CN103632046A - Power grid load flow calculation method - Google Patents
Power grid load flow calculation method Download PDFInfo
- Publication number
- CN103632046A CN103632046A CN201310584848.9A CN201310584848A CN103632046A CN 103632046 A CN103632046 A CN 103632046A CN 201310584848 A CN201310584848 A CN 201310584848A CN 103632046 A CN103632046 A CN 103632046A
- Authority
- CN
- China
- Prior art keywords
- node
- matrix
- admittance
- network
- voltage
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Abstract
The invention discloses a power grid load flow calculation method. The power grid load flow calculation method is based on establishment of a traditional nodal equation and an analytic theory, node power mapping is correlated with node earth impedance, a power grid load flow calculation model of three-dimensional network topology is formed by a traditional two-dimensional plane, node earth branches and an earth zero plane, the nodal equation and an admittance matrix are established according to a network topological structure chart, a common load flow calculation proposition is resolved into a linear part and a non-linear part, and then a pure algebraic operation is conducted. The power grid load flow calculation model has the novel advantages of being simple in principle, clear in process, efficient, accurate and the like. The power grid load flow calculation method has the advantages that the calculation speed is high, the required internal storage is small, a calculation result is good in reliability and credibility, adjustment of the ratio of transformation of a transformer can be processed, and the capacity of change of a system operation mode and the capacity of the cooperation of a system with other programs are strong.
Description
Technical field
The present invention relates to Regulation monitoring technical field, relate in particular to a kind of electric network swim computing method.
Background technology
It is a kind of substantially electric calculating of research power system mesomeric state ruuning situation that trend is calculated, the task that conventional Load Flow calculates is according to given service condition and network configuration, to determine the running status of whole system, as the power in the voltage on each bus (amplitude and phase angle), network distributes and power attenuation etc.The result that trend is calculated is the basis of Model for Stability Calculation of Power System and fault analysis.Therefore, to calculate be most widely used general, the most important a kind of electric computing of fundamental sum in electric system to trend.
The method that electric network swim calculates at present mainly be take Newton-Raphson calculating power system load flow as theoretical core, from the derivative the whole bag of tricks of this core system, be all to utilize the idea of pure mathematics to carry out the processing of trend calculating, actual electric network entity has been carried out to the mathematical abstractions of height, what set up is two dimensional surface network topology, use higher mathematics computing to resolve, algorithm principle is complicated and the realization of computer program structure has also been brought to great difficulty.
Summary of the invention
The object of this invention is to provide that a kind of principle is simple, process is clear, efficient accurately, be convenient to the electric network swim computing method of computer programing.
For realizing the technical scheme that object of the present invention adopts, be: a kind of electric network swim computing method are to carry out according to the following steps:
A. set up electric network swim computation model admittance matrix
By the grid nodes planar network that form associated with grid branch, to pay close attention to the intrinsic relation of node power and impedance ground, with impedance ground, planar network and basic unit's 0 plane are associated to formation cubic network, utilize network element parameter and associated intrinsic property, build electric network swim computation model admittance matrix, the storage of electric network swim computation model admittance matrix is defined as follows:
Y
11 Y
12 Y
13 …… Y
1n U
1 Y
10 Y
h1
Y
21 Y
22 Y
23 …… Y
2n U
2 Y
20 Y
h2
Y
31 Y
32 Y
33 …… Y
3n = U
3 Y
30 Y
h3
…… ……
Y
n1 Y
n2 Y
n3 …… Y
nn U
n Y
n0 Y
hn
Y in formula
11-Y
nnfor the element in bus admittance matrix; U
1-U
nfor node voltage row; Y
10-Y
n0the row of admittance over the ground for node; Y
h1-Y
hnfor the capable transadmittance of node;
In described electric network swim computation model admittance matrix, according to electrical network calculate trend to need balance node to be set to one or more, node self-admittance is synthetic by admittance over the ground and the transadmittance of each row, node voltage is listed as that all the other all insert null value except balance node is inserted magnitude of voltage, main transformer element adopts non-standard no-load voltage ratio model, and every three circle main transformer takies an isolated node, wherein, matrix element is all plural number, from the parsing that has been established to of matrix, all adopts complex operation;
B. the original state of network and linear parsing
Each node of electric network swim computation model is with default impedance access network nodes, impedance Z is got the admittance over the ground that becomes node after inverse, in admittance matrix, the self-admittance value of each diagonal element is synthetic by the transadmittance of this node and admittance over the ground, balance node voltage is write to matrix, adopt elimination method to resolve matrix, first by the row format of advancing at balance node place, method is that node self-admittance is made as to 1, other elements set to 0, recycling Gaussian elimination makes matrix become upper triangular matrix, then utilize elimination method to make matrix become unit matrix, in twice cancellation process, all use cancellation factor pair node voltage row to process, the voltage row that finally obtain are each node voltage value U of this computing
1, U
2... U
i(i=1～n),
C. using iterative method completes the nonlinear computation of trend
Trend calculating proposition is generally that Ps, the Qs value of setting balance node U value and each load bus is condition, asks each node voltage U
1, U
2... U
n, after the linear operation of network original state, obtain U
1 0, U
2 0... U
i 0(i=1～n), and set Ps, the Qs value of each load bus, enter following iterative process:
Iteration 2. is by the U of each load bus
1, Ps, Qs calculate the Y of admittance over the ground of each load bus successively
1 2, Y
2 2... Yi
2(i=1～n), by each node Yi
2substitution bus admittance matrix, carries out linear function parsing, obtains each node voltage U
1 2, U
2 2... Ui
2(i=1～n), the new value of the P of calculated load node, Q output, carries out ratio of precision with Ps, the Qs value set, determines whether to continue iteration, if desired enters iteration 3;
By that analogy, continue iteration until meet predetermined computational accuracy.
The method that electric network swim calculates at present mainly be take Newton-Raphson calculating power system load flow as theoretical core, from the derivative the whole bag of tricks of this core system, be all to utilize the idea of pure mathematics to carry out the processing of trend calculating, actual electric network entity has been carried out to the mathematical abstractions of height, conventionally set up two dimensional surface network topology, use higher mathematics computing to resolve.This computational algorithm principle is complicated, and the realization of computer program structure has also been brought to great difficulty.The present invention utilizes the concept of classical alternating circuit to build a kind of novel power flow algorithm, its modeling approach is: take the foundation of conventional node equation and analytic theory is basis, node power is hinted obliquely at and is associated with node earth impedance, and by conventional two-dimensional plane, node, branch road and the earth 0 plane form 3 D stereo network topology over the ground.Then, according to network topology structure, figure sets up modal equation, the scale of admittance matrix is dynamically adjusted according to topological diagram, and general trend is calculated to proposition be decomposed into two parts and carry out pure algebraic operation processing, linear segment and non-linear partial, the parsing of modal equation belongs to linear calculating section, and the iterative computation of each node load belongs to non-linear partial.Like this, utilize network element parameter and associated intrinsic property, build rational solution, trend is calculated and there is the characteristics such as principle is simple, process is clear, efficient accurate.Trend is calculated the general type of proposition: known network bus admittance matrix, a given node voltage (balance node) and each node turnover power S
1,s
2,s
ithe setting value of (i=1～n), asks each node voltage U
1,u
2,u
i(i=1～n).Be provided with n the operating network of node, its main operational factor has: U
0 1,u
0 2,u
0 i(i=1～n) is each node working voltage, S
0 1,s
0 2,s
0 i(i=1～n) is for each node injection or flow out power, Z
0 1,z
0 2,z
0 i(i=1～n) is each node earth impedance, and each node is each parameter Ui at any time, Ii, and Si, Zi meets the constraint of physical law.
After circuit original state is determined, for each node i, there is U
0 i, S
0 i, Z
0 ideng one group of operational factor of determining numerical relation, calculate target for equaling setting power S when node i operate power
itime corresponding node voltage U
i, parsing thinking is: compare initial power S
0 iwith setting power S
iif, S
0 i<S
i, node initial power need increase to S
iclose, use U
0 iand S
ican calculate the feedback impedance Z that makes node the increase of output power
1 i(feedback impedance is exactly the earth impedance of node, in circulation process of analysis, plays retroactive effect), with initial impedance Z
0 icompare Z
1 ithere is the trend of reducing; If S
0 i>S
i, node initial power need reduce to S
iclose, use U
0 iand S
ican calculate the feedback impedance Z that makes node power reduction
1 i, with initial impedance Z
0 icompare Z
1 ithere is increase tendency.Therefore, no matter S
0 iwith S
irelation how, can utilize U
0 iwith S
icalculate one and make initial power S
0 ito setting power S
iapproach the feedback impedance Z of variation
1 i.All nodes in network all have the characteristic that the adjusting by feedback impedance makes node operate power approach to setting power, from original state, utilize U
0 iand S
icalculate the feedback impedance Z of each node
1 1,z
1 2,z
1 i(i=1～n), substitution network equation directly solves new round node voltage U
1 1,u
1 2,u
1 i(i=1～n), now each node operate power S
1 1,s
1 2,s
1 i(i=1～n) is all to setting power S
1,s
2,s
i(i=1～n) convergence, then uses new round voltage U
1 iand S
isolve the feedback impedance Z of each node
2 i, substitution network equation directly solves each node voltage U
2 i, each node power convergence S more now
i, above-mentioned computing loops after m time, if node power S
m iwith S
idifference be less than the target error ε of setting, can think that computational accuracy meets the demands, each node voltage U
m 1,u
m 2,u
m i(i=1～n) is the trend calculating target that NDN model solution is separated out.A no matter how complicated electrical network, is undertaken after abstract representation by primary connection figure, has just formed a complicated alternating circuit, and from the viewpoint of alternating circuit, in arbitrary moment network, the operational factor of arbitrary node is all strictly obeyed the constraint of modal equation.Therefore, electric network swim computing method of the present invention have that principle is simple, process is clear, efficient accurate characteristic.
Electric network swim computing method of the present invention, take the foundation of conventional node equation and analytic theory is basis, node power is hinted obliquely at and is associated with node earth impedance, by conventional two-dimensional plane, node, branch road and the earth 0 plane form the electric network swim computation model of 3 D stereo network topology over the ground, according to network topology structure, figure sets up modal equation, the scale of admittance matrix is dynamically adjusted according to topological diagram, general trend is calculated to proposition and be decomposed into linear and non-linear two parts and carry out pure algebraic operation processing.So computation model of the present invention has the new characteristics such as principle is simple, process is clear, efficient accurate.Electric network swim computing method of the present invention also have that computing velocity is fast, internal memory needs little, result of calculation to have good reliability and credibility, and can process transformer voltage ratio adjustment, and the variation of system operation mode, the ability coordinating with other programs are strong.Electric network swim computation model modeling approach of the present invention is distinct, and convergence, precision, speed all show very well in contrast test.
Accompanying drawing explanation
Below in conjunction with accompanying drawing, the present invention is described further:
Fig. 1 is the block diagram of electric network swim computing method flow process in the present invention;
Fig. 2 is the 3 D stereo network chart that in the present invention, electric network swim calculates;
Fig. 3 is the process flow diagram of iterative computation in the present invention;
Fig. 4 is pure resistor element circuit trend calculation diagram in the present invention.
Embodiment
As shown in Figure 1, Figure 2, Figure 3 shows, electric network swim computing method of the present invention comprise set up the original state of electric network swim computation model admittance matrix 1, network and linear resolve 2,3 three steps of nonlinear computation that using iterative method completes trend, concrete grammar respectively:
A. set up electric network swim computation model admittance matrix 1
By node 4 planar network that form 5 associated with branch road, to pay close attention to the intrinsic relation of node power and impedance ground, with impedance ground 6, planar network 5 and basic unit's 0 plane 7 are associated to formation cubic network, utilize network element parameter and associated intrinsic property, build electric network swim computation model admittance matrix.In the forming process of admittance matrix, main transformer element adopts non-standard no-load voltage ratio model, and every three circle main transformer takies an isolated node.Electric network swim computation model admittance matrix storage of the present invention is defined as follows:
Y
11 Y
12 Y
13 …… Y
1n U
1 Y
10 Y
h1
Y
21 Y
22 Y
23 …… Y
2n U
2 Y
20 Y
h2
Y
31 Y
32 Y
33 …… Y
3n = U
3 Y
30 Y
h3
…… ……
Y
n1 Y
n2 Y
n3 …… Y
nn U
n Y
n0 Y
hn
Y in formula
11-Y
nnfor the element in the bus admittance matrix of classic method formation, U
1-U
nfor node voltage row, Y
10-Y
n0for the row of admittance over the ground of node 4, Y
h1-Y
hntransadmittance for node 4 row.In described electric network swim computation model admittance matrix, according to electrical network calculate trend to need balance node to be set to one or more, node self-admittance (diagonal element of traditional admittance matrix) is synthetic by admittance over the ground and the transadmittance of each row, node voltage is listed as that all the other all insert null value except balance node is inserted magnitude of voltage, main transformer element adopts non-standard no-load voltage ratio model, every three circle main transformer takies an isolated node, for coupled computer algorithm more easily, matrix element in electric network swim computation model is all plural number, from the parsing that has been established to of matrix, all adopt complex operation.
B. the original state of network and linear parsing 2
Each node of electric network swim computation model with default (initially) impedance (such as Z=1000+j500) access network nodes, impedance Z is got the admittance over the ground that becomes node after inverse, in admittance matrix, the self-admittance value of each diagonal element is synthetic by the transadmittance of this node and admittance over the ground, when the structural parameters of network are constant, the transadmittance value of diagonal element node remains constant, admittance changes with node load over the ground, balance node voltage is write to matrix, adopt improved elimination method to resolve matrix, first by the row format of advancing at balance node place, method is that node self-admittance is made as to 1, other elements set to 0, recycling Gaussian elimination makes matrix become upper triangular matrix, then utilize improvement elimination method to make matrix become unit matrix, in twice cancellation process, all to process with cancellation factor pair node voltage row, the voltage row that finally obtain are each node voltage value U of this computing
1, U
2... U
i(i=1～n).This process is exactly the linear calculating section in NDN power flow algorithm.
C. using iterative method completes the nonlinear computation 3 of trend
Trend calculating proposition is generally that Ps, the Qs value of setting balance node U value and each load bus is condition, asks each node voltage U
1, U
2... U
n.After the linear operation of network original state, obtain U
1 0, U
2 0... U
i 0(i=1～n), and the Ps of known each load bus, Qs setting value, enter following iterative process:
Iteration 2. is by the U of each load bus
1, Ps, Qs calculate the Y of admittance over the ground of each load bus successively
1 2, Y
2 2... Yi
2(i=1～n), by each node Yi
2substitution bus admittance matrix, carries out linear function parsing, obtains each node voltage U
1 2, U
2 2... Ui
2(i=1～n), the new value of the P of calculated load node, Q output, Ps, Qs carry out ratio of precision with setting, determine whether to continue iteration, if desired enter iteration 3.
By that analogy, iteration is until meet predetermined computational accuracy.Computational practice shows, generally through 5-6 iteration, numerical precision can reach the requirement of engineering calculation, while carrying out continuous in-circuit emulation under SCADA system coordinates, because system before each computing has good initial condition, so computational accuracy shows very remarkably, speed is also unrivaled.
Referring to Fig. 4, by a simple pure resistor element circuit trend computation process, electric network swim computing method seat of the present invention is further illustrated below:
Initial circuit is calculated as:
I＝110/（10＋100）＝1A
U
0＝1×100＝100V
P
f＝1×100＝100W
Above circuit parameter explanation: line resistance 9 R=10 Ω, load resistance 10 R=100 Ω, supply voltage 8 U=110V, solve load setting power P
fsduring=200W, current voltage 11 U
0deng main circuit operational factor.
Resolve thinking: use current voltage 11(U
0) and load setting power P
fscalculated load resistance 10(R), by load resistance 10(R) be newly worth substitution circuit and recalculate main operational factor current voltage 11(U
0) new value and P
fnew value.Compare P
fnew value and P
fserror whether meet the demands, otherwise continue to use current voltage 11(U
0) new value and P
fsnew value calculated load R, the calculating that iterates of substitution circuit, until P
fnew value meets accuracy requirement.Concrete iterative process is described below:
Iteration 1:U
0=100 V P
fs=200 W solve: R=50 Ω I=1.8333 A
| Iterations | U 1New value | P
fNew | Error % | |
| 1 | 91.6667 | 168.0525 | 15.97 |
Iteration 2:U
1=91.6667 V P
fs=200 W solve: R=42.0139 Ω I=2.1148 A
| Iterations | U 2New value | P fNew value | Error % |
| 2 | 88.8509 | 187.9018 | 6.04 |
Iteration 3:U
2=88.8509 V P
fs=200 W solve: R=39.4724 Ω I=2.2234 A
| Iterations | U 3New value | P
fNew | Error % | |
| 3 | 87.7629 | 195.132 | 2.43 |
Iteration 4:U
3=87.7629 V P
fs=200 W solve: R=38.5116 Ω I=2.2675 A
| Iterations | U 4New value | P fNew value | Error % |
| 4 | 87.325 | 198.0094 | 0.99 |
Iteration 5:U
4=87.325 V P
fs=200 W solve: R=38.1282 Ω I=2.2855 A
| Iterations | U 5New value | P fNew value | Error % |
| 5 | 87.142 | 199.163 | 0.41 |
Iteration 6:U
5=87.142 V P
fs=200 W solve: R=37.9686 Ω I=2.2931 A
| Iterations | U 6New value | P
fNew | Error % | |
| 6 | 87.0658 | 199.6506 | 0.17 |
Iteration 7:U
6=87.0658 V P
fs=200 W solve: R=37.9022 Ω I=2.2963 A
| Iterations | U 7New value | P
fNew | Error % | |
| 7 | 87.0348 | 199.858 | 0.07 |
Iteration 8:U
7=87.0348 V P
fs=200 W solve: R=37.8752 Ω I=2.2976 A
| Iterations | U 8New value | P fNew value | Error % |
| 8 | 87.022 | 199.9417 | 0.029 |
Computational practice shows can reach 0.5% by five iteration precisions, can meet the requirement of engineering calculation, and after ten interative computations, result precision can be up to 0.003%.
Electric network swim computation model of the present invention has following technical characteristic:
1. the convergence property of electric network swim computation model of the present invention: from tandem circuit theory, as load impedance Z
fequal hitching position network internal impedance Z
wtime, load can obtain peak power S
max, as long as the setting value of node power is less than S
max, whole calculating can restrain.That is to say while only having node power to be provided with the performance number that real network can not provide, the phenomenon that whole calculating just can occur not restrain, Software for Design can be found this phenomenon in time, and is easy to retrieve the Node configuration that causes that calculating does not restrain.The convergence property of NDN depends on practical structures and the service ability of network, is weak relevant to mathematical method.
2. in consolidated network, a plurality of balance node can be set: electric network swim computation model of the present invention is based on complicated alternating circuit tectonic network model, it by electrical network primary connection figure height abstract be equivalent alternating circuit, can add voltage source at arbitrary node, and power parameter (amplitude, initial phase angle) is adjustable, this specific character is calculated highly beneficial to the network simulation that has the trend simulation calculation of two above separate networks in region or contain isolated island.
3. Node configuration parameter: each node load is designed with impedance and power two classes arrange input interface, can select respectively, when node impedance is made as " 0 ", can carry out easily node short circuit calculation, and be high-speed, high-precision linear operation, a kind of model can carry out the various calculating in network stable state field.
4. the resolving of electric network swim computation model modal equation of the present invention and the structure of admittance matrix, it doesn't matter for element symmetry, adds the symmetric element of influence matrix in network, during as phase shifter or FACTS element, on not impact of the parsing of equation.This specific character adds the network operation that simulation capacity is provided to constructing the element of equivalent electrical circuit.Compare and have higher advantage with some algorithm that depends on admittance matrix structure and parameter symmetry in traditional power flow algorithm.
5. in the admittance matrix of electric network swim computation model of the present invention, except node over the ground admittance directly associated with node load, other structures and element are corresponding with the intrinsic parameter of network, when network parameter changes, can directly revise online element parameter.While having FACTS element in network, this model can provide trend solution more easily.
6. electric network swim computation model of the present invention provides the linear parameter of U, Z at node interface place, any element with the linear parameter interface of U, Z can directly pass through node access network, for nonlinear element, can pass through the method access network of node local iteration, diversified electrical equipment can be received, be contained to computation model of the present invention.
Electric power system tide calculates the method adopting and mainly contains following several: Gauss-Seidel method, impedance method, Newton-Raphson method and PQ decomposition method, the method extensively adopting is at present Newton method and PQ decomposition method, Newton method is the classical way that solves Nonlinear System of Equations in mathematics, there is good convergence, after sparse matrix analytic technique is successfully used, Newton method becomes electric power system tide and calculates basic, outstanding method.PQ decomposition method is on the basis of Newton method, in conjunction with the concrete feature of electric system, the Newton method of pure mathematics is improved, because the obvious raising of speed is applied rapidly.The trend Computing Principle of electric network swim computation model of the present invention is different from Newton method completely, uses the computing experience of alternating circuit, and electric power networks is the subset of alternating circuit, and computation model of the present invention is more suitable for the calculating of electric system.Specifically, computation model of the present invention is compared and is mainly contained following characteristics with the ripe computing method of tradition:
1. the saving of calculator memory, electric network swim computation model of the present invention is worked completely on the basis of admittance matrix, do not need to set up the update equation group of the huge Jacobian matrix of Newton method and PQ decomposition method, it uses improved admittance logic matrix and process tree technology, make admittance matrix boil down to n * Lmax matrix of former n * n, wherein Lmax be logic matrix capable in non-zero entry prime number and the maximal value of injecting number of elements sum, through research, find, the node scale of system is larger, more meets Lmax<<n.Process tree technology is a kind of new method that solves sparse matrix problem, in conjunction with logic matrix technology, Computer Storage can be carried out to height optimization, and this has prepared calculator memory technical support for large-scale with trend solution very large network.
2. computational convergence, precision and speed.Electric network swim computation model of the present invention does not need to arrange the voltage initial value of each node after being provided with balance node voltage, the convergence of electric network swim computation model of the present invention depends on the actual structure parameters of network and the settings of node power, so long as the running status that network can reach, electric network swim computation model of the present invention can be restrained under that state.And Newton method need to arrange the initial value of node voltage, disadvantageous initial value setting may cause computing not restrain or converge to the value that practical power systems can not be moved, PQ decomposition method is that Newton method is for the simplification of electric system feature, by test, show, when getting in touch with impedance between node while being obviously perception, PQ decomposition method can restrain, and is difficult for convergence when contact impedance is obviously resistive.In electric system, the above electrical network of 110kV can meet the condition of the contact impedance R<<X of PQ decomposition method requirement, the following electrical network of 110kV not necessarily meets this condition, calculating need to increase iterations or can not restrain at all, and the scope of application of PQ decomposition method is restricted.Newton method is a kind of trend solution of pure mathematics, and theory is very ripe, but before correct trend result occurs, needs the initial value of Empirical rules node, and departing from of initial value can cause the discrete of whole trend calculating.Computation model of the present invention is taken root in basic alternating-current circuit theory, and electric system is a subset of electric network swim computation model solution of the present invention.By evidence: the scheme Newton method that computation model of the present invention can the be restrained convergence surely that differs, and the problem that Newton method can restrain computation model one of the present invention is restrained surely.The convergence of computation model of the present invention depends on the actual motion ability of network, and computation model of the present invention is a kind of physics solution of trend computational problem.The precision of NDN can match in excellence or beauty with Newton method under identical iterations, generally carries out 5-6 iteration and can obtain satisfied result.The computing velocity of computation model of the present invention: its iterative resolution process is carried out on the basis of original admittance matrix, formation and the resolving of the update equation matrix of Newton method or PQ decomposition method have been crossed over, the physical meaning of computation model computation process of the present invention is clear, only use simple mathematics arithmetic, computing velocity can apply to real time environment, in the theoretical foundation of computation model of the present invention, can work out more efficient algorithm completely.
3. the versatility of model: electric network swim computation model of the present invention is from common alternating-current circuit theory, except solving similar electric power system tide computational problem, the concept of dealing with problems can be amplified the analysis of various alternating circuits, the essence of computation model of the present invention is the expert of an alternating circuit linear analysis, by iteration, solve the nonlinear problem of calculating for electric power system tide, Newton method and PQ decomposition method are born mainly for solving nonlinear problem, can not be directly with solving linear problem.NDN is a kind of comprehensive expert model that solves linear and nonlinear problem.After the model system of a network is set up, use same model can carry out conventional trend computational analysis, N-1 verification is calculated, network short calculating etc.Computation model of the present invention is a kind of intensive model, by further in-depth research, is likely evolved into basic model.
4. Newton method requires can only there be a balance node in a network, and when network occurs the isolated island of independent operating or the localized network of power down completely due to the variation of the method for operation, not having special processing to calculate cannot be by expection operation.The balance node quantity that computation model of the present invention can arrange is unrestricted, can a plurality of independently network of emulation in a model, for the localized network of complete power down, it can be crossed over automatically as white area.Computation model of the present invention is the sharp weapon of research power system operating mode, and it can simulate all methods of operation.Computation model of the present invention is applied to Voltage Stability Analysis, system optimization to be calculated and can obviously simplify and accelerate trend calculating link, the critical value of its calculating more approaches the actual motion ability of electrical network, for the stability analysis of electric system, may bring larger facility.
5. computation model of the present invention can be realized 0 voltage and starts to start electrical network, its network is a Utopian alternating circuit, do not use the concept completion system of electrical network rated voltage to calculate, certainly higher voltage also can be added on this desirable network, and the grid simulation that computation model of the present invention does not need special processing just can naturally complete roomy scope input voltage calculates.
Claims (1)
1. electric network swim computing method, is characterized in that: it is to carry out according to the following steps:
A. set up electric network swim computation model admittance matrix
By the grid nodes planar network that form associated with grid branch, to pay close attention to the intrinsic relation of node power and impedance ground, with impedance ground, planar network and basic unit's 0 plane are associated to formation cubic network, utilize network element parameter and associated intrinsic property, build electric network swim computation model admittance matrix, the storage of electric network swim computation model admittance matrix is defined as follows:
Y
11 Y
12 Y
13 …… Y
1n U
1 Y
10 Y
h1
Y
21 Y
22 Y
23 …… Y
2n U
2 Y
20 Y
h2
Y
31 Y
32 Y
33 …… Y
3n = U
3 Y
30 Y
h3
…… ……
Y
n1 Y
n2 Y
n3 …… Y
nn U
n Y
n0 Y
hn
Y in formula
11-Y
nnfor the element in bus admittance matrix; U
1-U
nfor node voltage row; Y
10-Y
n0the row of admittance over the ground for node; Y
h1-Y
hnfor the capable transadmittance of node;
In described electric network swim computation model admittance matrix, according to electrical network calculate trend to need balance node to be set to one or more, node self-admittance is synthetic by admittance over the ground and the transadmittance of each row, node voltage is listed as that all the other all insert null value except balance node is inserted magnitude of voltage, main transformer element adopts non-standard no-load voltage ratio model, and every three circle main transformer takies an isolated node, wherein, matrix element is all plural number, from the parsing that has been established to of matrix, all adopts complex operation;
B. the original state of network and linear parsing
Each node of electric network swim computation model is with default impedance access network nodes, impedance Z is got the admittance over the ground that becomes node after inverse, in admittance matrix, the self-admittance value of each diagonal element is synthetic by the transadmittance of this node and admittance over the ground, balance node voltage is write to matrix, adopt elimination method to resolve matrix, first by the row format of advancing at balance node place, method is that node self-admittance is made as to 1, other elements set to 0, recycling Gaussian elimination makes matrix become upper triangular matrix, then utilize elimination method to make matrix become unit matrix, in twice cancellation process, all use cancellation factor pair node voltage row to process, the voltage row that finally obtain are each node voltage value U of this computing
1, U
2... U
i(i=1～n),
C. using iterative method completes the nonlinear computation of trend
Trend calculating proposition is generally that Ps, the Qs value of setting balance node U value and each load bus is condition, asks each node voltage U
1, U
2... U
n, after the linear operation of network original state, obtain U
1 0, U
2 0... U
i 0(i=1～n), and set Ps, the Qs value of each load bus, enter following iterative process:
Iteration 1. is by the U of each load bus
0, Ps, Qs calculate the Y of admittance over the ground of each load bus successively
1 1, Y
2 1... Y
i 1(i=1～n), by each node Y
i 1substitution bus admittance matrix, carries out linear function parsing, obtains each node voltage U
1 1, U
2 1... Ui
1(i=1～n), the new value of the P of calculated load node, Q output, carries out ratio of precision with Ps, the Qs value set, determines whether to continue iteration, if desired enters iteration 2;
Iteration 2. is by the U of each load bus
1, Ps, Qs calculate the Y of admittance over the ground of each load bus successively
1 2, Y
2 2... Yi
2(i=1～n), by each node Yi
2substitution bus admittance matrix, carries out linear function parsing, obtains each node voltage U
1 2, U
2 2... Ui
2(i=1～n), the new value of the P of calculated load node, Q output, carries out ratio of precision with Ps, the Qs value set, determines whether to continue iteration, if desired enters iteration 3;
By that analogy, continue iteration until meet predetermined computational accuracy.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310584848.9A CN103632046A (en) | 2013-11-20 | 2013-11-20 | Power grid load flow calculation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310584848.9A CN103632046A (en) | 2013-11-20 | 2013-11-20 | Power grid load flow calculation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103632046A true CN103632046A (en) | 2014-03-12 |
Family
ID=50213085
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310584848.9A Pending CN103632046A (en) | 2013-11-20 | 2013-11-20 | Power grid load flow calculation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103632046A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105787604A (en) * | 2016-03-22 | 2016-07-20 | 山东大学 | Transmission-distribution-cooperative distribution network situation quick awareness method |
| CN107359616A (en) * | 2017-07-18 | 2017-11-17 | 国电南瑞科技股份有限公司 | One kind solves the not convergent method of Load flow calculation after large scale electric network mode adjusts |
| CN107957974A (en) * | 2017-11-29 | 2018-04-24 | 广西大学 | A kind of method for extracting polar coordinates power flow equation Jacobian matrix |
| CN109344361A (en) * | 2018-08-27 | 2019-02-15 | 南昌大学 | A kind of electric power system tide calculate in the method that is quickly formed of Jacobian matrix |
| CN110046450A (en) * | 2019-04-24 | 2019-07-23 | 深圳供电局有限公司 | A kind of Initialization Algorithms that the Newton Power Flow being suitble to the power grid containing hyperconductive cable calculates |
| CN110460059A (en) * | 2019-07-26 | 2019-11-15 | 国网山东省电力公司泰安供电公司 | A kind of electric power system tide calculation method |
| CN111724014A (en) * | 2019-03-20 | 2020-09-29 | 哈尔滨工业大学（威海） | Weighted-SALSA algorithm-based power system node importance calculation method |
| CN113364007A (en) * | 2021-06-07 | 2021-09-07 | 广东电网有限责任公司 | Two-dimensional admittance model construction method and device, terminal and readable storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2677384A1 (en) * | 2009-09-01 | 2011-03-01 | Oliver Romaniuk | Apparatus, methods and systems for parallel power flow calculation and power system simulation |
| CN102013680A (en) * | 2010-12-13 | 2011-04-13 | 大连海事大学 | Fast decoupled flow calculation method for power systems |
-
2013
- 2013-11-20 CN CN201310584848.9A patent/CN103632046A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2677384A1 (en) * | 2009-09-01 | 2011-03-01 | Oliver Romaniuk | Apparatus, methods and systems for parallel power flow calculation and power system simulation |
| CN102013680A (en) * | 2010-12-13 | 2011-04-13 | 大连海事大学 | Fast decoupled flow calculation method for power systems |
Non-Patent Citations (4)
| Title |
|---|
| 张国衡: "电力系统动态潮流计算及网络拓扑分析", 《中国优秀硕士学位论文全文数据库（工程期刊II辑）》 * |
| 李霜: "《电力系统》", 1 August 2006, 重庆大学出版社 * |
| 王尔铿: "《能量管理系统EMS》", 30 November 1998, 北京：科学出版社 * |
| 陈波: "基于负荷导纳模型法的静态电压稳定分析", 《中国优秀硕士学位论文全文数据库（工程科技II辑）》 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105787604A (en) * | 2016-03-22 | 2016-07-20 | 山东大学 | Transmission-distribution-cooperative distribution network situation quick awareness method |
| CN105787604B (en) * | 2016-03-22 | 2019-05-14 | 山东大学 | A kind of distribution situation method for quickly sensing of transmission & distribution collaboration |
| CN107359616B (en) * | 2017-07-18 | 2020-05-19 | 国电南瑞科技股份有限公司 | Method for solving problem of non-convergence of load flow calculation after large-scale power grid mode adjustment |
| CN107359616A (en) * | 2017-07-18 | 2017-11-17 | 国电南瑞科技股份有限公司 | One kind solves the not convergent method of Load flow calculation after large scale electric network mode adjusts |
| CN107957974A (en) * | 2017-11-29 | 2018-04-24 | 广西大学 | A kind of method for extracting polar coordinates power flow equation Jacobian matrix |
| CN107957974B (en) * | 2017-11-29 | 2020-12-18 | 广西大学 | Method for extracting polar coordinate tidal current equation Jacobi matrix |
| CN109344361A (en) * | 2018-08-27 | 2019-02-15 | 南昌大学 | A kind of electric power system tide calculate in the method that is quickly formed of Jacobian matrix |
| CN109344361B (en) * | 2018-08-27 | 2022-05-20 | 南昌大学 | Method for quickly forming Jacobian matrix in power system load flow calculation |
| CN111724014A (en) * | 2019-03-20 | 2020-09-29 | 哈尔滨工业大学（威海） | Weighted-SALSA algorithm-based power system node importance calculation method |
| CN110046450A (en) * | 2019-04-24 | 2019-07-23 | 深圳供电局有限公司 | A kind of Initialization Algorithms that the Newton Power Flow being suitble to the power grid containing hyperconductive cable calculates |
| CN110046450B (en) * | 2019-04-24 | 2023-04-07 | 深圳供电局有限公司 | Initial value selection method suitable for Newton method load flow calculation of superconducting cable-containing power grid |
| CN110460059A (en) * | 2019-07-26 | 2019-11-15 | 国网山东省电力公司泰安供电公司 | A kind of electric power system tide calculation method |
| CN113364007A (en) * | 2021-06-07 | 2021-09-07 | 广东电网有限责任公司 | Two-dimensional admittance model construction method and device, terminal and readable storage medium |
| CN113364007B (en) * | 2021-06-07 | 2023-06-23 | 广东电网有限责任公司 | Two-dimensional admittance model construction method, device, terminal and readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103632046A (en) | Power grid load flow calculation method | |
| JP5815382B2 (en) | Power flow analysis for balanced distribution systems | |
| Balamurugan et al. | Review of power flow studies on distribution network with distributed generation | |
| CN104037764B (en) | The rectangular coordinate Newton load flow calculation method that a kind of Jacobian matrix changes | |
| CN106356859B (en) | A kind of rectangular coordinate Newton load flow calculation method based on Matlab | |
| CN104298809B (en) | A kind of Nonlinear Modeling method for solving based on matrix exponetial electromagnetic transient simulation | |
| CN102420427B (en) | Method for calculating voltage stability margin of regional power grid by considering outer network equivalence | |
| CN102609598B (en) | Method for performing electromagnetic transient-state simulation to large power system | |
| CN102403720B (en) | Hyper-real-time setting method for superposed time sequences based on transient voltage safety margin | |
| CN103336866B (en) | Containing the disposal route of negative resistance branch road in a kind of electromagnetic transient simulation | |
| CN105514971A (en) | Flow calculation method suitable for microgrids in various operation modes | |
| CN103810646B (en) | Improved projection integral algorithm based active power distribution system dynamic simulation method | |
| CN103018534B (en) | Determine the method and system of harmonic voltage | |
| CN106532711A (en) | Newton method power flow calculation method which changes Jacobi matrix with iteration and node types | |
| CN104734148B (en) | Three-phrase power-distributing network continuation power flow analysis of distributed power supply | |
| CN104899396B (en) | A kind of algorithm quicksort tidal current computing method of correction factor matrix | |
| CN104113061B (en) | A kind of distribution network three-phase power flow method containing distributed power source | |
| CN105322546A (en) | Alternating current and direct current decoupling hybrid power flow algorithm | |
| CN105048468A (en) | Power transmission-distribution network integrating voltage stability assessment method based on distributed calculation | |
| CN109149665A (en) | Multi-rate simulating method and system for flexible direct current AC network associative simulation | |
| CN111181164B (en) | Improved master-slave split transmission and distribution cooperative power flow calculation method and system | |
| CN104578055B (en) | A kind of electric power system transient stability simulation method optimized based on network node numbering | |
| CN104967121A (en) | Large-scale electric power system node load flow computing method | |
| CN105119279A (en) | Distributed power supply planning method and system thereof | |
| CN106532712A (en) | Rectangular coordinate Newton method load flow calculation method for small-impedance-branch-containing power grid based on compensation method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140312 |