CN106777494A - A kind of Power System Reliability influence factor sensitivity computing method - Google Patents
A kind of Power System Reliability influence factor sensitivity computing method Download PDFInfo
- Publication number
- CN106777494A CN106777494A CN201611031272.3A CN201611031272A CN106777494A CN 106777494 A CN106777494 A CN 106777494A CN 201611031272 A CN201611031272 A CN 201611031272A CN 106777494 A CN106777494 A CN 106777494A
- Authority
- CN
- China
- Prior art keywords
- influence factor
- sensitivity
- index
- reliability
- link
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- 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
- G06N3/084—Backpropagation, e.g. using gradient descent
Abstract
The present invention provides a kind of influence factor sensitivity computing method of Power System Reliability, including：N bar links, the link is selected to meet following condition in the neutral net of reliability index：Each in each layer of neutral net to choose neuron one link of composition, the neuron being selected successively must be joined directly together successively；The sensitivity index of a certain influence factor in each bar link is calculated respectively, calculates sensitivity of the sensitivity index average of all links as the influence factor to reliability index.Its influence factor effectively can be divided into favorable factor and unfavorable factor by the present invention by the emulation experiment of real data, and according to contributed to the sensitivity of index, the sensitivity significance level of congeniality influence factor is carried out into quantization sequence.
Description
Technical field
The present invention relates to power information management domain, the influence factor more particularly, to Power System Reliability is sensitive
Degree computational methods.
Background technology
At present, the reliability of power system is closely related with the factor such as running environment and system operation planning mode, and shadow
The dynamic high dimensional nonlinear relation of complexity is often showed between the factor of sound and reliability index.It is traditional based on component reliability
The method of deduction system reliability, due to needing by historical statistics and metastable system architecture, to ensure correct description
Stochastic variable correlation and selection model structure, and nowadays, constantly decontrol and the demand such as new-energy grid-connected in electricity market
Under lasting stimulation, there is quick change, reliability of the tradition based on statistical inference in each regional distribution network system structure
Forecasting Methodology is difficult to effectively implement and applies.
The modeling pattern of data driven type is provided newly to set up the relation between fast-changing input data and output
Thinking.At present, reliability index modeling and the Numerical Predicting Method of data driven type have been researched and proposed.By what sword, Cheng Lin,
Sun Yuanzhang, Wang Peng were in 2008《Power system and protection》In, deliver《Meter and the Operation of Electric Systems reliability of weather forecasting
Property acute assessment》The power-system short-term reliability prediction model of meter and weather forecasting is proposed, using SVMs to temperature
Be predicted with wind speed, and derived the failure probability of system components to system cutting load probability, expected loss of load with
And the contribution degree of expected loss of energy.By Zhao Yuan, Zhou Niancheng, thank out to your grade in 2005 years《Electric power network technique》In deliver《Greatly
The sensitivity analysis of Model in Reliability Evaluation of Power Systems》Load-loss probability has been derived in one text, LOAD FREQUENCY has been lost and the electric power deficiency phase
The large power system reliability indexs such as prestige are to element availability, the sensitivity without validity, fault rate and repair rate.A.M.L.da
Silva,A.C.R.With L.C.Nascimento in 2014 up to being made in Rameau PMAPS international conferences
《Data calibration based on Monte Carlo simulation and evolutionary
optimization》One text is based on concept and sequential Meng Teka of the component reliability parameter to the sensitivity of Reliability Index
Lip river simulation algorithm, proposes a kind of component reliability parameters validation method.H.Ge and S.Asgarpoor, in the IEEE of 2014
Delivered on Transactions on Power Delivery《Reliability and Maintainability
Improvement of Substations With Aging Infrastructure》Propose that a kind of transformer station's availability is relative
In the sensitivity of its element availability, and each element is characterized with the normal value of the sensitivity for transformer station's reliability
Contribution.T.X.Zhu exists《A New Methodology of Analytical Formula Deduction and
Sensitivity Analysis of EENS in Bulk Power System Reliability Assessment》One text
In expect but delivery for composite electric Reliability Index, derived the sensitivity of element failure rate and repair rate.
R.Arya, S.Choube, L.Arya, and R.Shrivastava exist《Application of Sensitivity
Analysis for Improving Reliability Indices of a Radial Distribution System》In
Propose element failure rate and repair time optimization method based on component maintenance Cost Sensitivity Analysis of Industrial.On this basis, this hair
It is bright to propose a kind of distribution network reliability relation factor sensitivity computing method based on BP neural network.
The content of the invention
The present invention provides one kind and overcomes above mentioned problem or at least in part solution to the problems described above.
According to an aspect of the present invention, there is provided a kind of influence factor sensitivity computing method of Power System Reliability,
Including：Step 1, selects N bar links, the link to meet following condition in the neutral net of reliability index：In nerve net
Each in each layer of network to choose a neuron and constitute a link, the neuron being selected must successively direct phase successively
Even；Step 2, calculates the sensitivity index of a certain influence factor in each bar link respectively, calculates the sensitivity of all links
Sensitivity of the index average as the influence factor to reliability index.
The application proposes a kind of influence factor sensitivity computing method of Power System Reliability, and the method can effectively be known
Gei Ding not be under neural network model, the size and relative importance of the sensitivity of each relation factor of reliability index, are big
The management of system reliability and system Construction provide valuable numerical analysis tools under data environment.The present invention is by actual
Its influence factor effectively can be divided into favorable factor and unfavorable factor by the emulation experiment of data, and according to given index
Sensitivity contribution, the sensitivity significance level of congeniality influence factor is carried out into quantization sequence.
Brief description of the drawings
Fig. 1 is the overall flow figure schematic diagram according to the embodiment of the present invention；
Fig. 2 is the structural representation of the neutral net according to embodiment of the present invention reliability index；
Fig. 3 is the overall flow figure schematic diagram according to the embodiment of the present invention.
Specific embodiment
With reference to the accompanying drawings and examples, specific embodiment of the invention is described in further detail.Hereinafter implement
Example is not limited to the scope of the present invention for illustrating the present invention.
In Fig. 1, in a specific embodiment of the invention, show that a kind of influence factor of Power System Reliability is sensitive
Degree computational methods.On the whole, including：Step 1, selects N bar links, the link to expire in the neutral net of reliability index
It is enough to lower condition：It is each in each layer of neutral net to choose neuron one link of composition, the nerve being selected
Unit successively must be joined directly together successively；Step 2, calculates the sensitivity index of a certain influence factor in each bar link respectively,
Calculate sensitivity of the sensitivity index average of all links as the influence factor to reliability index.
In another specific embodiment of the invention, a kind of influence factor Calculation of Sensitivity side of Power System Reliability
Method, the data network is BP neural network.
In another specific embodiment of the invention, a kind of influence factor Calculation of Sensitivity side of Power System Reliability
Method, the reliability index can be：User's yearly average outage hours.
In another specific embodiment of the invention, the influence factor can be：Dielectric feeder ratio, interconnection ratio,
Segmentation and the quantity ratio of feeder line, unit segmental averaging number of users or average annual feeder line capacity-load ratio.
In Fig. 2, in another specific embodiment of the invention, a kind of influence factor sensitivity meter of Power System Reliability
Calculation method, as indicated, hidden layer, neuron number and transmission that the neutral net of the reliability index each network is included
Function.
In another specific embodiment of the invention, a kind of influence factor Calculation of Sensitivity side of Power System Reliability
Method, calculates the sensitivity index of a certain influence factor in each bar link, by following steps reality respectively described in step 2
It is existing.By continuous derivative rule, the reliability index and single influence factor under the neutral net of given reliability index are derived
Between parsing relational expression it is as follows：Symbol implication shown in a three-layer neural network similar to Figure 2, for comprising
One input layer (the 0th layer), M hidden layer and other neutral nets of output layer, the input and output of each level are led to
It is with relational model：
o(k+1)(j)=h(k+1)(n(k+1)(i)) (2)
Wherein, o(k+1)(j)=I (j),h(k)() represents the transmission function of kth layer.
Chain type Rule for derivation according to partial derivative, starts from last layer, i.e. M+1 layer, and from formula, (2)s carried out local derviation
Number asks for computing, you can to obtain：
Wherein I is reliability index, and F is influence factor, M be the reliability index neutral net in hidden layer
Number, i is neuron numbering in each layer.
In another specific embodiment of the invention, a kind of influence factor Calculation of Sensitivity side of Power System Reliability
Method, also includes before the step 1：Primary increasing is carried out to each influence factor data small sample of reliability using sample augmentation technology
Extensively, primary augmented matrix A is obtainedE；Based on primary augmented matrix AECarry out single output nerve network training.
In Fig. 3, carried out by taking user's yearly average outage hours (TOH) index as an example in another specific embodiment of the invention
The introduction of sensitivity computing method, shows a kind of influence factor sensitivity computing method of Power System Reliability.Including following
Step：(1) for reliability index TOH, three different BP neural network reliability models, each selected net are selected
Hidden layer, neuron number and the transmission function that network is included are as shown in table 1 below.
(2) Levenberg Marquardt are adopted using Matlab Neural Network Toolbox to each network
Optimization algorithms are trained, and other parameters use software kit default setting；Influence factor and reliability index
Definition and related raw statistical data are summarized in following table.
(3) after 100 times of expansions being carried out to initial data using cuclear density fitting technique, to each BP nerves, controlled training
Precision is less than 0.019 for the mean square deviation of sample predictions and actual value.
(4) according to given artificial neural network reliability model, according to the definition to " neuron link ", one is found
Bar meets the index TOH of condition and can turn " the neuron link " for feeder line ratio.Calculate the sensitivity of this neuron link
IndexAnd take the sensitivity average of all links and can turn for sensitivity of the feeder line ratio to index TOH as influence factor.
(5) for reliability index TOH, step in paragraph is independently repeated successively, you can obtain reliability index TOH
To its influence factor " dielectric feeder ratio ", " interconnection ratio ", " the quantity ratio of segmentation and feeder line " " unit segmental averaging user
The sensitivity of number " and " average annual feeder line capacity-load ratio ".
For each reliability index, independently repeat the above steps successively (three) to (five), you can obtain each reliability
Sensitivity of the property index to its influence factor.
In another specific embodiment of the invention, a kind of influence factor Calculation of Sensitivity side of Power System Reliability
Method, also includes before the step 1：Primary increasing is carried out to each influence factor data small sample of reliability using sample augmentation technology
Extensively, primary augmented matrix A is obtainedE；Based on primary augmented matrix AECarry out single output nerve network training.
In another specific embodiment of the invention, a kind of influence factor Calculation of Sensitivity side of Power System Reliability
Method, it is described that each neural metwork training is included：Levenberg Marquardt are adopted using Matlab Neural Network Toolbox
Optimization algorithms are trained, and other parameters use software kit default setting.
In another specific embodiment of the invention, a kind of influence factor Calculation of Sensitivity side of Power System Reliability
Method, primary augmentation is carried out in the step 1 using sample augmentation technology to each influence factor data small sample of reliability, bag
Include：Augmentation is carried out to small sample using cuclear density fitting technique.
In another specific embodiment of the invention, a kind of influence factor Calculation of Sensitivity side of Power System Reliability
Method, to each BP nerves, controlled training precision is less than 0.019 for the mean square deviation of sample predictions and actual value.
Finally, the present processes are only preferably embodiment, are not intended to limit the scope of the present invention.It is all
Within the spirit and principles in the present invention, any modification, equivalent substitution and improvements made etc. should be included in protection of the invention
Within the scope of.
Claims (9)
1. the influence factor sensitivity computing method of a kind of Power System Reliability, it is characterised in that including：
Step 1, selects N bar links, the link to meet following condition in the neutral net of reliability index：In neutral net
Each layer in it is each choose a neuron and constitute a link, the neuron being selected must successively direct phase successively
Even；
Step 2, calculates the sensitivity index of a certain influence factor in each bar link respectively, calculates the sensitivity of all links
Sensitivity of the index average as the influence factor to reliability index.
2. the method for claim 1, it is characterised in that the data network is BP neural network.
3. the method for claim 1, it is characterised in that the reliability index can be：User has a power failure hour every year
Number.
4. the method for claim 1, it is characterised in that the influence factor can be：Dielectric feeder ratio, interconnection ratio
Example, the quantity ratio of segmentation and feeder line, unit segmental averaging number of users or average annual feeder line capacity-load ratio.
5. the method for claim 1, it is characterised in that calculate a certain in each bar link described in step 2 respectively
The sensitivity index of influence factor, including：
WhereinIt is sensitivity index, I is reliability index, and F is influence factor, and M is the neutral net of the reliability index
The number of middle hidden layer, i is neuron numbering in each layer.
6. the method for claim 1, it is characterised in that also include before the step 1：Can using sample augmentation technology pair
Each influence factor data small sample by property carries out primary augmentation, obtains primary augmented matrix AE；Based on primary augmented matrix AEEnter
Row list output nerve network training.
7. method as claimed in claim 6, it is characterised in that described to include to each neural metwork training：Using Matlab god
Levenberg Marquardt optimization algorithms are adopted through network tool case to be trained, other parameters use software
Bag default setting.
8. method as claimed in claim 6, it is characterised in that using sample augmentation technology to reliability in the step 1
Each influence factor data small sample carries out primary augmentation, including：Augmentation is carried out to small sample using cuclear density fitting technique.
9. method as claimed in claim 6, it is characterised in that to each BP nerves, controlled training precision be sample predictions and
The mean square deviation of actual value is less than 0.019.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611031272.3A CN106777494B (en) | 2016-11-17 | 2016-11-17 | Method for calculating sensitivity of reliability influence factors of power system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611031272.3A CN106777494B (en) | 2016-11-17 | 2016-11-17 | Method for calculating sensitivity of reliability influence factors of power system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106777494A true CN106777494A (en) | 2017-05-31 |
| CN106777494B CN106777494B (en) | 2020-11-03 |
Family
ID=58971542
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611031272.3A Active CN106777494B (en) | 2016-11-17 | 2016-11-17 | Method for calculating sensitivity of reliability influence factors of power system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106777494B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110414086A (en) * | 2019-07-10 | 2019-11-05 | 北京华安中泰检测技术有限公司 | A kind of combined stress accelerated factor calculation method based on sensitivity |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101635457A (en) * | 2009-05-14 | 2010-01-27 | 国家电网公司 | Electric network parameter estimation method based on parameter sensitivity of state estimation residual error |
| CN103384069A (en) * | 2013-06-13 | 2013-11-06 | 广东电网公司电力科学研究院 | Method and system for adjusting sensitivity of offshore wind plant current collecting system |
| CN105932775A (en) * | 2016-05-23 | 2016-09-07 | 天津大学 | Analysis method for influence from information system to operational reliability of micro-grid |
-
2016
- 2016-11-17 CN CN201611031272.3A patent/CN106777494B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101635457A (en) * | 2009-05-14 | 2010-01-27 | 国家电网公司 | Electric network parameter estimation method based on parameter sensitivity of state estimation residual error |
| CN103384069A (en) * | 2013-06-13 | 2013-11-06 | 广东电网公司电力科学研究院 | Method and system for adjusting sensitivity of offshore wind plant current collecting system |
| CN105932775A (en) * | 2016-05-23 | 2016-09-07 | 天津大学 | Analysis method for influence from information system to operational reliability of micro-grid |
Non-Patent Citations (2)
| Title |
|---|
| 周竞: "发输电互联系统风险评估的解析模型研究", 《中国优秀硕士学位论文全文数据库（工程科技II辑）》 * |
| 宋云亭 等: "基于BP神经网络的城网供电可靠性预测方法", 《电网技术》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110414086A (en) * | 2019-07-10 | 2019-11-05 | 北京华安中泰检测技术有限公司 | A kind of combined stress accelerated factor calculation method based on sensitivity |
| CN110414086B (en) * | 2019-07-10 | 2023-01-17 | 北京华安中泰检测技术有限公司 | Sensitivity-based comprehensive stress acceleration factor calculation method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106777494B (en) | 2020-11-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Mohammed et al. | Available transfer capability calculation methods: A comprehensive review | |
| Nick et al. | Security constrained unit commitment with dynamic thermal line rating | |
| Wang et al. | Solving corrective risk-based security-constrained optimal power flow with Lagrangian relaxation and Benders decomposition | |
| CN105932723B (en) | A kind of grid structure Method for optimized planning of alternating current-direct current mixing micro-capacitance sensor | |
| Wang et al. | Reliability evaluation of integrated electricity–gas system utilizing network equivalent and integrated optimal power flow techniques | |
| Liang et al. | Probability-driven transmission expansion planning with high-penetration renewable power generation: A case study in northwestern China | |
| Qiu et al. | A risk-based approach to multi-stage probabilistic transmission network planning | |
| Awad et al. | Optimal distributed generation allocation and load shedding for improving distribution system reliability | |
| Zhao et al. | Resilient unit commitment for day-ahead market considering probabilistic impacts of hurricanes | |
| CN107908638A (en) | The operation of power networks efficiency rating method and system matched somebody with somebody are excavated based on big data | |
| Agajie et al. | Reliability enhancement and voltage profile improvement of distribution network using optimal capacity allocation and placement of distributed energy resources | |
| Mustafa et al. | Reactive power tracing in pool-based power system utilising the hybrid genetic algorithm and least squares support vector machine | |
| CN110601204A (en) | Photovoltaic grid-connected system probability power flow analysis method based on random variable state time sequence simulation | |
| Yi et al. | Optimal allocation of ESSs in active distribution networks to achieve their dispatchability | |
| Qiu et al. | Multi‐objective transmission expansion planning in a smart grid using a decomposition‐based evolutionary algorithm | |
| Guo et al. | Evidence-based approach to power transmission risk assessment with component failure risk analysis | |
| CN108694479A (en) | Consider the distribution network reliability prediction technique that weather influences time between overhaul | |
| Abdalla et al. | Generation expansion planning under correlated uncertainty of mass penetration renewable energy sources | |
| CN113569411B (en) | Disaster weather-oriented power grid operation risk situation awareness method | |
| Helseth et al. | A hydrothermal market model for simulation of area prices including detailed network analyses | |
| CN106777494A (en) | A kind of Power System Reliability influence factor sensitivity computing method | |
| Ivanov et al. | Intelligent monitoring and control in transmission and distribution networks | |
| Silva et al. | Evaluating the effectiveness of parametric and nonparametric energy consumption forecasts for a developing country | |
| Chi et al. | Multi‐stage coordinated dynamic VAR source placement for voltage stability enhancement of wind‐energy power system | |
| Jayaweera | Security enhancement with nodal criticality-based integration of strategic micro grids |
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 | ||
| CB03 | Change of inventor or designer information |
Inventor after: Zhang Jiangong Inventor after: Hu Qinghui Inventor after: Tian Hongxun Inventor after: Wang Honggang Inventor after: Wan Tao Inventor after: Li Haosong Inventor after: Li Jin Inventor after: Kang Taifeng Inventor after: Li Xin Inventor before: Zhang Jiangong Inventor before: Qing Hui Inventor before: Tian Hongxun Inventor before: Wang Honggang Inventor before: Wan Tao Inventor before: Li Haosong Inventor before: Li Jin Inventor before: Kang Taifeng Inventor before: Li Xin |
|
| CB03 | Change of inventor or designer information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |