CN107993012B - Time-adaptive online transient stability evaluation method for power system - Google Patents
Time-adaptive online transient stability evaluation method for power system Download PDFInfo
- Publication number
- CN107993012B CN107993012B CN201711260938.7A CN201711260938A CN107993012B CN 107993012 B CN107993012 B CN 107993012B CN 201711260938 A CN201711260938 A CN 201711260938A CN 107993012 B CN107993012 B CN 107993012B
- Authority
- CN
- China
- Prior art keywords
- time
- evaluation
- rbm
- transient stability
- hidden layer
- 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.)
- Expired - Fee Related
Links
- 238000011156 evaluation Methods 0.000 title claims abstract description 65
- 230000001052 transient effect Effects 0.000 title claims abstract description 51
- 238000012549 training Methods 0.000 claims abstract description 22
- 238000005259 measurement Methods 0.000 claims abstract description 10
- 238000013210 evaluation model Methods 0.000 claims abstract description 9
- 238000013507 mapping Methods 0.000 claims abstract description 8
- 238000000034 method Methods 0.000 claims description 33
- 230000008569 process Effects 0.000 claims description 21
- 230000000007 visual effect Effects 0.000 claims description 10
- 210000003061 neural cell Anatomy 0.000 claims description 7
- 230000004913 activation Effects 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 6
- 210000005036 nerve Anatomy 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 3
- 230000001537 neural effect Effects 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 238000004088 simulation Methods 0.000 claims description 3
- 238000013097 stability assessment Methods 0.000 claims description 3
- 230000003044 adaptive effect Effects 0.000 claims description 2
- 230000000750 progressive effect Effects 0.000 abstract description 3
- 239000011159 matrix material Substances 0.000 description 5
- 238000010801 machine learning Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Educational Administration (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- Theoretical Computer Science (AREA)
- Development Economics (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention relates to a time-adaptive online transient stability evaluation method for a power system, which selects measurable dynamic data of a power grid wide area measurement system with a time window with the length of T after fault clearing as input characteristics, utilizes DBN to extract the characteristics to obtain high-order characteristics, further utilizes GRU to perform offline training to obtain a mapping relation between the high-order characteristics and the transient stability of the power system, and inputs the input characteristics according to a time progressive sequence during online application until an accurate evaluation result is obtained or the time window length is up to stop. Therefore, the transient stability evaluation model of the power system provided by the invention has the characteristics of direct orientation to power grid measurable data, strong anti-noise interference capability, high evaluation precision and strong time self-adaption.
Description
Technical Field
The invention belongs to the field of power system automation, and relates to a time-adaptive online transient stability evaluation method for a power system.
Background
Transient stability is the ability of a power system to remain stable in the presence of large disturbances. Transient instability is a main factor of the events in many major power failure accidents occurring at home and abroad in the past, so that the maintenance of the transient stability of the power system still has important significance in the planning and operation of the power system. In order to meet the increasing power demand, loads are getting closer and closer to the power transfer capability, in which case the consequences of transient instability are extremely severe, and therefore online transient stability assessment of power systems is a necessary and promising approach.
The transient stability evaluation method of the power system based on machine learning is the method which has the most potential to be applied to online. At present, a method for transient stability evaluation based on machine learning mainly considers that after a fault occurs, a mapping relation between an input characteristic and a stable state is obtained by taking a dynamic response of a system as a characteristic, and emergency control is required to be adopted for adjustment once instability occurs. This method can achieve a high evaluation accuracy, but it generally requires a period of time after the fault is cleared before transient stability evaluation can be performed. The longer this time period, the higher the accuracy of the evaluation, but the time reserved for the dispatcher to make adjustments may be insufficient. The transient stability of the power system can be accurately evaluated in the early stage of fault clearing, so that the research on the online transient stability evaluation method of the time-adaptive power system is the key for solving the problem.
Disclosure of Invention
In order to solve the above problems, the present invention provides a time-adaptive online transient stability evaluation method for an electric power system. The method has the advantages of being oriented to bottom layer measurement data, strong in anti-noise interference capability, high in evaluation precision and adaptive to evaluation time, and is suitable for on-line transient stability evaluation of the power system.
The invention provides a time-adaptive online transient stability evaluation method for a power system, which is characterized by comprising the following steps of: forming high-level features X on measurable data of a power grid bottom layer by using a Deep Belief Network (DBN), training data pairs GRU (gate recovery Unit) of a time window T after fault clearing off in an off-line mode to obtain a mapping relation between the high-level features and transient stability, and sequentially carrying out on-line application on the high-level features X at each moment after fault clearing t And performing transient stability evaluation, and stopping evaluation when the evaluation confidence coefficient is met or the time window length is reached.
The technical scheme adopted by the invention is as follows:
a time-adaptive online transient stability evaluation method for a power system is characterized by comprising the following steps:
step 1: generating power grid dynamic data under a plurality of fault conditions by using a time domain simulation method, and selecting measurable data of a power grid wide area measurement system in a time window T after fault removal as an input feature set { x } 1 ,x 2 ,…,x t …x T The measurable data of the power grid wide area measurement system refers to the voltage of each bus of the power grid and the phase angle x thereof t ＝[v 1 ,v 2 ,…,v n ,θ 1 ,θ 2 ,…,θ n ] T Wherein n is the number of buses;
step 2: feature extraction is performed by using DBN to form a high-order feature set { X } 1 ,X 2 ,…,X t …X T Specifically, the DBN is stacked by a softmax layer and m layers of Restricted Boltzmann Machines (RBMs), the connection weight and offset of a first RBM are unsupervised and trained, so that the RBM can reconstruct input characteristics at the maximum probability, the output of the trained first RBM is used as the input of a next RBM for characteristic reconstruction until the unsupervised training of the mth RBM is completed, finally, the softmax layer is added for supervised training for network parameter fine adjustment, and the output of the mth RBM is high-order characteristics { X } 1 ,X 2 ,…,X t …X T }；
And step 3: performing off-line training on the GRU to obtain a mapping relation between high-order characteristics and transient stability of the power system, wherein during off-line training, the input is { X } 1 ,X 2 ,…,X t …X T Output is { y } 1 ,y 2 ,…,y t ,…,y T A training process consists of a forward propagation process and a parameter updating process, and an iteration number N and an upper error limit error are set;
and 4, step 4: applying a time self-adaptive transient stability evaluation model on line, and sequentially carrying out high-level feature X at each moment after fault clearing t Performing transient stability evaluation, stopping evaluation when the evaluation confidence coefficient is met or the time window length is reached, and applying the time self-adaptive transient stability evaluation model on line to obtain the evaluation result
Wherein alpha is a confidence factor, [0, alpha ] U (1-alpha, 1)]Is a confidence interval; using a chronologically progressive input sequence, in particular, inputting x to the model after the fault has cleared 1 Stopping evaluation if the evaluation result is within the confidence interval, otherwise continuing to input x 2 Combining the last time pair x 1 Evaluating the resulting "memory" h 1 Making an assessment of power system stability; until an evaluation result is obtained or the evaluation time window length T is reached.
In the above time-adaptive online transient stability evaluation method for the power system, in step 2, the unsupervised training process of a single RBM is as follows:
step 2.1, the units of the RBM hidden layer are mutually independent, and when the state v of the visual layer is given, the activation probability of the jth nerve unit of the hidden layer is
Wherein
b j Bias for the jth neural cell of the hidden layer, ω ij Is a RBM networkConnecting weights between the ith visual layer unit and the jth hidden layer unit;
2.2, the units of the RBM hidden layer are mutually independent, and when the state h of the hidden layer is given, the activation probability of the ith nerve unit of the visible layer is
Wherein a is i Biasing for the ith neural cell of the hidden layer;
step 2.3, for a number s of input sets { v 1 ,v 2 ,…,v s By maximizing the log-likelihood function of RBM on the input set
Obtaining a model parameter theta, namely extracting the output of the hidden layer as the characteristics of the visible layer, wherein P (v) k | h, θ) is the probability of a visible layer given the state h and parameter θ of the hidden layer, and is formulated as:
step 2.4, the parameter updating formula is as follows:
further, in the present invention, it is preferable that,
in the formula, v (0) For the original input set, v (l) Is to v (0) Obtained by carrying out Gibbs sampling in one step, rho is momentum term, eta is learning rate, b j Bias for the jth neural unit of the hidden layer, ω ij For the connection weight between the ith visual layer unit and the jth hidden layer unit of the RBM network, a i Biasing for the ith neural unit of the hidden layer.
In the above online transient stability evaluation method for a time-adaptive power system, step 3 specifically includes:
step 3.1: executing a loop i-1 to i-N;
step 3.2: executing a loop T ═ 1 to T ═ T;
step 3.3: carrying out a forward propagation process, W (z) ，U (z) Respectively, the input and the last hidden layer to the refresh gate z (r) ，U (r) Respectively representing the input and the connection matrix of the last hidden layer to the reset gate r, W h ，U h A connection matrix from hidden layer to output layer, W, representing input and last time instants, respectively out Is an output weight matrix;
step 3.4: performing a back propagation process, which is a process for updating the weight parameters, and defining a single sample error as
In the formula d t For the actual output value of the sample, the parameter variation amount per update is:
wherein
X t For input at time t, z t Updating the gate for time t, z t Reset gate for time t, h t Output layer for time t, y t Output value of s for time t t-1 For the memory at time t-1, r t+1 The gate is reset for time t +1,
resetting the gate error output for time t + 1;
step 3.5: when the number of iterations is larger than N or the total error of the sample
And when the temperature is less than error, stopping circulation. As shown in fig. 2.
The invention has the characteristics and beneficial effects that: according to the method, measurable dynamic data of a power grid wide area measurement system of a time window with the length of T after fault clearing is selected as input characteristics, DBN is used for extracting characteristics to obtain high-order characteristics, GRU is further used for conducting off-line training to obtain a mapping relation between the high-order characteristics and the transient stability of a power system, and when the method is applied on line, the input characteristics are input according to a time progressive sequence until an accurate evaluation result is obtained or the time window length is reached and stopped. Therefore, the transient stability evaluation model of the power system provided by the invention has the characteristics of direct orientation to power grid measurable data, strong anti-noise interference capability, high evaluation precision and strong time self-adaption, and particularly has the following advantages: (1) the DBN can abstract and express input data to obtain high-order characteristics, and can directly input measurable data to perform characteristic extraction so as to filter the influence of noise in a power grid, so that the DBN has excellent anti-interference capability; (2) supervised learning exists in the process of DBN feature extraction, so that transient stability oriented features can be formed, the evaluation precision of a GRU formed transient stability evaluation model is further improved, and the characteristic of high evaluation accuracy is achieved; (3) the larger the time window T, the higher the model evaluation accuracy. When the model is applied online, the model can evaluate the feature data of the next moment according to the evaluation result until the stability judgment can be made or the evaluation time is reached. In other words, if the stability determination can be made at an early stage, the dispatcher can be left more responsive, and thus the model has the characteristic of time adaptation.
The time-adaptive online transient stability evaluation method for the power system can be applied to online transient stability evaluation of the power system, can perform transient stability evaluation early after fault clearing, and strives for precious time for implementation measures of dispatchers. With the continuous development of the wide area measurement system of the power system, the measured data is gradually enriched, and the time-adaptive online transient stability evaluation method of the power system can play an increasingly important role in the dynamic security evaluation of the power system.
Drawings
FIG. 1 is a schematic flow diagram of the process of the present invention.
FIG. 2 is an online application of the time-adaptive transient stability assessment model of the power system according to the embodiment of the present invention.
Detailed description of the invention
According to the method, the characteristic extraction is carried out on the measurable data of the power system through the DBN, and then the GRU is used for forming the mapping relation between the high-order characteristic and the transient stability, so that the online transient stability evaluation of the power system with self-adaptive time is realized by utilizing the GRU characteristic in the application stage. The following is described in connection with the accompanying drawings and examples:
the technical scheme adopted by the invention is that the time-adaptive online transient stability evaluation method for the power system is characterized by comprising the following steps of:
step 1: generating power grid dynamic data under a plurality of fault conditions by using a time domain simulation method, and selecting measurable data of a power grid wide area measurement system in a time window T after fault removal as an input feature set { x } 1 ,x 2 ,…,x t …x T }；
Step 2: feature extraction is performed by using DBN to form a high-order feature set { X } 1 ,X 2 ,…,X t …X T }；
And 3, step 3: performing offline training on the GRU to obtain a mapping relation between high-order characteristics and transient stability of the power system;
and 4, step 4: applying a time self-adaptive transient stability evaluation model on line, and sequentially carrying out high-level feature X at each moment after fault clearing t And performing transient stability evaluation, and stopping evaluation when the evaluation confidence coefficient is met or the time window length is reached.
The measurable data of the power grid wide area measurement system in the step 1 refers to the voltage of each bus of the power grid and the phase angle x thereof t ＝[v 1 ,v 2 ,…,v n ,θ 1 ,θ 2 ,…,θ n ] T And n is the number of the buses.
Step 2, the DBN is stacked by a softmax layer and m layers of Restricted Boltzmann Machines (RBMs), the connection weight and the offset of the first RBM are trained without supervision, so that the RBM can reconstruct the input characteristics with the maximum probability, the output of the trained first RBM is used as the input of the next RBM for characteristic reconstruction,until the m-th RBM training finishes the unsupervised training, finally adding a softmax layer for carrying out the supervised training for carrying out the network parameter fine adjustment, and outputting the m-th RBM layer as the high-order characteristic { X } 1 ,X 2 ,…,X t …X T }. The single RBM unsupervised training process is as follows:
(1) the units of the RBM hidden layer are mutually independent, and when the state v of the visual layer is given, the activation probability of the jth nerve unit of the hidden layer is
Wherein
b is hidden layer bias, and omega is RBM network connection weight;
(2) the units of the RBM hidden layer are mutually independent, and when the state h of the hidden layer is given, the activation probability of the ith nerve unit of the hidden layer is
Where a is the hidden layer bias;
(3) for a number s of input sets { v 1 ,v 2 ,…,v s By maximizing the log-likelihood function of RBM on the input set
Obtaining a model parameter theta, namely extracting the output of the hidden layer as the characteristics of the visible layer, wherein P (v) k H, θ) is the probability of a visible layer given the state h and parameter θ of the hidden layer, expressed by the equation:
(4) the parameter update formula is as follows:
further, in the present invention, it is preferable that,
in the formula, v (0) For the original input set, v (l) Is to v (0) Obtained by carrying out Gibbs sampling in one step, rho is momentum term, eta is learning rate, b j Bias for the jth neural cell of the hidden layer, ω ij For the connection weight between the ith visual layer unit and the jth hidden layer unit of the RBM network, a i Biasing for the ith neural cell of the hidden layer;
step 3, off-line training is carried out on GRU, and input is { X 1 ,X 2 ,…,X t …X T Output is { y } 1 ,y 2 ,…,y t ,…,y T And a training process of the method consists of a forward propagation process and a parameter updating process, and an iteration number N and an error upper limit error are set, wherein the specific process is described as follows:
step 3.1: executing a loop i-1 to i-N;
step 3.2: executing a loop T-1 to T-T;
step 3.3: carrying out a forward propagation process, W (z) ，U (z) Respectively, the input and t-1 time hidden layer to the update gate z (r) ，U (r) Respectively representing the connections of the hidden layers to the reset gate r at the input and t-1 times, W h ，U h A connection matrix from hidden layer to output layer, W, representing the input and t-1 times, respectively out Is an output weight matrix.
(1) And updating the door at the moment t: z is a radical of t ＝σ(W (z) X t +U (z) s t-1 ) The refresh gate determines the memory s left at time t-1 t-1 ；
(2) Reset gate at time t: r is t ＝σ(W (r) X t +U (r) s t-1 ) The reset gate determines how to combine the new inputs x t And memory of time t-1 t-1 ；
(3) Output layer at time t:
(4) the output value at the time t is: y is t ＝σ(W out s t )。
Step 3.4: to perform a reverse directionA propagation process, which is a process of updating the weight parameters, defining a single sample error as
In the formula d t For the actual output value of the sample, the parameter variation amount per update is:
wherein
Wherein z is t Updating the gate for time t, z t Reset gate for time t, h t Output layer for time t, y t Output value at time t of s t-1 For the memory at time t-1, r t+1 Reset gate for time t +1, r t+1 Reset gate for time t +1
Step 3.5: when the number of iterations is larger than N or the total error of the sample
Is less than error The cycle is stopped.
The time-adaptive online transient stability evaluation method for the power system according to claim 1, characterized in that: step 4, applying the time self-adaptive transient stability evaluation model on line, wherein the evaluation result is
Wherein alpha is a confidence factor, [0, alpha ] U (1-alpha, 1)]Is the confidence interval. Using a time-progressive input sequence, in particular, inputting x to the model after the fault has cleared 1 Stopping evaluation if the evaluation result is within the confidence interval, otherwise continuing to input x 2 Combining the last time pair x 1 Evaluation of the resulting "memory" h 1 An assessment of power system stability is made. And the like until an evaluation result is obtained or the evaluation time window length T is reached.
Claims (2)
1. A time-adaptive online transient stability evaluation method for a power system is characterized by comprising the following steps:
step 1: generating power grid dynamic data under a plurality of fault conditions by using a time domain simulation method, and selecting measurable data of a power grid wide area measurement system in a time window T after fault removal as an input feature set { x } 1 ,x 2 ,…,x t ，…，x T The measurable data of the power grid wide area measurement system refers to the voltage of each bus of the power grid and the phase angle x thereof t ＝[v 1 ,v 2 ,…,v n ,θ 1 ,θ 2 ,…,θ n ] T Wherein n is the number of buses;
step 2: feature extraction is performed by using DBN to form a high-order feature set { X } 1 ,X 2 ,…,X t ，…，X T Specifically, the DBN is stacked by a softmax layer and m layers of Restricted Boltzmann Machines (RBMs), the connection weight and offset of a first RBM are unsupervised and trained, so that the RBM can reconstruct input characteristics at the maximum probability, the output of the trained first RBM is used as the input of a next RBM for characteristic reconstruction until the unsupervised training of the mth RBM is completed, finally, the softmax layer is added for supervised training for network parameter fine adjustment, and the output of the mth RBM is high-order characteristics { X } 1 ,X 2 ,…,X t ，…，X T }；
And step 3: performing off-line training on the GRU to obtain a mapping relation between high-order characteristics and transient stability of the power system, wherein during off-line training, the input is { X } 1 ,X 2 ,…,X t ，…，X T Output is { y } 1 ,y 2 ,…,y t ，…，y T A training process consists of a forward propagation process and a parameter updating process, and an iteration number N and an upper error limit error are set;
and 4, step 4: applying a time self-adaptive transient stability evaluation model on line, and sequentially carrying out high-level feature X at each moment after fault clearing t Performing transient stability evaluation to meet evaluation confidence or reach time window lengthThe evaluation is stopped, and the evaluation result of the online application time adaptive transient stability evaluation model is
Wherein alpha is a confidence factor, [0, alpha ] U (1-alpha, 1)]Is a confidence interval; with time-progressive input sequence, specifically, inputting X to the model after fault clearing 1 Stopping evaluation if the evaluation result is within the confidence interval, otherwise continuing inputting X 2 Combining the last time pair X 1 Evaluation of the resulting "memory" h 1 Making an assessment of power system stability; until an evaluation result is obtained or the evaluation time window length T is reached.
2. The time-adaptive online transient stability assessment method for the power system according to claim 1, wherein: in step 2, the single RBM unsupervised training process is as follows:
step 2.1, the units of the RBM hidden layer are mutually independent, and when the state v of the visual layer is given, the activation probability of the jth nerve unit of the hidden layer is
Wherein
b j Bias for the jth neural cell of the hidden layer, ω ij The connection weight between the ith visual layer unit and the jth hidden layer unit of the RBM network is given;
step 2.2, the units of the RBM hidden layer are mutually independent, and when the state h of the hidden layer is given, the activation probability of the ith nerve unit of the visual layer is
Wherein a is i Biasing for the ith neural cell of the hidden layer;
step 2.3, for a number s of input sets { v 1 ,v 2 ,…,v s At the input set by maximizing RBMLog likelihood function of
Obtaining a model parameter theta, namely extracting the output of the hidden layer as the characteristics of the visible layer, wherein P (v) k | h, θ) is the probability of a visible layer given the state h and parameter θ of the hidden layer, and is formulated as:
step 2.4, the parameter updating formula is as follows:
further, in the present invention,
in the formula, v (0) For the original input set, v (l) Is to v (0) Obtained by carrying out Gibbs sampling in one step, rho is momentum term, eta is learning rate, b i Offset of the ith neural unit of the visual layer, ω ij For the connection weight between the ith visual layer unit and the jth hidden layer unit of the RBM network, a i Biasing for the ith neural cell of the hidden layer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711260938.7A CN107993012B (en) | 2017-12-04 | 2017-12-04 | Time-adaptive online transient stability evaluation method for power system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711260938.7A CN107993012B (en) | 2017-12-04 | 2017-12-04 | Time-adaptive online transient stability evaluation method for power system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107993012A CN107993012A (en) | 2018-05-04 |
| CN107993012B true CN107993012B (en) | 2022-09-30 |
Family
ID=62035576
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711260938.7A Expired - Fee Related CN107993012B (en) | 2017-12-04 | 2017-12-04 | Time-adaptive online transient stability evaluation method for power system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107993012B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108763360A (en) * | 2018-05-16 | 2018-11-06 | 北京旋极信息技术股份有限公司 | A kind of sorting technique and device, computer readable storage medium |
| CN108960077A (en) * | 2018-06-12 | 2018-12-07 | 南京航空航天大学 | A kind of intelligent failure diagnosis method based on Recognition with Recurrent Neural Network |
| CN109239527A (en) * | 2018-09-07 | 2019-01-18 | 福州大学 | Distribution network failure recognition methods based on depth confidence network |
| CN109376939B (en) * | 2018-11-01 | 2021-08-10 | 三峡大学 | Power grid stability real-time prediction method based on adaptive neural network |
| CN110336277B (en) * | 2019-07-11 | 2021-03-30 | 福州大学 | Power system transient stability evaluation method based on deep belief network |
| CN112529684A (en) * | 2020-11-27 | 2021-03-19 | 百维金科(上海)信息科技有限公司 | Customer credit assessment method and system based on FWA _ DBN |
| CN114676717A (en) * | 2020-12-09 | 2022-06-28 | 电科云(北京)科技有限公司 | Bearing residual life prediction method, device and medium |
| CN113435113B (en) * | 2021-06-22 | 2024-06-11 | 清华大学 | Power system transient stability evaluation method and device |
| CN115775114A (en) * | 2022-12-29 | 2023-03-10 | 国网甘肃省电力公司经济技术研究院 | New energy station transient voltage stability evaluation method based on gated cycle unit |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103778467A (en) * | 2014-01-16 | 2014-05-07 | 天津大学 | Power system transient state stability estimation inputted characteristic quantity selection method |
| US8898093B1 (en) * | 2012-06-25 | 2014-11-25 | The Boeing Company | Systems and methods for analyzing data using deep belief networks (DBN) and identifying a pattern in a graph |
| CN105512808A (en) * | 2015-11-30 | 2016-04-20 | 武汉大学 | Power system transient stability assessment method based on big data |
| CN105608512A (en) * | 2016-03-24 | 2016-05-25 | 东南大学 | Short-term load forecasting method |
| CN106529476A (en) * | 2016-11-11 | 2017-03-22 | 重庆邮电大学 | Deep stack network-based electroencephalogram signal feature extraction and classification method |
| CN106709820A (en) * | 2017-01-11 | 2017-05-24 | 中国南方电网有限责任公司电网技术研究中心 | Power system load prediction method and device based on deep belief network |
| CN106991372A (en) * | 2017-03-02 | 2017-07-28 | 北京工业大学 | A kind of dynamic gesture identification method based on interacting depth learning model |
| US20170270410A1 (en) * | 2016-03-18 | 2017-09-21 | Michael James Denham | Neuromorphic training algorithm for a Restricted Boltzmann Machine. |
| CN107391852A (en) * | 2017-07-26 | 2017-11-24 | 清华大学 | Transient stability real time evaluating method and device based on depth confidence network |
-
2017
- 2017-12-04 CN CN201711260938.7A patent/CN107993012B/en not_active Expired - Fee Related
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8898093B1 (en) * | 2012-06-25 | 2014-11-25 | The Boeing Company | Systems and methods for analyzing data using deep belief networks (DBN) and identifying a pattern in a graph |
| CN103778467A (en) * | 2014-01-16 | 2014-05-07 | 天津大学 | Power system transient state stability estimation inputted characteristic quantity selection method |
| CN105512808A (en) * | 2015-11-30 | 2016-04-20 | 武汉大学 | Power system transient stability assessment method based on big data |
| US20170270410A1 (en) * | 2016-03-18 | 2017-09-21 | Michael James Denham | Neuromorphic training algorithm for a Restricted Boltzmann Machine. |
| CN105608512A (en) * | 2016-03-24 | 2016-05-25 | 东南大学 | Short-term load forecasting method |
| CN106529476A (en) * | 2016-11-11 | 2017-03-22 | 重庆邮电大学 | Deep stack network-based electroencephalogram signal feature extraction and classification method |
| CN106709820A (en) * | 2017-01-11 | 2017-05-24 | 中国南方电网有限责任公司电网技术研究中心 | Power system load prediction method and device based on deep belief network |
| CN106991372A (en) * | 2017-03-02 | 2017-07-28 | 北京工业大学 | A kind of dynamic gesture identification method based on interacting depth learning model |
| CN107391852A (en) * | 2017-07-26 | 2017-11-24 | 清华大学 | Transient stability real time evaluating method and device based on depth confidence network |
Non-Patent Citations (2)
| Title |
|---|
| Feature selection of power system transient stability assessment based on random;Chun Zhang 等;《2016 IEEE PES Asia-Pacific Power and Energy Engineering Conference (APPEEC)》;20161212;1264-1268 * |
| 基于机器学习的电力系统暂态稳定评估研究;叶圣永;《中国优秀博士学位论文全文数据库 工程科技Ⅱ辑（月刊）》;20100815(第8期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107993012A (en) | 2018-05-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107993012B (en) | Time-adaptive online transient stability evaluation method for power system | |
| CN112149316B (en) | Aero-engine residual life prediction method based on improved CNN model | |
| CN109308522B (en) | GIS fault prediction method based on recurrent neural network | |
| CN103226741B (en) | Public supply mains tube explosion prediction method | |
| CN102707256B (en) | Fault diagnosis method based on BP-Ada Boost nerve network for electric energy meter | |
| CN102722759B (en) | Method for predicting power supply reliability of power grid based on BP neural network | |
| CN109063247B (en) | Landslide disaster forecasting method based on deep belief network | |
| CN105354363A (en) | Fluctuation wind speed prediction method based on extreme learning machine | |
| CN111415010B (en) | Bayesian neural network-based wind turbine generator parameter identification method | |
| CN110826791A (en) | Hybrid wind power prediction method based on long-time and short-time memory neural network | |
| CN104951836A (en) | Posting predication system based on nerual network technique | |
| CN110942194A (en) | Wind power prediction error interval evaluation method based on TCN | |
| CN107563251B (en) | Fan fault diagnosis method based on extreme learning machine | |
| CN110807554A (en) | Generation method and system based on wind power/photovoltaic classical scene set | |
| CN108879732B (en) | Transient stability evaluation method and device for power system | |
| Nana et al. | Short-term wind speed prediction based on CNN_GRU model | |
| Shang et al. | Fault diagnosis method of rolling bearing based on deep belief network | |
| Li et al. | Application of ARIMA and LSTM in relative humidity prediction | |
| CN112632840A (en) | Power grid transient stability evaluation method based on adaptive differential evolution algorithm and ELM | |
| CN103399488B (en) | Multiple Model Control Method based on self study | |
| CN116227320A (en) | Double-fed fan control parameter identification method based on LSTM-IPSO | |
| CN109615109A (en) | Deep learning wind-powered electricity generation warning information analysis method based on Small Sample Database | |
| CN112712215A (en) | Wind power plant power prediction method | |
| CN115688574A (en) | Transient stability online evaluation method and system for power system | |
| CN112215279A (en) | Power grid fault diagnosis method based on immune RBF neural network |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20220930 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |