural network construction is constructed based on Python3.8 and Tensorflow2.8 versions, the front and back discrimination effects of the PageRank algorithm are shown in a table 1, and the oscillation discrimination precision is further improved after the PageRank algorithm is used.
Table 1: front-back discrimination effect comparison using PageRank algorithm
The foregoing has shown and described the basic principles, features, and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, and that the above embodiments and descriptions are merely illustrative of the principles of the present invention, and various changes and modifications may be made without departing from the spirit and scope of the invention, which is defined in the appended claims. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (6)
1. A novel intelligent calculation method for broadband oscillation damping coefficient of an electric power system is characterized by comprising the following steps:
s1, generating a sample, acquiring current data of a novel power system measuring point under renewable energy access through time domain simulation, dividing time sequence data into a plurality of data fragments according to oscillation discrimination time requirements, and further acquiring a data sample;
s2, sorting importance of measuring points based on PageRank algorithm, and sorting importance of topological nodes in the power system;
s3, measuring point damping coefficient calculation, and oscillation damping identification based on measured current data, so that the oscillation risk is estimated, and important information is provided for the subsequent deployment of the broadband oscillation suppression device;
s4, enhancing the system damping coefficient, and providing a broadband oscillation damping coefficient calculation method integrating PageRank and deep learning, selecting a key line, and combining the weight of the deep learning line to realize enhancing the system damping coefficient.
2. The intelligent calculation method of broadband oscillation damping coefficient of novel power system according to claim 1, wherein the step S2 specifically comprises the following steps:
s21, topology modeling: adopting PageRank algorithm, regarding the power system as an undirected graph, and representing the undirected graph by a binary matrix A;
s22, establishing a random Agent model: introducing a random Agent, traversing the power system by the Agent according to a topological connection relation, considering that energy is consumed at an original node when the Agent stays at the node in the traversing process, using a random browsing matrix P to represent a probability model, and using an element P of P ij Representing a probability of jumping from node i to node j;
s23, initializing PageRank parameters: initializing PageRank score v, setting hysteresis factor d, iteration times upper limit Imax, convergence threshold epsilon and counter n;
s24, increasing the value of a counter n, and starting to update the PageRank score when n does not reach the maximum iteration number;
s25, initializing PageRank score: the PageRank scores are regarded as node distribution over the whole power system, each node representing its relative importance by a PageRank score, these scores being regarded as a column vector v, v i A PageRank score representing node i;
s26, pageRank score updating: the importance of the nodes is converted into other node transmission weights through a PageRank algorithm, the PageRank score of one node depends on the number and the score of the nodes connected with the PageRank score, and the node transmission weight with the high PageRank score is more;
s27, detecting whether the PageRank value is converged, and if the convergence condition is met, ending the algorithm updating.
3. A novel power system broadband oscillation damping system according to claim 2The intelligent calculation method is characterized in that the node A in the binary matrix A in the S21 i The following constraints are satisfied:
wherein q is the maximum column number of the matrix, A ij Representing the line coefficients from node i to node j, A i For the ith row of the binary matrix, representing the node S i Connection with the rest nodes, A ii The binary matrix diagonal, i.e., the node itself, is represented, and the coefficient is set to 0 regardless of the node's self-loop.
4. The intelligent calculation method of broadband oscillation damping coefficient of novel power system according to claim 3, wherein the method of S26 is as follows:
assuming that the probability that the Agent does not stay in the original node is d epsilon [0,1], generating a random number x when updating PageRank distribution each time, and the updating process is as follows:
wherein: v k ,v k+1 The node PageRank score vector is obtained after the k and k+1 iterations, I is a unit matrix, P is a random browsing matrix, and A is a topological binary matrix.
5. The intelligent calculation method of the broadband oscillation damping coefficient of the novel power system according to claim 4, wherein the calculation formula of the damping coefficient in S3 is as follows:
wherein x (t) represents an oscillation signal, S is a modal amplitude, f is an oscillation frequency, and θ is a modal phase; t is t 0 Is the initial time of signal acquisition, t k Is the current sampling time; m is the length of the data window; ts is the sampling period; σ represents the damping coefficient and k is the number of sampling points.
6. The intelligent calculation method of broadband oscillation damping coefficient of a novel power system according to claim 5, wherein the step of enhancing the system damping coefficient in S4 is as follows:
an enhancement model of the self-encoder structure is designed, and the encoder and decoder formulas of the self-encoder are as follows:
ω＝f θ (x),x＝[σ,υ],
x′＝g φ (ω)；
in the formula, the output omega of the encoder is the line weight, the sum of the satisfaction is 1, l is the number of key lines, sigma is the damping coefficient of the key lines, v is the importance of the key lines, and x= [ sigma, v]Inputting a model; x' represents the decoder output; f (f) θ (x) Representing an encoder function; g φ (ω) represents a decoder function;
after obtaining the line weight omega, the damping coefficient of the obtained system is as follows
The enhancement model loss function includes a self-encoder loss and a damping coefficient loss, expressed as follows:
wherein: lambda is a constant value, and,as a loss function.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311631530.1A CN117669103A (en) | 2023-11-30 | 2023-11-30 | Novel intelligent calculation method for broadband oscillation damping coefficient of power system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311631530.1A CN117669103A (en) | 2023-11-30 | 2023-11-30 | Novel intelligent calculation method for broadband oscillation damping coefficient of power system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117669103A true CN117669103A (en) | 2024-03-08 |
Family
ID=90082023
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311631530.1A Pending CN117669103A (en) | 2023-11-30 | 2023-11-30 | Novel intelligent calculation method for broadband oscillation damping coefficient of power system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117669103A (en) |
-
2023
- 2023-11-30 CN CN202311631530.1A patent/CN117669103A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106055918B (en) | Method for identifying and correcting load data of power system | |
| CN108155648A (en) | Method for estimating state based on the infinite Extended Kalman filter of adaptive H | |
| TWI738597B (en) | Analog battery construction method and analog battery construction device | |
| Mirshekali et al. | Linear and nonlinear fault location in smart distribution network under line parameter uncertainty | |
| CN110514967B (en) | Fault positioning method for predicting arrival time of traveling wave head based on NARNN model | |
| CN109359665B (en) | Household appliance load identification method and device based on support vector machine | |
| CN109598052B (en) | Intelligent ammeter life cycle prediction method and device based on correlation coefficient analysis | |
| CN113484813B (en) | Intelligent ammeter fault rate prediction method and system under multi-environment stress | |
| CN113514739B (en) | IWOA-BP algorithm-based oil paper insulation aging evaluation method | |
| CN113190969A (en) | Lithium battery model parameter identification method based on information evaluation mechanism | |
| CN114358092B (en) | Method and system for online diagnosis of internal insulation performance of capacitor voltage transformer | |
| CN115270239A (en) | Bridge reliability prediction method based on dynamic characteristics and intelligent algorithm response surface method | |
| CN117233687B (en) | CVT initial error assessment method, medium and terminal based on historical data | |
| CN112287520B (en) | Oil paper insulation expansion Debye model parameter identification method based on two-step optimization method | |
| CN117669103A (en) | Novel intelligent calculation method for broadband oscillation damping coefficient of power system | |
| CN109038550B (en) | Power system self-repeatability index calculation method based on voltage static stability | |
| CN117310361A (en) | Power distribution network fault patrol positioning method based on intelligent perception and equipment image | |
| CN115688015A (en) | Voltage sag type identification method and device based on non-invasive load monitoring | |
| CN116449219A (en) | Lithium battery model parameter identification method for improving VFFRLS | |
| Ma et al. | Online assessment of failure probability for smart meters based on SARIMA-LTFRLS | |
| CN115856641A (en) | Method and device for predicting remaining charging time of battery and electronic equipment | |
| CN107069710B (en) | Power system state estimation method considering new energy space-time correlation | |
| CN115183884A (en) | Infrared temperature measurement compensation method and device of electric heating cooperative system | |
| Wang et al. | Harmonic state estimation based on graph signal processing | |
| CN117439068B (en) | Voltage sag estimation method, medium and system in large-scale power grid |
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 |