CN113394774A - Static voltage stability monitoring method based on deep neural network and impedance model margin - Google Patents
Static voltage stability monitoring method based on deep neural network and impedance model margin Download PDFInfo
- Publication number
- CN113394774A CN113394774A CN202110698905.0A CN202110698905A CN113394774A CN 113394774 A CN113394774 A CN 113394774A CN 202110698905 A CN202110698905 A CN 202110698905A CN 113394774 A CN113394774 A CN 113394774A
- Authority
- CN
- China
- Prior art keywords
- impedance
- neural network
- deep neural
- margin
- load
- 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
Images
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/10—Power transmission or distribution systems management focussing at grid-level, e.g. load flow analysis, node profile computation, meshed network optimisation, active network management or spinning reserve management
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a static voltage stability monitoring method based on a deep neural network and an impedance model margin. The method comprises the following steps: establishing a deep neural network model; setting a determined load level parameter, and calculating an impedance mode margin algorithm under the parameter to obtain impedance mode margin values of all system load nodes; sequencing the impedance mode margin values of all the load nodes, and finding out relatively weak nodes for observation; randomly setting the overall load level parameters of the system, calculating the impedance model margin value of the weak node, bringing multiple groups of input parameters and impedance model margin values into a deep neural network to iteratively learn the characteristics of the weak node, and stopping iteration until an expected deep neural network model is obtained; the method comprises the steps of collecting input data in real time, predicting an impedance modulus tolerance value of a current weak node through a deep neural network, and judging the current system voltage stable state according to the impedance modulus tolerance value of the current weak node.
Description
Technical Field
The invention relates to the technical field of power grid safety, in particular to a static voltage stability monitoring method based on a deep neural network and an impedance model margin.
Background
With the continuous increase of load demands, the interconnection scale of the power grid is continuously increased, and the uncertain distributed energy sources in the power system are greatly accessed and the intelligent degree is continuously improved, so that the voltage stability problem caused by reactive power which cannot be transmitted in a large range is more prominent. Therefore, the efficient and accurate online voltage stability evaluation system is particularly important for preventing a large-scale power failure accident. One of the key technologies is to provide a voltage stabilization online monitoring method capable of quickly, accurately and adaptively adapting to the system state change.
A load margin index evaluation system based on continuous power flow, which is widely used in the current online voltage stability evaluation system, needs a long time of offline analysis and calculation, and has the defects that the mode of the average increase of the load of the whole network does not conform to the actual disturbance mode, weak link information cannot be given, the change of the running state of the system can not be quickly adapted, and the like, so that the requirement of online monitoring cannot be met. This is a disadvantage of the prior art.
In view of the above, the present application provides a static voltage stability monitoring method based on a deep neural network and an impedance model margin; it is necessary to solve the above-mentioned defects and shortcomings existing in the prior art.
Disclosure of Invention
The invention aims to disclose a static voltage stability monitoring method based on a deep neural network and an impedance model margin, so as to solve the technical problem.
In order to achieve the above object, a static voltage stability monitoring method based on a deep neural network and an impedance model margin is disclosed, comprising:
step S1, establishing a deep neural network;
step S2, setting a determined load level parameter, and calculating an impedance mode margin algorithm under the parameter to obtain impedance mode margin values of all system load nodes;
s3, sequencing the impedance module margin values of all the load nodes, and finding out a relatively weak node for observation;
step S4, randomly setting the overall load level parameters of the system, calculating the impedance model margin value of the weak node, bringing multiple groups of input parameters and impedance model margin values into the deep neural network to iteratively learn the characteristics of the weak node, and stopping iteration until an expected deep neural network model is obtained;
and step S5, acquiring input data in real time, predicting the impedance modulus tolerance value of the current weak node through the deep neural network, and judging the current system voltage stable state according to the impedance modulus tolerance value of the current weak node.
2. The method according to claim 1, wherein the step S2 specifically includes:
and S21, the load level parameter k is a load proportion coefficient. When k equals 1, the system is in the ground state; when the k value exceeds the limit load proportionality coefficient kmaxWhen the system transmission load exceeds the system transmission load limit, the system load flow calculation is not converged and cannot be solved;
s22, calculating the impedance mode margin:
defining current data acquired by a phasor measurement unit as a current mode I, and defining voltage data acquired by the phasor measurement unit as a voltage mode U;
definition of ZiLDThe equivalent impedance mode is U/I and is a static equivalent impedance mode of an electric load end;
definition of ZiTHEVdU/dI, which is a comprehensive dynamic equivalent impedance mode;
defining the impedance mode margin value as:
3. the method according to claim 1, wherein the step S3 specifically includes:
s31, determining the number n of the relatively weak nodes according to actual requirements or relevant regulations;
s32, the sorting of the impedance mode margin values is carried out under the ground state of the system, namely under the condition that the load proportionality coefficient is 1.
4. The method according to claim 1, wherein the step S4 specifically includes:
s41, the input of the deep neural network model is defined as voltage data collected by the phasor measurement unit, active load data collected by the phasor measurement unit, voltage phase angle data collected by the phasor measurement unit and reactive load data; the output of the deep neural network model is defined as the impedance modulus margin value of the weak node being monitored.
5. The method according to claim 1, wherein the step S5 specifically includes:
s51, collecting the sensor data of each weak node in real time as the input data of the deep neural network, and predicting the real-time impedance model tolerance value according to the deep neural network model trained in S4; when the margin value of the impedance mode is less than the threshold etaminAnd when the voltage is 0.1, judging that the voltage of the power grid system is unstable.
As a general technical concept, the present invention further provides a static voltage stability monitoring method based on a deep neural network and an impedance mode margin, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of any one of the above methods when executing the computer program.
The invention has the following beneficial effects:
according to the method, the weak nodes of the power grid system are considered, the voltage stability margin state of the whole power grid is indicated by the local impedance mode margin information of the weak nodes, and the monitoring difficulty of the voltage stability margin of the system is reduced; in addition, the invention simplifies the solving process of the impedance model margin algorithm by a deep neural network method, improves the calculation efficiency, is beneficial to online application and has good application potential.
The present invention will be described in further detail below with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate embodiments of the invention and, together with the description, serve to explain the invention and not to limit the invention. In the drawings:
fig. 1 is a schematic flow chart of a static voltage stability monitoring method based on a deep neural network and an impedance model margin according to an embodiment of the present invention.
Fig. 2 is a power grid structure topology and weak node positioning diagram according to the embodiment of the present invention.
Detailed Description
The embodiments of the invention will be described in detail below with reference to the drawings, but the invention can be implemented in many different ways as defined and covered by the claims.
The embodiment discloses a static voltage stability monitoring method based on a deep neural network and an impedance model margin, as shown in fig. 1, including:
and step S1, establishing a deep neural network.
And step S2, setting a determined load level parameter, and calculating an impedance mode margin algorithm under the parameter to obtain impedance mode margin values of all system load nodes.
In this step, preferably, specifically, the method includes:
and S21, the load level parameter k is a load proportion coefficient. When k equals 1, the system is in the ground state; when the k value exceeds the limit load proportionality coefficient kmaxWhen the system transmission load exceeds the system transmission load limit, the system load flow calculation is not converged and cannot be solved;
s22, calculating the impedance mode margin:
defining current data acquired by a phasor measurement unit as a current mode I, and defining voltage data acquired by the phasor measurement unit as a voltage mode U;
definition of ZiLDThe equivalent impedance mode is U/I and is a static equivalent impedance mode of an electric load end;
definition of ZiTHEVdU/dI, which is a comprehensive dynamic equivalent impedance mode;
defining the impedance mode margin value as:
and step S3, sequencing the impedance module margin values of all the load nodes, and finding out the relatively weak nodes for observation.
In this step, preferably, specifically, the method includes:
s31, determining the number n of the relatively weak nodes according to actual requirements or relevant regulations;
s32, the sorting of the impedance mode margin values is carried out under the ground state of the system, namely under the condition that the load proportionality coefficient is 1.
And step S4, randomly setting the overall load level parameters of the system, calculating the impedance model margin value of the weak node, bringing multiple groups of input parameters and impedance model margin values into the deep neural network to iteratively learn the characteristics of the weak node, and stopping iteration until the expected deep neural network model is obtained.
In this step, preferably, specifically, the method includes:
s41, the input of the deep neural network model is defined as voltage data collected by the phasor measurement unit, active load data collected by the phasor measurement unit, voltage phase angle data collected by the phasor measurement unit and reactive load data; the output of the deep neural network model is defined as the impedance modulus margin value of the weak node being monitored.
And step S5, acquiring input data in real time, predicting the impedance modulus tolerance value of the current weak node through the deep neural network, and judging the current system voltage stable state according to the impedance modulus tolerance value of the current weak node.
In this step, preferably, specifically, the method includes:
s51, collecting the sensor data of each weak node in real time as the input data of the deep neural network, and predicting the real-time impedance model tolerance value according to the deep neural network model trained in S4; when the margin value of the impedance mode is less than the threshold etaminAnd when the voltage is 0.1, judging that the voltage of the power grid system is unstable.
Further, regarding the above steps, the present embodiment takes an IEEE 118 node system as an example to further explain and verify the method of the present invention.
First, a deep neural network model is built. Then, the calculation of the impedance mode tolerance value in the ground state, i.e., when the load level factor k is 1, is performed for all nodes of the IEEE 118 node system shown in fig. 2. And determining six nodes with relatively smaller impedance modulus margin values as system weak nodes, namely a node No. 29, a node No. 28, a node No. 41, a node No. 115, a node No. 114 and a node No. 53. After weak nodes are determined, a load level coefficient k value is set randomly, voltage data, active load data, voltage phase angle data and reactive load data of the six weak nodes under a large batch of random load levels are obtained and used as input of deep neural network model training, and corresponding impedance model margin values are calculated and used as output of the deep neural network model training. And stopping iteration after the precision of the neural network model reaches a preset value through repeated iterative training of the deep neural network model. And finally, collecting sensor data of each observation point in real time to verify the method, wherein the trained neural network predicts the accuracy of an unknown sample, as shown in the table I:
Therefore, the method can utilize the information of a plurality of weak nodes, the efficiency of voltage stability monitoring is guaranteed through a deep neural network method, and the stability level of the power grid can be monitored quickly and accurately through the method of fusing the deep neural network and the impedance model margin algorithm.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (6)
1. A static voltage stability monitoring method based on a deep neural network and an impedance model margin is characterized by comprising the following steps:
step S1, establishing a deep neural network;
step S2, setting a determined load level parameter, and calculating an impedance mode margin algorithm under the parameter to obtain impedance mode margin values of all system load nodes;
s3, sequencing the impedance module margin values of all the load nodes, and finding out a relatively weak node for observation;
step S4, randomly setting the overall load level parameters of the system, calculating the impedance model margin value of the weak node, bringing multiple groups of input parameters and impedance model margin values into the deep neural network to iteratively learn the characteristics of the weak node, and stopping iteration until an expected deep neural network model is obtained;
and step S5, acquiring input data in real time, predicting the impedance modulus tolerance value of the current weak node through the deep neural network, and judging the current system voltage stable state according to the impedance modulus tolerance value of the current weak node.
2. The method according to claim 1, wherein the step S2 specifically includes:
and S21, the load level parameter k is a load proportion coefficient. When k equals 1, the system is in the ground state; when the k value exceeds the limit load proportionality coefficient kmaxWhen the system transmission load exceeds the system transmission load limit, the system load flow calculation is not converged and cannot be solved;
s22, calculating the impedance mode margin:
defining current data acquired by a phasor measurement unit as a current mode I, and defining voltage data acquired by the phasor measurement unit as a voltage mode U;
definition of ZiLDThe equivalent impedance mode is U/I and is a static equivalent impedance mode of an electric load end;
definition of ZiTHEVdU/dI, which is a comprehensive dynamic equivalent impedance mode;
defining the impedance mode margin value as:
3. the method according to claim 1, wherein the step S3 specifically includes:
s31, determining the number n of the relatively weak nodes according to actual requirements or relevant regulations;
s32, the sorting of the impedance mode margin values is carried out under the ground state of the system, namely under the condition that the load proportionality coefficient is 1.
4. The method according to claim 1, wherein the step S4 specifically includes:
s41, the input of the deep neural network model is defined as voltage data collected by the phasor measurement unit, active load data collected by the phasor measurement unit, voltage phase angle data collected by the phasor measurement unit and reactive load data; the output of the deep neural network model is defined as the impedance modulus margin value of the weak node being monitored.
5. The method according to claim 1, wherein the step S5 specifically includes:
s51, collecting the sensor data of each weak node in real time as the input data of the deep neural network, and predicting the real-time impedance model tolerance value according to the deep neural network model trained in S4; when the margin value of the impedance mode is less than the threshold etaminAnd when the voltage is 0.1, judging that the voltage of the power grid system is unstable.
6. A static voltage stability monitoring method based on a deep neural network and an impedance model margin, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method according to any one of claims 1 to 5 when executing the computer program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110698905.0A CN113394774A (en) | 2021-06-23 | 2021-06-23 | Static voltage stability monitoring method based on deep neural network and impedance model margin |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110698905.0A CN113394774A (en) | 2021-06-23 | 2021-06-23 | Static voltage stability monitoring method based on deep neural network and impedance model margin |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113394774A true CN113394774A (en) | 2021-09-14 |
Family
ID=77623545
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110698905.0A Pending CN113394774A (en) | 2021-06-23 | 2021-06-23 | Static voltage stability monitoring method based on deep neural network and impedance model margin |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113394774A (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105811404A (en) * | 2016-03-22 | 2016-07-27 | 山东大学 | Stable situation monitoring method for quiescent voltage of distribution network with synergic transmission and distribution |
| CN111628501A (en) * | 2020-06-18 | 2020-09-04 | 国网山东省电力公司济南供电公司 | AC/DC large power grid transient voltage stability assessment method and system |
-
2021
- 2021-06-23 CN CN202110698905.0A patent/CN113394774A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105811404A (en) * | 2016-03-22 | 2016-07-27 | 山东大学 | Stable situation monitoring method for quiescent voltage of distribution network with synergic transmission and distribution |
| CN111628501A (en) * | 2020-06-18 | 2020-09-04 | 国网山东省电力公司济南供电公司 | AC/DC large power grid transient voltage stability assessment method and system |
Non-Patent Citations (1)
| Title |
|---|
| 颜廷鑫等: "基于神经网络的法向阻抗模裕度快速计算方法", 《电网技术》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106532778B (en) | Method for calculating maximum access capacity of distributed photovoltaic grid connection | |
| CN109063276B (en) | Wind power plant dynamic equivalent modeling method suitable for long-time domain wind speed fluctuation | |
| CN110601250B (en) | Method for evaluating stability of static voltage of access point based on new energy power station | |
| CN110503153B (en) | Photovoltaic system fault diagnosis method based on differential evolution algorithm and support vector machine | |
| CN112734128A (en) | 7-day power load peak value prediction method based on optimized RBF | |
| CN112287605B (en) | Power flow checking method based on graph convolution network acceleration | |
| CN112883632A (en) | Lithium battery equivalent circuit model parameter identification method based on improved ant colony algorithm | |
| US20220243347A1 (en) | Determination method and determination apparatus for conversion efficiency of hydrogen production by wind-solar hybrid electrolysis of water | |
| CN117370744A (en) | Dynamic cleaning method and system for abnormal power consumption data of power consumer | |
| CN110571788A (en) | static voltage stability domain boundary coefficient calculation method based on dynamic equivalent circuit | |
| CN109447512B (en) | Large power grid reliability assessment method based on uniform design | |
| CN110489852A (en) | Improve the method and device of the wind power system quality of data | |
| CN111127242A (en) | Power system reliability dynamic real-time assessment method based on small sample data | |
| CN116896093A (en) | Online analysis and optimization method for grid-connected oscillation stability of wind farm | |
| CN113394774A (en) | Static voltage stability monitoring method based on deep neural network and impedance model margin | |
| CN113128130B (en) | Real-time monitoring method and device for judging stability of direct-current power distribution system | |
| CN107069710B (en) | Power system state estimation method considering new energy space-time correlation | |
| CN112581311B (en) | Method and system for predicting long-term output fluctuation characteristics of aggregated multiple wind power plants | |
| CN110571791B (en) | Optimal configuration method for power transmission network planning under new energy access | |
| CN113627655A (en) | Power distribution network pre-disaster fault scene simulation prediction method and device | |
| CN110207871B (en) | Method, device, storage medium and system for stress prediction of wind turbine generator | |
| CN109376494A (en) | A kind of Probabilistic Load Flow based on electric system determines method and system | |
| CN117390418B (en) | Transient stability evaluation method, system and equipment for wind power grid-connected system | |
| CN114418194B (en) | Tower damage prediction method and device based on data driving and model driving | |
| CN111798049B (en) | Voltage stability evaluation method based on integrated learning and multi-target planning |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20210914 |