CN102157938A - Method for online identifying lower-voltage stability nodes of power system - Google Patents

Abstract

The invention discloses a method for online identifying lower-voltage stability nodes of a power system, which relates to the online identification of the lower-voltage stability nodes and belongs to the field of power system safety analysis and monitoring. By the method, the shortcoming of low accuracy or great calculated amount of the conventional lower-voltage stability node identification method is overcome. The method comprises the following steps of: 1, online calling corresponding system network impedance matrix parameters from a database according to real-time network topology structure conditions; 2, calculating equivalent impedance parameters when viewed from each load node by utilizing the voltage and current measurement information of each load node; and 3, calculating the ratio of equivalent impedance to load impedance of each load node, and performing sequencing according to the ratios to realize the online identification of the lower-voltage stability nodes of the system.

Description

Power system voltage stabilization weak node ONLINE RECOGNITION method

Technical field

The present invention relates to voltage and stablize the weak node ONLINE RECOGNITION, belong to power system security analysis, supervision field.

Background technology

In recent years, worldwide the large area blackout that causes owing to voltage collapse happened occasionally, and had caused very big economic loss, had caused the great attention of national governments and Utilities Electric Co..In order to guarantee the safe and stable operation of electric power system, for the user provides high-quality reliable electric energy, the power-management centre needs the ruuning situation of real-time electric power monitoring system.The voltage of electric power system is stablized accident, is normally caused by one or several weak node, then spreads gradually to be the total system accident.Therefore, studying and invent a kind of effective index, technology or means, for the in-service monitoring and the control of power system voltage stabilization weak node provides service, is important content and the task that realizes the power system safety and stability operation.

So far, voltage is stablized the great attention that weak node ONLINE RECOGNITION and monitoring problem have caused power department, and its method mainly contains following several:

1) pattern analysis (modal analysis) method: this method according to the size of characteristic vector, is come the vulnerability of recognition node voltage by the minimum singular value (or characteristic value) and the characteristic vector thereof of solving system Jacobian matrix.Because system's Jacobian matrix is carried out characteristic value decomposition needs data of the whole network and bigger amount of calculation, and singular value/characteristic value has strong nonlinearity and changes very fast near system's critical point.Therefore, slow, the poor linearity of the computational speed of this method.

2) voltage stability domain degree index method: this method need be utilized continuous tide or other instrument, calculates the distance of current operating point to collapse point, utilizes this range index to come the relative vulnerability of recognition node voltage then.Therefore, this method need be carried out repeatedly trend calculating usually, and amount of calculation is very big, is difficult to satisfy the requirement of monitoring in real time.

3) sensitivity index method: this method is utilized the variation relation between some physical quantity in the system (being the approximate differential relation) based on power flow equation, studies the vulnerability of node voltage.This method needs the flow data of the whole network equally, and the sensitivity relation of a certain operating point only this point and near establishment, sensitivity index has strong nonlinearity near collapse point, thereby this method is applied to voltage and stablizes the weak node ONLINE RECOGNITION and have problems sometimes.

4) based on the modulus of impedance ratio method of Dai Weinan equivalence: node Dai Weinan equivalent impedance parameter accurately fast identification be a key factor of this method of restriction.So far, various methods of attempting identification electric power system Dai Weinan equivalent impedance all have limitation in essence, and identification result is often inaccurate, sometimes even the result of mistake can occur.

Summary of the invention

The purpose of this invention is to provide a kind of power system voltage stabilization weak node ONLINE RECOGNITION method, stablize weak node recognition methods poor accuracy or the big shortcoming of amount of calculation to overcome existing voltage.

It may further comprise the steps: one, according to real-time network topology structure situation, call corresponding grid impedance matrix parameter from lane database is online; Two, utilize the voltage and current measurement information of each load bus, calculate the equivalent impedance parameter of seeing into from each load bus; Three, calculate the equivalent impedance of each load bus and the ratio of load impedance, sort, realize ONLINE RECOGNITION system's weak node according to size.

Because the progress of phasor measurement technology, be subjected in recent years paying close attention to comparatively widely based on the modulus of impedance ratio method of synchronous measure.The present invention continues to use this thinking, and by the parsing to electric power system Dai Weinan equivalent parameters, the voltage that has proposed a kind of parsing is stablized weak node ONLINE RECOGNITION method.This method is directly utilized the network impedance data and the node measurement information of discontinuity surface when single, can realize voltage stablize weak node fast, accurately discern.A large amount of simulation results show that this method physical significance is distinct, computational speed is fast, accuracy is high, the linearity is good.

Description of drawings

Fig. 1 is the parsing equivalent network schematic diagram of multiport decoupling zero proposed by the invention.Fig. 2 is each the load bus vulnerability index schematic diagram that utilizes the inventive method to calculate.Fig. 3 utilizes each load bus that the inventive method calculates relative contribution index schematic diagram to load bus 42 vulnerability.Fig. 4 is the contrast schematic diagram that utilizes each load bus vulnerability index that the pattern analysis method calculates.Fig. 5 is the power system network structural representation of a broad sense.

Embodiment

Embodiment one: specify present embodiment below in conjunction with Fig. 1 to Fig. 4.Present embodiment may further comprise the steps: one, according to real-time network topology structure situation, call corresponding grid impedance matrix parameter from lane database is online; Two, utilize the voltage and current measurement information of each load bus, calculate the equivalent impedance parameter of seeing into from each load bus; Three, calculate the equivalent impedance of each load bus and the ratio of load impedance, sort, realize ONLINE RECOGNITION system's weak node according to size.Identify each voltage and stablize after the weak node, can be at each weak node, cause the principal element of this node weakness in line computation and identification, carry out corresponding idle regulation and control or load unloading measure.

In first step, the network impedance matrix under the various topological structures can calculated off-line; During practical application, can directly from offline database, call, saved the computational process and the computing time of network impedance matrix according to ONLINE RECOGNITION to network topology structure.

In second step, the present invention is based on the equivalent notion of the multiport network that proposed, the complicated electric power system equivalence is one group sees into from each load bus and the simple equivalent circuit of approximate decoupling zero as shown in Figure 1, to utilize the diagonal entry Z of network impedance matrix _IiWith off diagonal element Z _Ij, and the measurement electric current of each load bus, calculate equivalent impedance:

$Z_{\mathrm{eq},i}=Z_{\mathrm{ii}}+\underset{j=1,i\≠j}{\overset{n}{\mathrm{\Σ}}}{Z}_{\mathrm{ij}}\frac{I_j}{I_i}(i=1,\·\·\·,n)---\left(1\right)$

In the formula, n is the load point number, I _iAnd I _jBe respectively the measurement electric current of load bus i and j.From formula (1) as can be seen, this equivalence impedance has tangible analytical expression, only load current I _jBigger and corresponding Z _IjAlso bigger load, to the load i the voltage vulnerability bigger contribution is just arranged.Therefore, method proposed by the invention only needs the measurement current information of a small amount of load bus (to those bigger load buses of system weak node contribution), can realize to Dai Weinan equivalent impedance parameter be similar in line computation.

In the 3rd step, utilize the node current and the voltage metric data of each load, calculate the impedance Z of corresponding load _i, computing impedance ratio then,

$r_i=\left|\frac{Z_{\mathrm{eq},i}}{Z_i}\right|(i=1,\·\·\·,n)---\left(2\right)$

Discover Z _{Eq, i}Satisfy a kind of impedance matching condition of special shape, under ideal system maximum power transfer condition, each load impedance Z _iWith Z _{Eq, i}Be complementary.Therefore, the present invention proposes to utilize as the formula (2) impedance ratio index to describe the relative vulnerability of each load bus.A large amount of simulation results confirm that this impedance ratio index can accurately be discerned the relative vulnerability of each load bus, and computational speed is fast, the linearity good; And physical significance distinct, have an analyticity.

Present embodiment can also comprise the 4th step, utilizes the described index r of formula (3) _{I, j}, cause the principal element of this node weakness in line computation and identification:

$r_i=\left|\underset{j=1}{\overset{n}{\mathrm{\Σ}}}\frac{Z_{\mathrm{ij}}{I}_j}{V_{\mathrm{Li}}}\right|=\left|\frac{1}{V_{\mathrm{Li}}}\right|\left|\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{r}_{i,j}\right|---\left(3\right)$

In the formula, r _{I, j}Can be used for describing the relative contribution of each load bus to load i vulnerability:

r _i，j＝Z _ijI _j (4)

From formula (4) as can be seen, those load currents (I only _j) bigger and corresponding Z _IjAlso bigger load, to the load i the voltage vulnerability bigger contribution is just arranged.Therefore, utilize the contribution degree index shown in the formula (4), (vulnerability that reduces load i must be from those load currents (I can ONLINE RECOGNITION to go out to reduce the crucial control measure of system's vulnerability _j) bigger and corresponding Z _IjAlso bigger load bus and respective lines are set about, by carrying out reactive power compensation or cutting load is realized at respective lines switching reactive-load compensation equipment with at corresponding load bus).And, utilize this contribution degree index also can discern which load bus disposes PMU (phase amount measurement unit) for needs critical load node (only those have the load bus of significant contribution degree need dispose PMU to system's weak node) with resolving.

By above analysis as can be seen, tool of the present invention has an enormous advantage: method proposed by the invention has tangible analyticity, do not need to carry out the identification of trend calculating and Dai Weinan equivalent impedance, efficiently solve conventional method and depend on trend calculating or this technical difficult problem of Dai Weinan equivalent parameters identification; Directly utilize the network impedance matrix data and the node measurement information of discontinuity surface when single, can realize that voltage stablizes the ONLINE RECOGNITION of weak node, have that principle is distinct, result of calculation is accurate, computational speed is fast and simple operation and other advantages; And the various control measure that cause the key factor of respective nodes weakness and improve the node vulnerability also can be identified with resolving.A large amount of simulation results show: method computational speed proposed by the invention is fast, the linearity good, result of calculation is accurate; Compare with putative pattern analysis method, method proposed by the invention can identify and corresponding each weak node of all critical modes, and the ranking results consistency is fine, and is not subjected near this constraints limit of collapse point.

Below being one embodiment of the present of invention, is that example is carried out test simulation and analysis with IEEE 118 node systems, and concrete steps and result are as follows:

1) supposes the network impedance matrix of IEEE 118 node systems under various topological structures calculated off-line, the present invention only needs network topology structure (network configuration is intact or the have individual line to stop transport) situation according to current operation, therefore directly from database, call corresponding network impedance matrix data, saved computing time and amount of calculation that other method needs computing network matrix or Jacobian matrix.

2) data of utilizing trend to calculate replace the real-time measurement data of each load bus, comprise the voltage and current measurement information of each heavy load node;

3) utilize formula (2) to calculate the impedance ratio index of each load bus, the result as shown in Figure 2.As can be seen from the figure, the vulnerability of each load bus has tangible regionality, and loads as can be seen and 42 be weak node.

4) be example with weak node 42, utilize formula (4) to calculate the relative contribution of each load bus to its vulnerability, the result as shown in Figure 3.As can be seen from the figure, near several loads of only loading 42 have important influence to the vulnerability of this node; And can find out that the contribution degree maximum of this weak node self is at this cutting load or attempt to reduce the impedance matrix diagonal entry corresponding with this point and can improve its vulnerability.

Above result all matches with the actual of electric power system with discovery.In order further to confirm the validity of institute of the present invention extracting method, Fig. 4 has provided the result of calculation of utilizing the pattern analysis method, and pattern 1 is the most serious pattern among the figure, and pattern 2 is a time serious pattern, and other is arranged in order.Comparison diagram 2 and Fig. 4 as can be seen, method result of calculation proposed by the invention is accurate, can identify and corresponding each weak node of all critical modes, and the ranking results consistency is fine.Compare with the pattern analysis method, method proposed by the invention, avoided the selection problem of a plurality of critical modes, can resolve each weak node and the origin cause of formation of ground recognition system, can realize better that voltage stablizes the in-service monitoring and the control of weak node, advantage such as have that notion is distinct, principle simple, accuracy height, computational speed be fast as a result.

Embodiment two: specify present embodiment below in conjunction with Fig. 5.The difference of present embodiment and execution mode one is: present embodiment be explain in detail the present invention based on theoretical foundation, promptly, it how is one group of multiport equivalent circuit of seeing into approximate decoupling zero from each load bus with the complicated electric power system equivalence, its theoretical derivation process is as follows: broad sense power system network as shown in Figure 5 (this network comprises n load bus, a m generating node and some contact nodes), can write out nodal voltage equation as the formula (11).

$\left[\begin{array}{c}-I_L\\ 0\\ I_G\end{array}\right]=\left[\begin{array}{ccc}{Y}_{\mathrm{LL}}& Y_{\mathrm{LT}}& Y_{\mathrm{LG}}\\ Y_{\mathrm{TL}}& Y_{\mathrm{TT}}& Y_{\mathrm{TG}}\\ Y_{\mathrm{GL}}& Y_{\mathrm{GT}}& Y_{\mathrm{GG}}\end{array}\right]\left[\begin{array}{c}{V}_L\\ V_T\\ V_G\end{array}\right]---\left(11\right)$

In the formula: Y representative system admittance matrix; V and I be representation node voltage and node current respectively; Subscript L, T and G represent load bus, contact node and generating node respectively.

According to formula (11), it is as follows that we can derive the load bus voltage equation:

V _L＝E _eq-ZI _L (12)

In the formula:

$Z={(Y_{\mathrm{LL}}-Y_{\mathrm{LT}}{Y}_{\mathrm{TT}}^{-1}{Y}_{\mathrm{TL}})}^{-1}---\left(13\right)$

$E_{\mathrm{eq}}=Z(Y_{\mathrm{LT}}{Y}_{\mathrm{TT}}^{-1}{Y}_{\mathrm{TG}}-Y_{\mathrm{LG}})V_G---\left(14\right)$

Make E _{Eq, i}Represent E _EqI element, V _LiAnd I _LiRepresent the voltage and current of load i respectively, then formula (12) can be rewritten as:

V _Li＝E _eq，i-Z _eq，iI _Li (i＝1，2，…，n) (15)

$Z_{\mathrm{eq},i}=Z_{\mathrm{ii}}+\underset{j=1,i\≠j}{\overset{n}{\mathrm{\Σ}}}{Z}_{\mathrm{ij}}\frac{I_{\mathrm{Lj}}}{I_{\mathrm{Li}}}(i=1,\·\·\·,n)---\left(16\right)$

In the formula, Z _IjRepresent the capable j column element of i of matrix Z,

Like this, just can obtain a multiport decoupling zero equivalent circuit (as shown in Figure 1) of seeing into from each load bus.

Claims (5)

1. power system voltage stabilization weak node ONLINE RECOGNITION method is characterized in that it may further comprise the steps: one, according to real-time network topology structure situation, call corresponding grid impedance matrix parameter from lane database is online; Two, utilize the voltage and current measurement information of each load bus, calculate the equivalent impedance parameter of seeing into from each load bus; Three, calculate the equivalent impedance of each load bus and the ratio of load impedance, sort, realize ONLINE RECOGNITION system's weak node according to size.

2. power system voltage stabilization weak node ONLINE RECOGNITION method according to claim 1 is characterized in that in the first step, the network impedance matrix calculated off-line under the various topological structures; During ONLINE RECOGNITION,, directly from offline database, call according to ONLINE RECOGNITION to network topology structure.

3. power system voltage stabilization weak node ONLINE RECOGNITION method according to claim 1, it is characterized in that in second step, the complicated electric power system equivalence is one group sees into the multiport equivalent circuit of approximate decoupling zero, utilize the diagonal entry Z of network impedance matrix from each load bus _IiWith off diagonal element Z _Ij, and the measurement electric current of each load bus, calculate equivalent impedance:

$Z_{\mathrm{eq},i}=Z_{\mathrm{ii}}+\underset{j=1,i\≠j}{\overset{n}{\mathrm{\Σ}}}{Z}_{\mathrm{ij}}\frac{I_j}{I_i}(i=1,\·\·\·,n)---\left(1\right)$

In the formula, n is the load point number, I _iAnd I _jBe respectively the measurement electric current of load bus i and j.

4. power system voltage stabilization weak node ONLINE RECOGNITION method according to claim 1 is characterized in that in third step, utilizes the electric current and the voltage metric data of each load bus, calculates the impedance Z of corresponding load _i, computing impedance ratio then,

$r_i=\left|\frac{Z_{\mathrm{eq},i}}{Z_i}\right|(i=1,\·\·\·,n).---\left(2\right)$

5. power system voltage stabilization weak node ONLINE RECOGNITION method according to claim 1 is characterized in that it also comprises the 4th step: utilize the described r of formula (3) _{I, j}Index causes the principal element of respective nodes weakness in line computation and identification:

$r_i=\left|\underset{j=1}{\overset{n}{\mathrm{\Σ}}}\frac{Z_{\mathrm{ij}}{I}_j}{V_{\mathrm{Li}}}\right|=\left|\frac{1}{V_{\mathrm{Li}}}\right|\left|\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{r}_{i,j}\right|---\left(3\right)$

In the formula, r _{I, j}Can be used for describing the relative contribution of each load bus to load i vulnerability:

r _i，j＝Z _ijI _j。(4)

Images