CN113488997B - Micro-grid critical energy evaluation method based on disturbance response multi-scale features - Google Patents

Micro-grid critical energy evaluation method based on disturbance response multi-scale features Download PDF

Info

Publication number
CN113488997B
CN113488997B CN202110764255.5A CN202110764255A CN113488997B CN 113488997 B CN113488997 B CN 113488997B CN 202110764255 A CN202110764255 A CN 202110764255A CN 113488997 B CN113488997 B CN 113488997B
Authority
CN
China
Prior art keywords
disturbance
micro
time
grid
critical energy
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
Application number
CN202110764255.5A
Other languages
Chinese (zh)
Other versions
CN113488997A (en
Inventor
欧阳静
秦龙
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang University of Technology ZJUT
Original Assignee
Zhejiang University of Technology ZJUT
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Zhejiang University of Technology ZJUT filed Critical Zhejiang University of Technology ZJUT
Priority to CN202110764255.5A priority Critical patent/CN113488997B/en
Publication of CN113488997A publication Critical patent/CN113488997A/en
Application granted granted Critical
Publication of CN113488997B publication Critical patent/CN113488997B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • H02J3/001—Methods to deal with contingencies, e.g. abnormalities, faults or failures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • 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
    • H02J3/24—Arrangements for preventing or reducing oscillations of power in 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
    • H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Molecular Biology (AREA)
  • Artificial Intelligence (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Power Engineering (AREA)
  • Supply And Distribution Of Alternating Current (AREA)

Abstract

The invention discloses a disturbance response multi-scale feature-based micro-grid critical energy evaluation method, which specifically comprises the following steps: 1) Acquiring and preprocessing meteorological and electric energy quality real-time data of a microgrid system; 2) Judging a disturbance source and disturbance time of the microgrid; 3) Extracting and fusing the disturbance response multi-scale features of the micro-grid; 4) Evaluating the critical energy of the microgrid system in real time; the method comprehensively considers the virtual power angle transient characteristics of the voltage source inversion type distributed power supply based on the virtual synchronous generator technology, adopts the convolutional neural network method to extract the disturbance response characteristics of the microgrid, adopts the series characteristic fusion method to fuse the deep characteristics and the shallow characteristics, is favorable for fully utilizing the detailed information and the response information of the disturbance response characteristics in the critical energy evaluation process, and improves the critical energy evaluation precision.

Description

Micro-grid critical energy evaluation method based on disturbance response multi-scale features
Technical Field
The invention relates to the field of disturbance stability analysis of a micro-grid source load containing distributed energy, in particular to a micro-grid critical energy evaluation method based on disturbance response multi-scale characteristics.
Background
The micro-grid system is used as an effective organization form of distributed energy sources capable of running in parallel and off-grid, and various distributed energy sources accessed by power electronic type interfaces, such as wind power, photovoltaic and the like, are included. Due to weak inertia caused by a power electronic type interface and distributed power supply output fluctuation caused by meteorological factors, a micro-grid system is very easily influenced by source load disturbance, and the stability problem is caused.
With the proposal of a double-carbon target, the permeability of distributed energy represented by wind power generation and photovoltaic power generation in a microgrid also tends to be higher, and the influence of the disturbance problem of the 'source load' on the stability of the microgrid is more prominent.
The dynamic stability of the power system is judged from the energy perspective, the calculation speed is high compared with that of a traditional time domain simulation method, and the system stability margin can be obtained. The existing system critical energy evaluation method aims at a large power grid system, wherein an inertia link is provided by a synchronous generator, a micro-grid is dominated by a voltage source inversion type distributed power supply based on a virtual synchronous generator technology, and the output characteristic of the micro-grid is different from that of the synchronous generator. And the methods such as a dominant unstable equilibrium point method, a dominant unstable equilibrium point method based on a stable domain boundary, a potential energy interface method and the like all need to adopt a plurality of off-line simulations, and the requirements of real-time on-line stability analysis are difficult to meet. In conclusion, the determination of the critical energy of the microgrid system is a key and difficult point of the dynamic stability analysis of the microgrid under the source load disturbance.
The wide-area measurement and high-speed communication technology of the power quality parameters improves the observability of the micro-grid system and provides a prerequisite for real-time online evaluation of dynamic stability. Therefore, how to evaluate the critical energy of the microgrid system under the source charge disturbance in real time becomes an urgent problem to be solved.
Disclosure of Invention
The invention aims to overcome the defects that most of the conventional system critical energy evaluation methods use a synchronous generator as an inertia link, or need multiple off-line simulation and are difficult to evaluate in real time, provides a disturbance response multi-scale feature-based micro-grid critical energy evaluation method, and comprehensively considers the virtual power angle transient characteristics of a voltage source inversion type distributed power supply based on the virtual synchronous generator technology and the system disturbance response multi-scale feature fusion.
The invention achieves the above purpose through the following technical scheme: a micro-grid critical energy assessment method based on disturbance response multi-scale features specifically comprises the following steps:
1) Real-time data acquisition and pretreatment of meteorological and electric energy quality of microgrid system
Acquiring real-time data of weather and power quality and filtering the data by using a micro-grid weather data acquisition system and a power quality data wide-area measurement system; on the basis, the meteorological data and the power quality data of each electrical node are normalized, and finally, a time sequence is formed according to the sampling time;
2) Micro-grid disturbance source and disturbance moment judgment
When meteorological data such as irradiance, wind speed and temperature and humidity are subjected to sudden change, and meanwhile, one of electric energy quality data such as voltage, current, active data and reactive data of distributed power supply nodes is subjected to sudden change, namely disturbance generated by output fluctuation of the distributed power supply; when the meteorological data change is stable, the power quality data of the electrical nodes in the microgrid system are suddenly changed, and the load increase and load decrease or the disturbance of electrical faults may occur; recording the time when the power quality data mutates and an electrical node, wherein the time is the disturbance time, and the electrical node is a disturbance source;
3) Micro-grid disturbance response multi-scale feature extraction and fusion
Calculating a virtual power angle of a voltage source inversion type distributed power supply based on a virtual synchronous generator technology in the microgrid system; extracting disturbance response characteristics of the microgrid by adopting a Convolutional Neural Network (CNN) -based method, designing a convolution operator kernel for source load disturbance fault risk identification, and performing multi-scale fusion on deep response information and shallow detail characteristics in the characteristic extraction process to obtain a final disturbance response characteristic sequence;
4) Micro-grid system critical energy real-time evaluation
Training the disturbance response characteristic sequence obtained in the step 3) by using a gate control cycle unit network GRU algorithm, and establishing a micro-grid system critical energy real-time evaluation model; and sequentially inputting the disturbance response characteristic sequence according to a time sequence until the output of the model is finally converged, outputting a real-time evaluation value of the critical energy of the microgrid system, and finally performing inverse normalization on the real-time evaluation value to obtain an actual evaluation result.
Further, the step of extracting and fusing the disturbance response multi-scale features of the microgrid in the step 3) comprises the following steps:
1.1 The virtual power angle calculation method of the voltage source inversion type distributed power supply based on the virtual synchronous generator technology comprises the following steps:
when the output of the voltage source inversion type distributed power supplyVirtual power angle when current is not saturated
Figure GDA0003777856980000031
When the output current is saturated, the virtual power angle
Figure GDA0003777856980000032
Wherein, delta is the virtual power angle of the voltage source inversion type distributed power supply, P 0 For the inverter output power, P nm Maximum output power in unsaturated state, P sm The maximum output power in the saturated state; and arranging the virtual power angles obtained by calculation into a time sequence according to the serial numbers of the electrical nodes where the distributed power supplies are located and the sampling time.
1.2 Dividing the time sequence of the power quality node where the disturbance source in the step 1) is located from the disturbance moment to the next according to the disturbance moment and the disturbance source obtained in the step 2); the input data of the micro-grid disturbance response characteristic extraction model are the time sequence of the power quality data of the disturbance source after disturbance occurs and the virtual power angle time sequence obtained in the step 1.1).
1.3 Inputting the electric energy quality data time sequence of the disturbance source obtained in the step 1.2) and the virtual power angle time sequence obtained in the step 1.1) into a micro-grid disturbance response characteristic extraction model, and performing characteristic extraction on the electric energy quality data time sequence and the virtual power angle time sequence of the disturbance source after disturbance occurs by using a convolutional neural network method; with the increase of the number of the convolution layers, the extracted disturbance information feature levels are different, shallow layer features comprise voltage/current out-of-limit, amplitude fluctuation, frequency fluctuation, phase angle deviation and other detailed information, deep layer features are small in resolution and have rich response information, and the extraction of disturbance response features can be assisted.
1.4 The convolution operator kernel of the 'source load' disturbance fault risk identification adopted by the convolution neural network method in the step 1.3) is as follows:
Figure GDA0003777856980000033
where K is the number of feature map channels, f is the size of the convolution kernel, l is the number of convolution layers, and s 0 Is the convolution step size, b is the amount of deviation,
Figure GDA0003777856980000034
in order to convolve the input matrix with each other,
Figure GDA0003777856980000035
is a weight matrix; l.. I e {0,1 l+1 }，
Figure GDA0003777856980000036
L l+1 The dimension of the (l + 1) th layer convolution output is defined, and p is the number of filling layers; the parameters of the weight matrix are trained in the characteristic learning process, and automatic updating is realized.
1.5 Carrying out multi-scale fusion on the deep response information and the shallow detail characteristics obtained in the step 1.3) in the identification process, specifically fusing the deep characteristics containing the response information and the shallow characteristics containing the detail information, and carrying out oversampling on the deep characteristics by 2 times in order to ensure the dimension consistency of the two characteristics; and then, fusing the deep-layer features and the shallow-layer features by adopting a series feature fusion concat method to obtain a micro-grid disturbance response feature sequence.
Further, the real-time evaluation of the critical energy of the microgrid system in the step 4) comprises the following specific steps:
2.1 Training and evaluating the disturbance response characteristic sequence of the micro-grid obtained in the step 3) by using a gated cyclic neural network GRU algorithm, establishing a micro-grid system critical energy real-time evaluation model, and relieving the problem of gradient disappearance or gradient explosion existing in long-time sequence learning of a convolutional neural network algorithm; inputting the model, namely obtaining a disturbance response characteristic sequence of the microgrid in the step 3), and outputting a real-time evaluation value of the critical energy of the microgrid system; and performing inverse normalization on the real-time evaluation value to obtain an actual critical energy evaluation result of the microgrid system.
The invention has the beneficial effects that:
(1) The virtual power angle of the voltage source inversion type distributed power supply based on the virtual synchronous generator technology is adopted to replace the angular speed of the traditional synchronous generator to be used as the data input of the critical energy evaluation of the microgrid system, so that the evaluation result is more consistent with the actual running condition of the microgrid;
(2) The method adopts a convolutional neural network method to extract the characteristics of the electric energy quality data time sequence and the virtual power angle time sequence of the disturbance source, characterizes the disturbance response characteristics from the angles of shallow characteristics and deep characteristics, enriches the disturbance characteristic information from the aspect of data and improves the identification precision of the disturbance characteristics;
(3) According to the method, the deep-layer characteristics and the shallow-layer characteristics are fused by adopting a series characteristic fusion concat method in the disturbance response characteristic identification process, so that the detailed information and response information of the disturbance response characteristics are fully utilized in the critical energy evaluation process, and the critical energy evaluation precision is improved;
(4) According to the invention, the gated cyclic neural network GRU algorithm is adopted to train and evaluate the disturbance response characteristic sequence of the microgrid, so that the problem that the gradient disappears or the gradient explodes when the data length of the convolutional neural network algorithm is larger can be effectively relieved, and the further improvement of the critical energy evaluation result is realized.
Drawings
FIG. 1 is a flow chart of a micro-grid critical energy assessment method based on disturbance response multi-scale features.
Detailed Description
The invention will be described in detail with reference to the accompanying drawings: the present embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation manner and a specific operation process are given, but the protection scope of the present invention is not limited to the following embodiments.
Referring to fig. 1, a micro-grid critical energy assessment method based on disturbance response multi-scale features includes the following steps:
1) Meteorological and electric energy quality real-time data acquisition and pretreatment of micro-grid system
Acquiring real-time data of weather and power quality and filtering the data by using a micro-grid weather data acquisition system and a power quality data wide-area measurement system; on the basis, the meteorological data and the power quality data of each electrical node are normalized, and finally, a time sequence is formed according to the sampling time;
2) Micro-grid disturbance source and disturbance moment judgment
When meteorological data such as irradiance, wind speed and temperature and humidity are subjected to sudden change, and meanwhile, one of electric energy quality data such as voltage, current, active data and reactive data of distributed power supply nodes is subjected to sudden change, namely disturbance generated by output fluctuation of the distributed power supply; when the meteorological data change is stable, the power quality data of the electrical nodes in the microgrid system are suddenly changed, and loading and unloading or electrical faults may occur at the moment; recording the time when the power quality data mutates and an electrical node, wherein the time is the disturbance time, and the electrical node is a disturbance source;
3) Micro-grid disturbance response multi-scale feature extraction and fusion
Calculating a virtual power angle of a voltage source inversion type distributed power supply based on a virtual synchronous generator technology in the micro-grid system; extracting disturbance response characteristics of the microgrid by adopting a Convolutional Neural Network (CNN) -based method, designing a convolution operator kernel for source load disturbance fault risk identification, and performing multi-scale fusion on deep response information and shallow detail characteristics in the characteristic extraction process to obtain a final disturbance response characteristic sequence;
4) Micro-grid system critical energy real-time assessment
Training the disturbance response characteristic sequence obtained in the step 3) by using a gate control cycle unit network GRU algorithm, and establishing a micro-grid system critical energy real-time evaluation model; and sequentially inputting the disturbance response characteristic sequence according to a time sequence until the output of the model is finally converged, outputting a real-time evaluation value of the critical energy of the microgrid system, and finally performing inverse normalization on the real-time evaluation value to obtain an actual evaluation result.
The wind-solar-storage combined micro-grid system comprises distributed power supplies including wind power and photovoltaic power, wherein the operation mode of one part of the distributed power supplies is a voltage source based on a virtual synchronous generator technology; the load comprises a dragging load represented by an induction motor, an RLC type daily load and the like, and a storage battery energy storage system, and under the condition that the output fluctuation of a distributed power supply or the switching of a high-power load is used as a disturbance source, the micro-grid critical energy evaluation modeling is taken as an example.
The invention relates to a microgrid critical energy evaluation method considering multi-scale characteristics of a voltage source virtual power angle and disturbance response based on a virtual synchronous generator technology, which is shown in figure 1 and comprises the following steps:
step 1, acquiring and preprocessing meteorological and electric energy quality real-time data of a microgrid system
Acquiring real-time weather and electric energy quality data of each node by using a micro-grid weather and electric energy quality data wide-area measurement system, then carrying out preprocessing such as data filtering, storage and the like, carrying out normalization on the basis, and arranging into a time sequence according to sampling moments;
step 2, judging disturbance source and disturbance moment of the microgrid
Firstly, judging whether the output of the distributed power supply fluctuates according to whether the meteorological data and the power quality data of the distributed power supply nodes are mutated simultaneously; according to the fact that the meteorological data change is stable, the power quality data of the electrical nodes in the microgrid system are suddenly changed, and the fact that disturbance such as loading and unloading or electrical faults possibly occurs at the moment is judged; recording the disturbance time t and disturbance source d 1 ,d 2 ...,d n (where n is the number of disturbance sources);
step 3, extracting and fusing multi-scale features of disturbance response of the microgrid
Firstly, calculating a virtual power angle of a voltage source inverter type distributed power supply based on a virtual synchronous generator technology; when the output current of the voltage source inversion type distributed power supply is not saturated, the virtual power angle is as follows:
Figure GDA0003777856980000061
when the output current of the voltage source inversion type distributed power supply is not saturated, the virtual power angle is as follows:
Figure GDA0003777856980000071
wherein delta ' is a virtual power angle P ' of the voltage source inversion type distributed power supply ' 0 Is inverter output power, P' nm Is the maximum output power, P ', in a non-saturated state' sm The maximum output power in the saturated state; and arranging the virtual power angles obtained by calculation into a time sequence according to the serial number of the electrical node where the distributed power supply is located and the sampling time.
Secondly, the disturbance source d is determined by the disturbance time t obtained in the step 2 1 ,d 2 ...,d n Dividing the time sequence of the power quality node where the disturbance source is located in the step 1) from the disturbance moment t to serve as a dividing basis; and the input data of the micro-grid disturbance response characteristic extraction model are the time sequence of the power quality data of the disturbance source after disturbance occurs and the virtual power angle time sequence obtained in the last step.
Thirdly, designing the kernel of the convolution operator for identifying the source load disturbance fault risk as follows:
Figure GDA0003777856980000072
where K is the number of feature map channels, f is the size of the convolution kernel, l is the number of convolution layers, and s 0 Is the convolution step size, b is the amount of deviation,
Figure GDA0003777856980000073
in order to convolve the input matrix with each other,
Figure GDA0003777856980000074
is a weight matrix; l.. I e {0,1 l+1 }，
Figure GDA0003777856980000075
L l+1 The dimension of the (l + 1) th layer convolution output is shown, and p is the number of filling layers; the parameters of the weight matrix are trained in the characteristic learning process, and automatic updating is realized; performing feature extraction on the power quality data time sequence and the virtual power angle time sequence of the disturbance source after the disturbance occurs by using a convolutional neural network method to obtain shallow features of disturbance informationAnd deep features;
and finally, performing 2-time oversampling on the deep features, and fusing the deep features and the shallow features by adopting a series feature fusion concat method to obtain a micro-grid disturbance response feature sequence.
Step 4, evaluating critical energy of the microgrid system in real time
Training and evaluating the micro-grid disturbance response characteristic sequence obtained in the step 3 by using a gated recurrent neural network GRU algorithm, and establishing a micro-grid system critical energy real-time evaluation model; inputting the model, namely obtaining a disturbance response characteristic sequence of the microgrid in the step 3, and outputting a real-time evaluation value of the critical energy of the microgrid system; and performing inverse normalization on the real-time evaluation value to obtain an actual critical energy evaluation result of the microgrid system.
The above embodiments are only preferred embodiments of the present invention, and are not intended to limit the technical solutions of the present invention, so long as the technical solutions can be realized on the basis of the above embodiments without creative efforts, which should be considered to fall within the protection scope of the patent of the present invention.

Claims (2)

1. A micro-grid critical energy assessment method based on disturbance response multi-scale features is characterized by comprising the following steps: the method specifically comprises the following steps:
1) Meteorological and electric energy quality real-time data acquisition and pretreatment of micro-grid system
Acquiring real-time data of weather and power quality and filtering the data by using a micro-grid weather data acquisition system and a power quality data wide-area measurement system; on the basis, the meteorological data and the power quality data of each electrical node are normalized, and finally, a time sequence is formed according to the sampling time;
2) Micro-grid disturbance source and disturbance moment judgment
When the meteorological data are mutated, the power quality data of the distributed power supply nodes are also mutated, and the power quality data are the disturbance generated by the output fluctuation of the distributed power supply; when the meteorological data change is stable, the power quality data of the electrical nodes in the microgrid system are suddenly changed, and the load increase and the load decrease or the disturbance of electrical faults may occur; recording the time when the power quality data mutates and an electrical node, wherein the time is the disturbance time, and the electrical node is a disturbance source;
3) Micro-grid disturbance response multi-scale feature extraction and fusion
Calculating a virtual power angle of a voltage source inversion type distributed power supply based on a virtual synchronous generator technology in the microgrid system; extracting disturbance response characteristics of the microgrid by adopting a Convolutional Neural Network (CNN) -based method, designing a convolution operator kernel for source charge disturbance fault risk identification, and performing multi-scale fusion on deep response information and shallow detail characteristics in the characteristic extraction process to obtain a final disturbance response characteristic sequence, wherein the method comprises the following specific steps of:
1.1 The virtual power angle calculation method of the voltage source inversion type distributed power supply based on the virtual synchronous generator technology comprises the following steps: when the output current of the voltage source inversion type distributed power supply is not saturated, the virtual power angle
Figure FDA0003799263330000011
When the output current is saturated, the virtual power angle
Figure FDA0003799263330000012
Wherein, delta is the virtual power angle of the voltage source inversion type distributed power supply, P 0 For the inverter output power, P nm Maximum output power in the unsaturated state, P sm The maximum output power in the saturated state; arranging the virtual power angles obtained by calculation into a time sequence according to the serial numbers of the electrical nodes where the distributed power supplies are located and the sampling time;
1.2 Dividing the time sequence of the power quality node where the disturbance source in the step 1) is located from the disturbance moment to the next according to the disturbance moment and the disturbance source obtained in the step 2); the input data of the micro-grid disturbance response characteristic extraction model are a power quality data time sequence of a disturbance source after disturbance occurs and a virtual power angle time sequence obtained in the step 1.1);
1.3 Inputting the electric energy quality data time sequence of the disturbance source obtained in the step 1.2) and the virtual power angle time sequence obtained in the step 1.1) into a micro-grid disturbance response characteristic extraction model, and performing characteristic extraction on the electric energy quality data time sequence and the virtual power angle time sequence of the disturbance source after disturbance occurs by using a convolutional neural network method; along with the increase of the number of the convolution layers, the extracted disturbance information features are different in level, shallow features comprise voltage/current out-of-limit, amplitude fluctuation, frequency fluctuation, phase angle shift and other detailed information, deep features are small in resolution and have rich response information, and the extraction of disturbance response features can be assisted;
1.4 The convolution operator kernel of the 'source load' disturbance fault risk identification adopted by the convolution neural network method in the step 1.3) is as follows:
Figure FDA0003799263330000021
where K is the number of feature map channels, f is the size of the convolution kernel, l is the number of convolution layers, and s 0 Is the convolution step size, b is the amount of deviation,
Figure FDA0003799263330000022
in order to convolve the input matrix with each other,
Figure FDA0003799263330000023
is a weight matrix; l.. I e {0,1 l+1 }，
Figure FDA0003799263330000024
L l+1 The dimension of the (l + 1) th layer convolution output is shown, and p is the number of filling layers; the parameters of the weight matrix are trained in the characteristic learning process, and automatic updating is realized;
1.5 Carrying out multi-scale fusion on the deep response information and the shallow detail characteristics obtained in the step 1.3) in the identification process, specifically fusing the deep characteristics containing the response information and the shallow characteristics containing the detail information, and carrying out oversampling on the deep characteristics by 2 times in order to ensure the dimension consistency of the two characteristics; then, fusing the deep layer characteristics and the shallow layer characteristics by adopting a series characteristic fusion concat method to obtain a micro-grid disturbance response characteristic sequence;
4) Micro-grid system critical energy real-time assessment
Training the disturbance response characteristic sequence obtained in the step 3) by using a gate control cycle unit network GRU algorithm, and establishing a micro-grid system critical energy real-time evaluation model; and sequentially inputting the disturbance response characteristic sequence according to a time sequence until the output of the model is finally converged, outputting a real-time evaluation value of the critical energy of the microgrid system, and finally performing inverse normalization on the real-time evaluation value to obtain an actual evaluation result.
2. The microgrid critical energy evaluation method based on disturbance response multi-scale features of claim 1, characterized in that: the real-time evaluation of the critical energy of the microgrid system in the step 4) comprises the following specific steps:
2.1 Training and evaluating the disturbance response characteristic sequence of the micro-grid obtained in the step 3) by using a gated recurrent neural network GRU algorithm, establishing a real-time evaluation model of critical energy of the micro-grid system, and relieving the problem of gradient disappearance or gradient explosion existing in long-time sequence learning of a convolutional neural network algorithm; inputting the model, namely obtaining a disturbance response characteristic sequence of the microgrid in the step 3), and outputting a real-time evaluation value of the critical energy of the microgrid system; and performing inverse normalization on the real-time evaluation value to obtain an actual critical energy evaluation result of the micro-grid system.
CN202110764255.5A 2021-07-06 2021-07-06 Micro-grid critical energy evaluation method based on disturbance response multi-scale features Active CN113488997B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110764255.5A CN113488997B (en) 2021-07-06 2021-07-06 Micro-grid critical energy evaluation method based on disturbance response multi-scale features

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110764255.5A CN113488997B (en) 2021-07-06 2021-07-06 Micro-grid critical energy evaluation method based on disturbance response multi-scale features

Publications (2)

Publication Number Publication Date
CN113488997A CN113488997A (en) 2021-10-08
CN113488997B true CN113488997B (en) 2022-10-04

Family

ID=77941406

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110764255.5A Active CN113488997B (en) 2021-07-06 2021-07-06 Micro-grid critical energy evaluation method based on disturbance response multi-scale features

Country Status (1)

Country Link
CN (1) CN113488997B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114422515B (en) * 2021-12-13 2023-09-15 广东电网有限责任公司 Edge computing architecture design method and system suitable for power industry
CN116957361B (en) * 2023-07-27 2024-02-06 中国舰船研究设计中心 Ship task system health state detection method based on virtual-real combination

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105932717A (en) * 2016-06-30 2016-09-07 东南大学 Grid-connected and off-grid smooth handover control method of micro-grids based on disturbance observer
CN111950913A (en) * 2020-08-17 2020-11-17 浙江工业大学 Micro-grid power quality comprehensive evaluation method based on node voltage sensitivity
CN112017070A (en) * 2020-07-17 2020-12-01 中国电力科学研究院有限公司 Method and system for evaluating transient stability of power system based on data enhancement

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160247065A1 (en) * 2006-02-14 2016-08-25 Power Analytics Corporation Systems and Methods for Real-Time Forecasting and Predicting of Electrical Peaks and Managing the Energy, Health, Reliability, and Performance of Electrical Power Systems Based on an Artificial Adaptive Neural Network

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105932717A (en) * 2016-06-30 2016-09-07 东南大学 Grid-connected and off-grid smooth handover control method of micro-grids based on disturbance observer
CN112017070A (en) * 2020-07-17 2020-12-01 中国电力科学研究院有限公司 Method and system for evaluating transient stability of power system based on data enhancement
CN111950913A (en) * 2020-08-17 2020-11-17 浙江工业大学 Micro-grid power quality comprehensive evaluation method based on node voltage sensitivity

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
Storage_Free_Smart_Energy_Management_for_Frequency_Control_in_a_Diesel-PV-Fuel_Cell-Based_Hybrid_AC_Microgrid;P. C. Sekhar;《IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS》;20160831;第1657-1671页 *
基于节点电压灵敏度的微电网电能质量综合动态评价方法;欧阳静;《高技术通讯》;20210215;第199-206页 *
微电网脆弱性预评估方法;周磊等;《电力需求侧管理》;20180120(第01期);第20-24页 *

Also Published As

Publication number Publication date
CN113488997A (en) 2021-10-08

Similar Documents

Publication Publication Date Title
CN111046581B (en) Power transmission line fault type identification method and system
CN113488997B (en) Micro-grid critical energy evaluation method based on disturbance response multi-scale features
CN111914486B (en) Power system transient stability evaluation method based on graph attention network
CN108832619A (en) Transient stability evaluation in power system method based on convolutional neural networks
CN112330165B (en) Power grid transient stability evaluation method and system based on feature separation type neural network
CN111478314B (en) Transient stability evaluation method for power system
CN106980073A (en) A kind of two ends of electric transmission line fault recorder data matching process based on convolutional neural networks
CN114021433B (en) Construction method and application of dominant instability mode identification model of power system
CN111415059B (en) Practical model machine construction and online application method
CN112200694A (en) Dominant instability mode identification model construction and application method based on graph neural network
CN114169249B (en) Artificial intelligent identification method for high-resistance ground fault of power distribution network
CN111553112A (en) Power system fault identification method and device based on deep belief network
LU500551B1 (en) Virtual load dominant parameter identification method based on incremental learning
CN105932673B (en) A kind of external network equivalent method of the large-scale interconnected power system containing extensive new energy
CN114330571B (en) Power system dominant instability mode identification method and system based on transfer learning
CN114169395A (en) Construction method and application of dominant instability mode recognition model of power system
CN115469184A (en) New energy transmission line fault identification method based on convolutional network
CN112564107A (en) Transient stability assessment method for power system
CN116131313A (en) Explanatory analysis method for association relation between characteristic quantity and transient power angle stability
CN117498379A (en) New energy station broadband frequency coupling impedance on-line identification modeling method and system
CN116432023A (en) Novel power system fault classification method based on sample transfer learning
CN113092934B (en) Single-phase earth fault judgment method and system based on clustering and LSTM
Li et al. Visualization of CNN transient voltage classification based on feature recognition and enhancement
CN116204771A (en) Power system transient stability key feature selection method, device and product
Bai et al. An evolution strategy of GAN for the generation of high impedance fault samples based on Reptile algorithm

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