CN106599337A - Power grid frequency simulation parameter identification method based on simplex method - Google Patents
Power grid frequency simulation parameter identification method based on simplex method Download PDFInfo
- Publication number
- CN106599337A CN106599337A CN201610889021.2A CN201610889021A CN106599337A CN 106599337 A CN106599337 A CN 106599337A CN 201610889021 A CN201610889021 A CN 201610889021A CN 106599337 A CN106599337 A CN 106599337A
- Authority
- CN
- China
- Prior art keywords
- parameter
- group
- value
- simulation
- frequency
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention belongs to the field of an electrical power system and the automation technology thereof, and discloses a power grid frequency simulation parameter identification method based on a simplex method. The method comprises the following steps of: firstly, utilizing a multi-parameter sensitivity analysis method to determine a simulation parameter group to be identified; then, utilizing a simplex algorithm based on a numerical analysis method to intelligently guide a parameter optimizing direction; and utilizing the simulation result of the electromechanical transient simulation program BPA (Bonneville Power Administration) of the electrical power system to calculate a target function value so as to quickly and reliably search an optimal parameter combination. By use of the method, one set of parameters which enable the simulation result to have a smallest error compared with a practical response result can be identified, the accuracy of power grid frequency simulation in the future is improved, simulation time consumption can be shortened, and the optimization efficiency of the parameter identification can be greatly improved.
Description
Technical field
The invention belongs to Power System and its Automation technical field, is specifically related to a kind of electrical network based on simplex method
Frequency simulation parameter discrimination method.
Background technology
September 19 days 21 in 2015:58:Occur when 02, brocade Soviet Union DC bipolar block, East China Power Grid power shortage reach 490
Ten thousand kilowatts, before failure, system frequency is 49.97Hz, and after 12s, the whole network frequency is minimum falls to 49.56Hz, falls amplitude for 0.41Hz.
Under the premise of the identical method of operation, the frequency calculated based on the electromechanical transient simulation program (BPA) under canonical parameter falls width
Value is only 0.17Hz.Although current accident does not cause very serious consequence, but reflect the current frequency of electrical network and imitate
True precision is barely satisfactory, has much room for improvement.
With building up for extra-high voltage alternating current-direct current serial-parallel power grid, when there is commutation failure, locking failure in DC transmission system,
Active power shortage is huge, and this kind of system low frequency problem lost after big power supply directly depends on the devices such as prime mover and speed regulator
Response.For power system, either calculated off line analysis, or on line real-time monitoring, it is necessary to set up accurately,
On the basis of believable numerical model or equivalent parameters.Therefore, in this case it is desirable to based on meeting actual frequency response
System mathematic model, finds out the high relevant parameter of sensitivity, obtains each parameter concrete numerical value by recognizing, and then the ginseng that will be recognized
Number is applied in simulation calculation, is easy to traffic control personnel to carry out the formulation of emergent control measure.
At present, both at home and abroad the dynamic modeling that identification technique is used for frequency response system is made slow progress, some power systems
The master pattern of the frequency response system relevant device that simulation software (such as BPA, PSS/E etc.) is provided, especially determines once to adjust
The devices such as prime mover and speed regulator of frequency ability, use typical model and " representative value " that not experiment proves that, and actual
In system, the Primary frequency control ability of unit reaches far away this ideal value, directly results in the system frequency under this high-power vacancy
Rate response simulation result is significantly optimistic.
It is the important measures for improving simulation Credibility using actual measurement system perturbed Newton method and checking frequency emulation relevant parameter.
The research meanses of parameter identification are broadly divided into two kinds of analytical technique and numerical analysis method.The former emphasizes the side using mathematical analyses
Method, lists equation and real system is simulated, carry out dimensionality reduction to system features matrix by some mathematical methods, finally solve, obtain
Obtain the corresponding relation between parameter and aim parameter；The latter is mainly affected on track by using each parameter in system, compares actual measurement
Solve problem is carried out in track, checking emulation.The impacts such as the Primary frequency control ability due to reflecting unit, load factor are all non-linear
Dynamic model, this complicated non-linear dynamical behavior are difficult analysis mode and express, and cause analytical analysis and method for designing
There is larger difficulty.And numerical analysis method has certain advantage, and fortune in terms of the identification of nonlinear dynamical model parameter
With many methods must be compared.
Patent at present in terms of frequency emulates relevant parameter discrimination method is less.Patent application《Based on simplex method
Nuclear power generating sets prime mover and its governor parameter identification method》(application number CN201210252089.1) is there is provided a kind of base
In nuclear power generating sets prime mover and its governor parameter discrimination method of simplex method, but the method can only be entered for single unit
Line parameter is recognized, and is applied to that bulk power grid obvious operand in East China is huge, and efficiency is too low.Patent application《Frequency locus and population
The governor parameter discrimination method that algorithm combines》(application number CN201310236183.2) is proposed based on Particle Swarm Optimization
Method is adjusted to governor parameter so that simulation frequency curve and practical frequency curve are as close possible to have ignored big net completely
Impact of the middle load factor to frequency, not with general applicability.Patent application《Power system prime mover and its speed regulator actual measurement
Modeling method》(application number CN201010107126.0) proposes the modeling method based on field measurement, although ensure that mould
The accuracy of type, parameter, ensure that the simulation accuracy of power system, but at present for East China Power Grid, all units
Field measurement can't realize that and the accuracy of the site test results under big frequency need to be investigated.
The content of the invention
Goal of the invention：In order to overcome the deficiencies in the prior art, the present invention to provide a kind of using simplex method generation
New one group of parameter group, the method ignore the diversity of each unit and load in net, by the Primary frequency control ability of the whole network unit with it is negative
Lotus level is fitted to unified model and unified parameters, and the accuracy with frequency simulation result gives a kind of determination and treats as target
Identified parameters and the method for realizing the optimum identification of parameter, so that it is determined that the relevant parameter that powernet is emulated with offline frequency, pole
The earth improves the accuracy of frequency emulation of the electrical network under large disturbances failure.
Technical scheme：For achieving the above object, a kind of mains frequency simulation parameter based on pure algorithm of the invention
Discrimination method, comprises the following steps：
1) operating condition when collecting grid simulation data and being occurred according to power grid accident adjusts the grid simulation data, institute
Stating grid simulation data includes power flow files and stablizes file, and the stable file includes the unit related to frequency emulation and bears
Lotus model；
Ripple is recorded in practical frequency response after acquisition causes the concrete large disturbances failure of mains frequency fluctuation and breaks down
Data；
2) determining affects the n dimension parameter space X=(x of frequency simulation result1,x2,…,xn), the parameter space is and waits to distinguish
Know parameter, and the span in the practical situation arrange parameter space according to electrical network；
3) determine initial simplex：The frequency emulation relevant parameter of electrical network adopts representative value, and each Parameter Typical is made
For one group of initial value parameter group, other n groups parameter group is produced in n dimension parameter spaces based on the initial value parameter group, then has n+1
Group parameter group, the n+1 groups parameter group constitutes initial simplex, if iterationses k=0；
4) each parameter group is obtained in n+1 group parameter group identical using the emulation of Electrical-Machanical Transient Simulation of Power System program BPA
Fault disturbance under corresponding frequency response results curve Y, calculate the corresponding target function value of each parameter group in n+1 group parameter group
J, finds out maximum target functional value JHWith minimum target functional value JLAnd calculate maximum target functional value JHWith minimum target functional value
JLBetween difference, if the difference meets convergence criterion, stop identification；Step 5 is entered otherwise)；
5) iterationses k=k+1 is set, if iterationses k is not up to iterationses higher limit K, is given birth to using simplex method
Cheng Xinyi group parameter group, new one group of parameter group be used for replacement step 4) described in maximum target functional value JHCorresponding parameter
Group, ultimately forms new simplex, and return to step is 4)；If iterationses k reaches iterationses higher limit K, parameter identification
Failure, method ends.
Further, the step 2) middle determination n dimension parameter space X=(x1,x2,…,xn), using based on multiparameter spirit
The analysis method of sensitivity, comprises the following steps：
2-a) the measured value according to each unit and load model parameters in electrical network, it is determined that emulating phase with frequency in stable file
The each unit for closing and the maximum and minima of the parameter of load model, if diThe parameter of i-th unit or load model is represented,
dimaxRepresent diMaximum, diminRepresent diMinima, then the span of the parameter of i-th unit or load model be
[dimin,dimax], wherein i=1,2 ..., t, t be parameter sum；
The intermediate value of each unit and the parameter of load model is substituted into into phantom 2-b), emulation obtains the frequency after fault disturbance
Extreme value f of rate responseminWith steady-state value f after 60 secondss, the intermediate value of the parameter of each unit and load model be equal to its maximum and
The half of minima sum, i.e., for diFor, its intermediate value is equal to
2-c) for each parameter, m equally distributed independent random number is generated in span, i.e., for diFor
Symbiosis is into m equally distributed independent random number dij, wherein j=1,2 ..., m；Then for all parameters, generate altogether
mtGroup set of parameter values；Respectively by the mtGroup set of parameter values substitutes into phantom, and emulation draws every group of set of parameter values identical
Fault disturbance under corresponding frequency response extreme value fuminWith steady-state value f after 60 secondsus, wherein u=1,2 ..., mt；
2-d) the m that will be generated respectivelytGroup set of parameter values simulation result and step 2-b) in intermediate value emulation knot
Fruit is compared, according to formula (1) parameter value Q, while the median for taking all Q-value is defined as base value θ, if Q is ＞ θ,
Think that this group of set of parameter values is acceptable, this group of set of parameter values is distributed in acceptable group；Otherwise it is assumed that be can not
Receive, this group of set of parameter values is distributed in unacceptable group, so as to complete mtThe distribution of group set of parameter values：
Q=| fumin-fmin|+|fus-fs| formula (1)；
2-e) compare the distribution situation of each parameter, for arbitrary parameter, defining that its j-th independent random number be distributed in can
The number of acceptance group is pj, the number for being distributed in unacceptable group is qj, while sensitive parameter δ is defined, ifThen for the parameter, acceptable group is similar with unacceptable component cloth, then show that the parameter is unwise
Sense；Conversely, then showing the parameter sensitivity；If the sum of whole sensitive parameters is calculated as n, then n dimension parameter space X to be identified are constituted
=(x1,x2,…,xn)。
Further, the step 3) in other n groups parameter group is generated based on the initial value parameter group, specifically include following
Step：N is tieed up into parameter space X=(x1,x2,…,xn) in each Parameter Typical it is as one group of initial value parameter group and this is first
In initial value parameter group, each Parameter Typical deviates the Parameter Typical most in replacing with the corresponding parameter of the Parameter Typical successively
Remote measured value, individually replaces a parameter, every time so as to form other n groups parameter group.
Further, the step 4) according to formula (2) calculating target function value J：
Wherein, α is sampling number, yiFor i-th point value of frequency response results curve Y, aiFor PMU frequencies after failure
I-th point value of response curve a；
If maximum target functional value is JH, minimum target functional value is JL, draw JHParameter group be XH, draw JLParameter
Group is set to XL, concurrently set convergenceIfThen stop identification, display parameters are recognized successfully, parameter group XL
For optimized parameter group.
Further, the step 5) in generate new one group of parameter group using simplex method, specifically include following steps：Will
Maximum target functional value JHCorresponding parameter group XHReflection, obtains reflection parameters group XR, and calculate reflection parameters group XRIt is corresponding
Target function value JR, by generating compression parameters group or expansion parameters group come alternative parameter group XH, form new with other specification group
Simplex.
Further, in the stable file, the unit related to frequency emulation includes prime mover, speed regulator and boiler main
Steam pressure model.
Beneficial effect：The present invention compared with the prior art, this have the advantage that：
1st, the present invention proposes distinguishing for relevant parameter for the not accurate enough situation of current mains frequency emulation relevant parameter
Knowledge method, first with multiparameter Sensitivity Analysis Method, determines simulation parameter group to be identified, then using based on numerical value point
The simplex algorithm intelligent guidance parameter optimization direction of analysis method, reuses the imitative of Electrical-Machanical Transient Simulation of Power System program (BPA)
True result calculating target function value, searches out best parameter group so as to fast and reliable；
2nd, reliable phantom is provided to calculate with offline safety and stability online, can improve electrical network in large disturbances event
Frequency simulation accuracy under barrier, the formulation for the corresponding frequency urgent control measure of East China Power Grid provide support, lift online peace
Full stability analysis, early warning and control decision level；
3rd, the present invention is entered using the nonlinear parameter optimization method that simplex algorithm and electromechanical transient simulation program combine
Line frequency emulates the parameter identification of correlation model, so as to pick out a set of so that simulation result is minimum with real response resultant error
Parameter, so as to improve the accuracy of the emulation of mains frequency from now on.
4th, using the algorithm, emulation can be shortened and is taken, greatly improve the optimization efficiency of parameter identification.
Description of the drawings
Fig. 1 is the flow chart of the inventive method.
Fig. 2 is the response curve feature extraction schematic diagram that the inventive method is applied in East China Power Grid " 919 " accident.
Specific embodiment
Below in conjunction with the accompanying drawings the present invention is further described.
Embodiment：
Reference Fig. 1, a kind of mains frequency simulation parameter discrimination method based on pure algorithm proposed by the present invention, first
Start Data Preparation into step S1, operation work when specifically collecting grid simulation data and occurring according to power grid accident
Condition adjusts the grid simulation data, and wherein grid simulation data include power flow files and stablize file, wherein stably file includes
The unit related to frequency emulation and load model, need in addition to obtain the concrete large disturbances failure for causing mains frequency fluctuation and
Practical frequency response recorder data after breaking down；
Subsequently into step S2, it is determined that affecting the n dimension parameter space X=(x of frequency simulation result1,x2,…,xn), the ginseng
Number space is parameter (parameter i.e. to be optimized) to be identified, can be according to the value model in the practical situation arrange parameter space of electrical network
Enclose.Wherein, determination method for parameter to be identified, using the analysis method based on multiparameter sensitivity, method is as follows：
Measured values of the S21 according to each unit and load model in electrical network, determines related to frequency emulation in stable file
Each unit parameter and load model parameters in each parameter maximum and minima, use di(i=1,2 ..., t) respectively
Each unit parameter and load model parameters are represented, wherein t is parameter sum, and the maximum and minima of each parameter are expressed as
dimaxAnd dimin, diSpan be [dimin,dimax]；
S22 is by the intermediate value of each parameterPhantom is substituted into, will Phantom is substituted into, extreme value f of the frequency response after obtaining fault disturbance is emulatedmin
With steady-state value f after 60 secondss；
S23 generates m equally distributed independent random number d respectively for each parameter in generation in spanij(j
=1,2 ..., m), the value of general m takes 3-5, and statistical representation isWhereinIt is expressed as ginseng
Number d1It is [d in span1min,d1max] interior m equally distributed independent random number of generation respectively, by that analogy,Table
It is shown as parameter dtIt is [d in spantmin,dtmax] interior m equally distributed independent random number of generation respectively.Then for all
For parameter, m is generated altogethertGroup set of parameter values, such as [d11,d21,...,dt1] it is expressed as one group of set of parameter values.Respectively will
The m of generationtGroup set of parameter values substitutes into phantom, show that every group of set of parameter values is disturbed in identical failure using BPA emulation
Extreme value f of corresponding frequency response under dynamicuminWith steady-state value f after 60 secondsus, wherein (u=1,2 ..., mt)；
The m that S24 will be generated respectivelytThe simulation result of the intermediate value in simulation result and step S22 of group set of parameter values enters
Row compares, while the median for taking all Q-value defines base value θ, according to formula (1) parameter value Q, if Q is ＞ θ, then it is assumed that should
Group set of parameter values is acceptable, and this group of set of parameter values is distributed in acceptable group；Otherwise it is assumed that be it is unacceptable,
This group of set of parameter values is distributed in unacceptable group：
Q=| fumin-fmin|+|fus-fs| formula (1)；
The distribution situation of the more each parameters of S25, for arbitrary parameter, defines its j-th independent random number and is distributed in and can connect
The number organized is pj, the number for being distributed in unacceptable group is qj, while sensitive parameter δ is defined, if
Then for the parameter, acceptable group is similar with unacceptable component cloth, then show that the parameter is insensitive；Conversely, then showing
The parameter sensitivity.If the sum of whole sensitive parameters is n, n dimension parameter space X=(x to be identified are constituted1,x2,…,xn)；Its
In, sensitive parameter δ is the integer more than zero, and its value depends on the simulation accuracy that practical study needs, and required precision is higher, and δ takes
Value is less.
Subsequently into step S3, initial simplex is determined, the frequency emulation relevant parameter of current electrical network adopts representative value,
Each Parameter Typical can be expressed as into (x ' as one group of initial value parameter group in the present invention1,x′2,…,x′n), it is first based on this
Initial value parameter group generates other n groups parameter group, then have n+1 group parameter group, and the n+1 groups parameter group constitutes initial simplex, if
Iterationses k=0；Other n groups parameter group is generated based on the initial value parameter group, specifically by initial value parameter group (x '1,x
′2,…,x′n) in each Parameter Typical individually replace every time, formed one group of other parameter group.Parameter Typical is replaced during replacement
Be changed to the corresponding parameter of the Parameter Typical it is actually measured deviate the farthest measured value of the Parameter Typical, so generation altogether
N+1 group parameter group, for example, can be expressed asWherein x1It is parameter x1Representative value x '1Corresponding ginseng
Number x1Actually measured deviation Parameter Typical x '1Farthest measured value, by x '1Replace with x "1, by that analogy；Then n+1 groups
Parameter group constitutes initial simplex；
Subsequently into step S4, judge whether to reach convergence criterion, it is imitative using Electrical-Machanical Transient Simulation of Power System program BPA
Each parameter group corresponding frequency response results curve Y under identical fault disturbance in n+1 group parameter group is really obtained, n+1 is calculated
In group parameter group, the corresponding target function value J of each parameter group, finds out maximum target functional value JHWith minimum target functional value JLAnd count
Calculate maximum target functional value JHWith minimum target functional value JLBetween difference, if the difference meets convergence criterion, stop excellent
Change；Step 5 is entered otherwise), wherein judge whether that the method for reaching convergence criterion is as follows, according to formula (2) calculating target function
Value J：
Wherein, α is sampling number, yiFor i-th point value of frequency response results curve Y, aiFor PMU frequencies after failure
I-th point value of response curve a；If the target function value for finally giving is less, the model parameter emulation knot set up is illustrated
Fruit is less with the error of fitting of measured curve, and the precision of frequency phantom is also higher；
By the maximum target functional value J for obtainingHCorresponding parameter group is set to XH, minimum target functional value JLCorresponding parameter
Group is set to XL, concurrently set convergenceIfThen stop optimization, display parameters are recognized successfully, parameter group XL
As optimized parameter group, otherwise into S5；Wherein, convergenceDepending on the simulation accuracy that practical study needs, required precision
It is higher, convergenceIt is less.
Step S5：Iterationses k=k+1 is made, if iterationses higher limit K that iterationses k not up to sets, utilizes
Simplex method generates new one group of parameter group, new one group of parameter group be used for replacement step 4) described in maximum target functional value JHIt is right
Parameter group X answeredH, new simplex is ultimately formed, and return to step is 4)；If iterationses k reaches iterationses higher limit K,
Then parameter optimization failure, method ends, wherein generating new one group of parameter group using simplex method, specifically include following steps：
By maximum target functional value JHCorresponding parameter group XHReflection, obtains reflection parameters group XR, and calculate reflection parameters group XRCorrespondence
Target function value JR, by generating compression parameters group or expansion parameters group XNCarry out alternative parameter group XH, formed with other specification group
New simplex, for example：In n+1 group parameter groupIn, if by [x '1,x′2,…,x′n] it is set to maximum target
Functional value JHCorresponding parameter group XH, then by [x '1,x′2,…,x′n] replace with parameter group XN, regenerate new n+1 group parameters
Group is new simplex, reenters step S4, parameter group is optimized, obtain optimized parameter group, in optimized parameter group
Lower simulation frequency, it is better.
Unit related to frequency emulation in aforementioned stable file includes prime mover, speed regulator and boiler main in the present invention
Steam pressure model.
With reference to Fig. 2, it is that the response curve feature extraction that the inventive method is applied in East China Power Grid " 919 " accident is illustrated
Figure, it can be seen that the response curve of actual measurement is bent with the frequency response under the optimized parameter recognized by the inventive method
Line is fitted substantially, and the response curve under canonical parameter is with the response curve surveyed and misfits, and error is larger, furtherly
Understand the accuracy that mains frequency emulation is improve using the inventive method.
The above is only the preferred embodiment of the present invention, it should be pointed out that：For those skilled in the art come
Say, under the premise without departing from the principles of the invention, can also make some improvements and modifications, these improvements and modifications also should be regarded as
Protection scope of the present invention.
Claims (6)
1. a kind of mains frequency simulation parameter discrimination method based on simplex method, it is characterised in that comprise the following steps：
1) operating condition when collecting grid simulation data and being occurred according to power grid accident adjusts the grid simulation data, the electricity
Network simulation data include power flow files and stablize file, and the stable file includes the unit related to frequency emulation and load mould
Type；
Practical frequency after acquisition causes the concrete large disturbances failure of mains frequency fluctuation and breaks down responds recorder data；
2) determining affects the n dimension parameter space X=(x of frequency simulation result1,x2,…,xn), the parameter space is ginseng to be identified
Number, and the span in the practical situation arrange parameter space according to electrical network；
3) determine initial simplex：The frequency emulation relevant parameter of electrical network adopts representative value, using each Parameter Typical as one
Group initial value parameter group, produces other n group parameter group in n dimension parameter spaces based on the initial value parameter group, then has n+1 groups ginseng
Array, the n+1 groups parameter group constitute initial simplex, if iterationses k=0；
4) it is former in identical each parameter group in n+1 group parameter group to be obtained using the emulation of Electrical-Machanical Transient Simulation of Power System program BPA
Corresponding frequency response results curve Y under barrier disturbance, calculates the corresponding target function value J of each parameter group in n+1 group parameter group, looks for
Go out maximum target functional value JHWith minimum target functional value JLAnd calculate maximum target functional value JHWith minimum target functional value JLIt
Between difference, if the difference meets convergence criterion, stop identification；Step 5 is entered otherwise)；
5) iterationses k=k+1 is set, if iterationses k is not up to iterationses higher limit K, is generated using simplex method new
One group of parameter group, new one group of parameter group be used for replacement step 4) described in maximum target functional value JHCorresponding parameter group, most
The simplex of end form Cheng Xin, and return to step is 4)；If iterationses k reaches iterationses higher limit K, parameter identification fails,
Method ends.
2. the mains frequency simulation parameter discrimination method based on simplex method according to claim 1, it is characterised in that：Institute
State step 2) middle determination n dimension parameter space X=(x1,x2,…,xn), the analysis method based on multiparameter sensitivity is adopted, including
Following steps：
2-a) the measured value according to each unit and load model parameters in electrical network, it is determined that related to frequency emulation in stable file
The maximum and minima of the parameter of each unit and load model, if diRepresent the parameter of i-th unit or load model, dimax
Represent diMaximum, diminRepresent diMinima, then the span of the parameter of i-th unit or load model be
[dimin,dimax], wherein i=1,2 ..., t, t be parameter sum；
The intermediate value of each unit and the parameter of load model is substituted into into phantom 2-b), emulation obtains the frequency sound after fault disturbance
Extreme value f answeredminWith steady-state value f after 60 secondss, the intermediate value of the parameter of each unit and load model is equal to its maximum and minimum
The half of value sum, i.e., for diFor, its intermediate value is equal to
2-c) for each parameter, m equally distributed independent random number is generated in span, i.e., for diFor symbiosis
Into m equally distributed independent random number dij, wherein j=1,2 ..., m；Then for all parameters, m is generated altogethertGroup
Set of parameter values；Respectively by the mtGroup set of parameter values substitutes into phantom, and emulation draws every group of set of parameter values in identical
Extreme value f of corresponding frequency response under fault disturbanceuminWith steady-state value f after 60 secondsus, wherein u=1,2 ..., mt；
2-d) the m that will be generated respectivelytThe simulation result and step 2-b of group set of parameter values) in the simulation result of intermediate value carry out
Relatively, according to formula (1) parameter value Q, while the median for taking all Q-value is defined as base value θ, if Q is ＞ θ, then it is assumed that should
Group set of parameter values is acceptable, and this group of set of parameter values is distributed in acceptable group；Otherwise it is assumed that be it is unacceptable,
This group of set of parameter values is distributed in unacceptable group, so as to complete mtThe distribution of group set of parameter values：
Q=| fumin-fmin|+|fus-fs| formula (1)；
2-e) compare the distribution situation of each parameter, for arbitrary parameter, define its j-th independent random number be distributed in it is acceptable
The number of group is pj, the number for being distributed in unacceptable group is qj, while sensitive parameter δ is defined, ifThen
For the parameter, acceptable group is similar with unacceptable component cloth, then show that the parameter is insensitive；Conversely, then showing this
Parameter sensitivity；If the sum of whole sensitive parameters is calculated as n, then n dimension parameter space X=(x to be identified are constituted1,x2,…,xn)。
3. the mains frequency simulation parameter discrimination method based on simplex method according to claim 1, it is characterised in that：Institute
State step 3) in other n groups parameter group is generated based on the initial value parameter group, specifically include following steps：N is tieed up into parameter space X
=(x1,x2,…,xn) in each Parameter Typical as one group of initial value parameter group, and by each parameter allusion quotation in the initial value parameter group
Offset deviates the farthest measured value of the Parameter Typical in replacing with the corresponding parameter of the Parameter Typical successively, individually replace every time
A parameter is changed, so as to form other n groups parameter group.
4. the mains frequency simulation parameter discrimination method based on simplex method according to claim 1, it is characterised in that：Institute
State step 4) according to formula (2) calculating target function value J：
Wherein, α is sampling number, yiFor i-th point value of frequency response results curve Y, aiFor PMU frequency responses after failure
I-th point value of curve a；
If maximum target functional value is JH, minimum target functional value is JL, draw JHParameter group be XH, draw JLParameter group set
For XL, concurrently set convergenceIfThen stop identification, display parameters are recognized successfully, parameter group XLFor most
Excellent parameter group.
5. the mains frequency simulation parameter discrimination method based on simplex method according to claim 1, it is characterised in that：Institute
State step 5) in generate new one group of parameter group using simplex method, specifically include following steps：By maximum target functional value JHIt is right
Parameter group X answeredHReflection, obtains reflection parameters group XR, and calculate reflection parameters group XRCorresponding target function value JR, pass through
Generate compression parameters group or expansion parameters group carrys out alternative parameter group XH, new simplex is formed with other specification group.
6. the mains frequency simulation parameter discrimination method based on pure algorithm according to claim 1 and 2, its feature exist
In：In the stable file, the unit related to frequency emulation includes prime mover, speed regulator and boiler main steam pressure model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610889021.2A CN106599337B (en) | 2016-10-12 | 2016-10-12 | Power grid frequency simulation parameter identification method based on simplex method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610889021.2A CN106599337B (en) | 2016-10-12 | 2016-10-12 | Power grid frequency simulation parameter identification method based on simplex method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106599337A true CN106599337A (en) | 2017-04-26 |
| CN106599337B CN106599337B (en) | 2020-04-14 |
Family
ID=58555857
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610889021.2A Active CN106599337B (en) | 2016-10-12 | 2016-10-12 | Power grid frequency simulation parameter identification method based on simplex method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106599337B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112464437A (en) * | 2020-10-30 | 2021-03-09 | 海南电网有限责任公司电力科学研究院 | Parameter identification method of electric power simulation model |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6437325B1 (en) * | 1999-05-18 | 2002-08-20 | Advanced Research And Technology Institute, Inc. | System and method for calibrating time-of-flight mass spectra |
| CN101615217A (en) * | 2008-06-27 | 2009-12-30 | 富士通株式会社 | Different design shape to similar characteristics is classified/device shown and method |
| CN102146812A (en) * | 2010-02-09 | 2011-08-10 | 浙江省电力公司 | Actual-measurement modeling method for prime mover and speed governor thereof of electric power system |
| CN102799785A (en) * | 2012-07-19 | 2012-11-28 | 广东电网公司电力科学研究院 | Prime motor based on simplex method of nuclear power unit and method for identifying parameters of speed regulator of prime motor |
-
2016
- 2016-10-12 CN CN201610889021.2A patent/CN106599337B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6437325B1 (en) * | 1999-05-18 | 2002-08-20 | Advanced Research And Technology Institute, Inc. | System and method for calibrating time-of-flight mass spectra |
| CN101615217A (en) * | 2008-06-27 | 2009-12-30 | 富士通株式会社 | Different design shape to similar characteristics is classified/device shown and method |
| CN102146812A (en) * | 2010-02-09 | 2011-08-10 | 浙江省电力公司 | Actual-measurement modeling method for prime mover and speed governor thereof of electric power system |
| CN102799785A (en) * | 2012-07-19 | 2012-11-28 | 广东电网公司电力科学研究院 | Prime motor based on simplex method of nuclear power unit and method for identifying parameters of speed regulator of prime motor |
Non-Patent Citations (2)
| Title |
|---|
| 武诚 等: "《基于响应特性的原动机及调速器建模与参数辨识》", 《电工技术学报》 * |
| 黄杰 等: "《面向风电场SSO抑制的TCSC参数电磁暂态智能优化方法》", 《电网技术》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112464437A (en) * | 2020-10-30 | 2021-03-09 | 海南电网有限责任公司电力科学研究院 | Parameter identification method of electric power simulation model |
| CN112464437B (en) * | 2020-10-30 | 2022-11-08 | 海南电网有限责任公司电力科学研究院 | Parameter identification method of electric power simulation model |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106599337B (en) | 2020-04-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106655159B (en) | New energy power station primary frequency modulation capability test system and test method thereof | |
| WO2016026355A1 (en) | Voltage sag simulation and evaluation method of active power distribution grid | |
| CN103887815B (en) | Based on wind energy turbine set parameter identification and the Dynamic Equivalence of service data | |
| CN103944507B (en) | Photovoltaic-power-station low-voltage penetrating performance evaluation method based on inverter model test | |
| CN103973203B (en) | A kind of online equivalent modeling method in large-sized photovoltaic power station suitable in security and stability analysis | |
| CN104599189A (en) | Power grid planning scheme risk evaluation method considering power system operation mode | |
| Fang et al. | Application of gray relational analysis to k-means clustering for dynamic equivalent modeling of wind farm | |
| CN103246806A (en) | Operation risk evaluation method comprising wind- power plant electric system | |
| CN103198184B (en) | A kind of low-frequency oscillation character noise-like identification method in electric power system | |
| WO2021035681A1 (en) | Lagrange multiplier-based optimal load curtailment algorithm and application thereof | |
| CN106786608B (en) | A kind of uncertain harmonic flow calculation method suitable for distributed generation resource access | |
| CN104156889A (en) | Wind farm performance evaluation system based on WAMS data and evaluation method thereof | |
| CN104616061A (en) | Island detection method based on wavelet packet logarithmic energy entropy and genetic algorithm optimization | |
| CN104638671A (en) | Island detection method based on introducing particle swarm optimization and phase disturbance | |
| CN109428327A (en) | Power grid key branch and leading stable mode recognition methods and system based on response | |
| CN103078343B (en) | Evaluation method for impact of wind power integration on power grid transient state as well as medium and long term reliability | |
| CN105956760B (en) | Intelligent power distribution network situation perception method based on multivariate time-space information modeling | |
| CN105701265B (en) | A kind of double-fed fan motor unit modeling method and device | |
| CN106056305A (en) | Power generation system reliability rapid assessment method based on state clustering | |
| CN106599337A (en) | Power grid frequency simulation parameter identification method based on simplex method | |
| CN106329515B (en) | A kind of horizontal determination method of electric network reliability based on static reliability probability level | |
| CN102957166B (en) | Method for quickly calculating wind-power allocation ratio based on trajectory sensitivity | |
| CN114139812A (en) | Monitoring device optimal configuration method for subsynchronous oscillation monitoring | |
| CN106253328A (en) | A kind of analysis method that wind farm grid-connected point is carried out reasonable plant-site selection | |
| CN106100003A (en) | The voltage security assessments method and system of power system |
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 |