Key area identification and power grid oscillation risk early warning method and device based on node modal voltage distribution

Abstract

The invention provides a key area identification and power grid oscillation risk early warning method and device based on node modal voltage distribution in order to evaluate the stability of a multi-ring network wind and light distributed access system and further analyze an instability scene of the multi-ring network wind and light distributed access system, and the key area identification and power grid oscillation risk early warning method and device based on node modal voltage distribution are combined according to linear transformation of a research system node voltage equation and a discrete state space state variable. And the modal quantity response condition of each node of the system is calculated, and a node modal voltage distribution coefficient is defined as an index, so that a key area and a line are positioned, and system operation personnel are helped to strengthen monitoring on a local power grid and take related countermeasures at a proper position. According to the method, verification is carried out in combination with an actual multi-ring network wind and light distributed access system, and the effectiveness of the method is verified through time domain simulation and eigenvalue analysis.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of small signal stability analysis of large systems, and particularly relates to a key area identification and power grid oscillation risk early warning method and device based on node modal voltage distribution. BACKGROUND

[0002] With the rapid development of "double high" power systems, the penetration rate of wind and solar distributed power sources based on converter interfaces continues to rise, and the dynamic characteristics of the system show a strong power electronic trend. Under this background, wideband oscillation, as an external manifestation of the interaction of multi-time scale control loops, has become a new and prominent problem threatening the safe operation of the power grid. Existing research mainly focuses on impedance modeling and stability criterion derivation of single converter device, or analyzes the oscillation characteristics of small-scale systems through simplified equivalent methods. However, in actual regional-level systems with multi-loop network structure, the interaction between dynamic coupling of device clusters and network topology and power flow distribution is not clear.

[0003] And in actual large systems, the change of system structure and power flow characteristics is closely related to the stability of new energy grid connection, therefore, it is urgent to develop a method for identifying key oscillation risk areas to help system operators strengthen the monitoring of local power grids and take appropriate measures. The present application provides a key area identification and power grid oscillation risk early warning method based on node modal voltage distribution, which can effectively calculate the node modal voltage distribution according to the actual large system network, thereby locating the key areas and lines, and is suitable for risk early warning of current actual large system wideband oscillation research. SUMMARY

[0004] The present application aims to solve the above-mentioned problems of the prior art, and provides a key area identification and power grid oscillation risk early warning method and device based on node modal voltage distribution, which can identify the key oscillation risk area of a complex system, help system operators strengthen the monitoring of local power grids and take appropriate measures. It can correctly quantify the small signal stability of a complex system, and effectively solve the problem that the influence of the change of structure and power flow characteristics on the stability of new energy grid connection is not clear, and the interaction between dynamic coupling of device clusters and network topology and power flow distribution is not clear.

[0005] A key area identification and power grid oscillation risk early warning method based on node modal voltage distribution, comprising the following steps:

[0006] Step 1: Construct a discrete state space model of the whole system to obtain a discrete state space coefficient matrix A composed of coefficients in the historical current source of each element d , B d , C d , D d , wherein Ad Let B be the state matrix. d Given the input matrix, C d For the output matrix, D d For direct matrix transmission;

[0007] Step 2: Construct the system admittance matrix G and branch-node correlation matrix L for the entire system. t ;

[0008] Step 3: Simultaneously solve the discrete state-space coefficient matrix A obtained in Step 1. d B d C d D d Obtain the discrete state space state matrix A D Then, by diagonalizing and decomposing the state variables, the relationship between the state variables and the modal quantities is obtained by expanding the corresponding state variables into linear transformations.

[0009] Step 4: Simultaneously solve the discrete state-space coefficient matrix obtained in Step 1, the system admittance matrix G obtained in Step 2, and the branch-node correlation matrix L. t Based on the relationship between the state variables and modal quantities obtained in step three, the relationship between node voltage and modal quantities is obtained, and thus the response of the i-th mode in each node voltage is obtained. The voltage x-axis component and y-axis component are superimposed to obtain the voltage oscillation component of the i-th mode in node p. Based on the voltage oscillation component Obtain the node distribution coefficient Based on node distribution coefficient Locate key areas and routes, and deploy relevant measuring devices based on the located key areas and routes to obtain the operating status of the key areas. When the operating status is at the threshold edge, issue an early warning and take corresponding measures.

[0010] Furthermore, step one includes:

[0011] Based on the system under study, a discrete state-space model of the entire system is constructed: First, the relevant parameters and operating conditions of the system components are input to establish the discrete state-space model of each component, thereby obtaining the discrete state-space coefficient matrix A of the system. d B d C d D d The discretized state-space model of the individual element includes a single-port model, an AC two-port model, and a DC two-port model.

[0012] in:

[0013] The iterative expression for the historical current term in the single-port model and its relationship with the unit output current are as follows:

[0014] (1);

[0015] The iterative expression of the history current term of the AC two-port model and the relationship between the history current term and the unit output current are as follows:

[0016] (2);

[0017] The iterative expression of the history current term of the DC two-port model and the relationship between the history current term and the unit output current are as follows:

[0018] (3);

[0019] In the formula, h(t) represents a state variable, the system discrete state space coefficient matrix A d , B d , C d , and D d are diagonal block matrices composed of coefficient matrices in the history current sources of respective elements, wherein D d is a diagonal block matrix composed of coefficient matrices D d-s , D d-L , D d-h1 , and D d-h2 , the subscripts d-s, d-L, d-h1, d-h2, d-h3, and d-h4 represent coefficient matrices under different elements; U xy and i xy represent the voltage and the current of the model port, and the subscripts sxy, Lxy, 1xy, and 2xy are used to distinguish different elements.

[0020] Further, step two specifically comprises:

[0021] According to the discrete circuit network, a corresponding system admittance matrix G and a branch-node incidence matrix Lt are obtained:

[0022] (4);

[0023] (5);

[0024] In the formula, the diagonal element value G ii of the system admittance matrix G is equal to the sum of the equivalent conductances of all elements connected to node i, and the non-diagonal element G ij is equal to the negative value of the equivalent conductance of the element directly connected between nodes i and j; the branch-node incidence matrix L tIn the middle, the elements at the positions corresponding to the nodes connected by the single-port element are set to the unit matrix I2, the elements at the positions corresponding to the two nodes connected by the alternating-current two-port element are set to I2 and -I2 respectively, the elements at the positions corresponding to the two nodes connected by the direct-current two-port element are set to I2, and the elements at the other positions are all set to 0.

[0025] Further, step three specifically includes:

[0026] The discrete state matrix A of the alternating-current system D The calculation expression is:

[0027] (6) ;

[0028] The diagonalization decomposition is performed on the discrete state matrix A D

[0029] (7) ;

[0030] In the formula, Λ is a diagonal matrix composed of all eigenvalues of the discrete state matrix A D , wherein A D =A d +B d L t G -1 (-L t ) T D d。

[0031] V and W are left and right eigenvector matrices corresponding to the discrete eigenvalues respectively, and satisfy V T W=I, wherein I is a unit matrix with the same dimension as the state matrix;

[0032] The same linear transformation is performed on the state variables to obtain modal quantities:

[0033] (8) ;

[0034] The system discrete state space equation expressed by the modal quantities is obtained by combining formula (6) and formula (7):

[0035] (9).

[0036] Further, step four specifically includes:

[0037] In the discrete state space, the historical current source is a function of the discrete state variable h(t) of the element, and the relationship between the node injected current and the discrete state variable of the element is established according to this:

[0038] (10) ;

[0039] ​where the superscript 'T' denotes the transpose of a matrix, D d is a diagonal block matrix composed of coefficient matrices in the historical current sources of all elements, and h(t) is a full system state vector containing discrete state variables of all elements;

[0040] The node conductance matrix G in step two and the branch-node incidence matrix L t The relationship between the system node voltage and the discrete state variable is obtained:

[0041] (11) ;

[0042] The mathematical relationship between the modal quantity and the discrete state variable in step three is combined to obtain the relationship between the system node voltage and the modal quantity:

[0043] (12) ;

[0044] In the formula, S is defined as the node voltage observability matrix of the system;

[0045] The i-th column of the matrix S represents the response of the i-th mode in the node voltage. Since the node voltage and branch current are converted to the unified xy rotating coordinate system in the full system state space modeling, the x-axis component and the y-axis component are superimposed to obtain the voltage oscillation component of the i-th mode in the node p :

[0046] (13) ;

[0047] In the formula, the letters in the subscript parentheses represent the matrix element index number;

[0048] The node voltage distribution coefficient is defined as follows:

[0049] (14) ;

[0050] According to the calculation results of the node voltage distribution coefficient, the power grid interval composed of nodes with relatively large relative amplitude is determined as the main oscillation risk area of the system, and then the key area and line are located, helping the system operator to strengthen the monitoring of the local power grid and take relevant measures at the appropriate position, realizing the power grid oscillation risk early warning.

[0051] A key area identification and power grid oscillation risk early warning device based on node modal voltage distribution, comprising:

[0052] A first matrix acquisition module for constructing a discrete state space model of the full system to obtain a discrete state space coefficient matrix A d , B d , C d , Dd where A d is the state matrix, B d is the input matrix, C d is the output matrix, and D d is the direct transfer matrix.

[0053] The second matrix obtaining module is configured to construct a system admittance matrix G and a branch-node association matrix L t of the whole system.

[0054] The state-mode relationship obtaining module is configured to simultaneously obtain the discrete state space coefficient matrices A d , B d , C d , D d , obtain a discrete state space state matrix A D , and diagonalize and decompose the discrete state space state matrix A t to obtain a relationship between the state variables and modal quantities.

[0055] The early warning module is configured to simultaneously obtain the discrete state space coefficient matrices, the system admittance matrix G and the branch-node association matrix L t , and the relationship between the state variables and the modal quantities, obtain a relationship between the node voltages and the modal quantities, and further obtain a response of the i-th mode in each node voltage, superimpose the voltage x-axis component and the voltage y-axis component to obtain a voltage oscillation component of the i-th mode in the node p , obtain a node distribution coefficient based on the voltage oscillation component , and locate a critical area and a line based on the node distribution coefficient , arrange a related measuring device based on the located critical area and line to obtain an operating condition of the critical area, and give an early warning and take corresponding measures when the operating condition is on the threshold edge.

[0056] Further, the first matrix obtaining module is specifically configured to: construct a discrete state space model of the whole system based on the researched system; first, input related parameters and operating conditions of system elements to establish a discrete state space model of a single element, and obtain discrete state space coefficient matrices A d , B d , C d , D d of the system; the discrete state space model of the single element includes a single-port model, an alternating current two-port model, and a direct current two-port model.

[0057] wherein:

[0058] The iterative expression of the historical current term of the single-port model and the relationship between the historical current term and the output current of the unit are as follows:

[0059] (1);

[0060] The iterative expression of the history current term of the AC two-port model and the relationship between the history current term and the unit output current are as follows:

[0061] (2);

[0062] The iterative expression of the history current term of the DC two-port model and the relationship between the history current term and the unit output current are as follows:

[0063] (3);

[0064] In the formula, h(t) represents a state variable, the system discrete state space coefficient matrix A d , B d , C d , and D d are respectively composed of diagonal block matrices of coefficient matrices in the history current sources of the elements, wherein D d is a diagonal block matrix composed of coefficient matrices D d-s , D d-L , D d-h1 , and D d-h2 , the subscripts d-s, d-L, d-h1, d-h2, d-h3, and d-h4 respectively represent the coefficient matrices under different elements; U xy and i xy represent the voltage and current of the model port, and the subscripts sxy, Lxy, 1xy, 2xy, and 3xy are used to distinguish different elements.

[0065] Further, the second matrix obtaining module is specifically configured to: obtain a corresponding system admittance matrix G and a branch-node association matrix Lt according to the discrete circuit network.

[0066] (4);

[0067] (5);

[0068] In the formula, the diagonal element value G ii of the system admittance matrix G is equal to the sum of the equivalent conductances of all elements connected to node i, the non-diagonal element G ij is equal to the negative value of the equivalent conductance of the element directly connected between nodes i and j; in the branch-node association matrix L t , the elements corresponding to the nodes connected by the single-port element are set as an identity matrix I2, the elements corresponding to the two nodes connected by the AC two-port element are respectively set as I2 and -I2, the elements corresponding to the two nodes connected by the DC two-port element are both set as I2, and the elements at the remaining positions are all set as 0.

[0069] Further, a mode relationship obtaining module, specifically configured to: diagonalize and decompose a discrete state matrix A of the alternating current system D The calculation expression is:

[0070] (6) ;

[0071] Diagonalize and decompose the discrete state matrix A D :

[0072] (7) ;

[0073] In the formula, Λ is a diagonal matrix composed of all eigenvalues of the discrete state matrix A D , wherein A D =A d +B d L t G -1 (-L t ) T D d。

[0074] V and W are left and right eigenvector matrices corresponding to the discrete eigenvalues, respectively, and satisfy V T W=I, wherein I is a unit matrix with the same dimension as the state matrix;

[0075] Make the same linear transformation on the state variables to obtain modal quantities:

[0076] (8) ;

[0077] The system discrete state space equation expressed by the modal quantities is obtained by combining formula (6) and formula (7):

[0078] (9).

[0079] Further, a warning module, specifically configured to: in the discrete state space, the historical current source is a function of the element discrete state variable h(t), and accordingly, the relationship between the node injected current and the element discrete state variable is established:

[0080] (10) ;

[0081] In the formula, the superscript 'T' represents the transpose of the matrix, D d is a diagonal block matrix composed of the coefficient matrix in each element historical current source, and h(t) is a full-system state vector containing all element discrete state variables;

[0082] The relationship between the system node voltage and the discrete state variable is obtained by combining the node conductance matrix G and the branch-node association matrix L t :

[0083] (11) ;

[0084] The mathematical relationship of the modal quantity and the discrete state variable is obtained again, and the relationship between the system node voltage and the modal quantity is obtained:

[0085] (12) ;

[0086] In the formula, S is defined as the node voltage observability matrix of the system;

[0087] The i-th column of the matrix S represents the response of the i-th mode in each node voltage, and since the node voltage and branch current are converted into the unified xy rotating coordinate system in the state space modeling of the whole system, the x-axis component and the y-axis component are superimposed to obtain the voltage oscillation component of the i-th mode in the node p :

[0088] (13) ;

[0089] In the formula, the subscript letter in the parentheses represents the index number of the matrix element;

[0090] The node voltage distribution coefficient is defined as follows:

[0091] (14) ;

[0092] According to the calculation results of the node voltage distribution coefficient, the power grid interval formed by the nodes with relatively large relative amplitude is delimited as the main oscillation risk area of the system, and then the key area and the line are located, helping the system operator to strengthen the monitoring of the local power grid and take relevant measures at the appropriate position, realizing the power grid oscillation risk early warning.

[0093] The beneficial effects of the present application are: the present application explores the key area identification and power grid oscillation risk early warning method based on node modal voltage distribution, according to the research on the linear transformation of the system node voltage equation and the discrete state space state variable, the modal quantity response of each node of the system is calculated, and the node modal voltage distribution coefficient is defined as an index, so as to locate the key area and the line, help the system operator to strengthen the monitoring of the local power grid and take relevant measures at the appropriate position. BRIEF DESCRIPTION OF DRAWINGS

[0094] Figure 1 It is a schematic diagram of the system discrete circuit network containing diversified power elements of the present application;

[0095] Figure 2 It is a schematic diagram of the branch-node correlation matrix form of the present application;

[0096] Figure 3 A network frame structure diagram of a research system of the present application;

[0097] Figure 4 A 220kV each node common mode modal distribution histogram of the present application;

[0098] Figure 5 A 220kV each node different mode modal distribution histogram of the present application;

[0099] Figure 6 A 220kV each node common mode modal distribution heat map of the present application;

[0100] Figure 7 A 220kV each node different mode modal distribution heat map of the present application;

[0101] Figure 8 A time domain simulation result of the present application;

[0102] Figure 9 A flow chart of a key area identification and power grid oscillation risk early warning method based on node modal voltage distribution provided by the present application. DETAILED DESCRIPTION

[0103] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0104] Please refer to Figure 9 The present application provides a key area identification and power grid oscillation risk early warning method based on node modal voltage distribution, which comprises the following steps:

[0105] Step one, in the discrete state space, since the discrete state variable h(t) of each power element is a linear combination of the previous time state variable h(t-Δt) and the input voltage U(t-Δt), therefore the key to generating the discrete state matrix of the whole system by combining the models of different elements is to eliminate the intermediate variable U(t-Δt). Considering that any power element can be expressed in the form of a discrete equivalent circuit, if the equivalent circuits of all elements in the system are interconnected according to the actual topological structure, then the discrete circuit network of the whole system can be obtained as shown in Figure 1 .

[0106] Specifically, a discrete state space model of the whole system is constructed based on the studied system: first, relevant parameters and operating conditions of system elements are input to establish a discrete state space model of a single element, and a discrete state space coefficient matrix A d , B d , C d , D d of the system is obtained; the discrete state space model of the single element includes a single-port model, an alternating-current two-port model and a direct-current two-port model;

[0107] wherein:

[0108] The iterative expression of the historical current term of the single-port model and the relationship between the historical current term and the output current of the unit are as follows:

[0109] (1);

[0110] The iterative expression of the historical current term of the alternating-current two-port model and the relationship between the historical current term and the output current of the unit are as follows:

[0111] (2);

[0112] The iterative expression of the historical current term of the direct-current two-port model and the relationship between the historical current term and the output current of the unit are as follows:

[0113] (3);

[0114] In the formula, h(t) represents a state variable, and the system discrete state space coefficient matrix A d , B d , C d , D d is composed of the coefficient matrix of each element historical current source, wherein D d is a diagonal block matrix composed of the coefficient matrix D d-s , D d-L , D d-h1 , D d-h2 of each element historical current source, and the subscripts d-s, d-L, d-h1, d-h2, d-h3 and d-h4 represent the coefficient matrix under different elements; U xy and i xy represent the voltage and current of the model port, and the subscripts sxy, Lxy, 1xy and 2xy are used to distinguish different elements.

[0115] Step two, for a complex power system, which can be represented as a discrete circuit network containing p nodes and q branches, the node conductance matrix G with a dimension of 2p×2p and the branch-node association matrix L with a dimension of 2q×2p can be obtained according to the basic circuit knowledge twhere the diagonal elements of the conductance matrix G are the sum of the conductances of all elements connected to node i, the off-diagonal elements G ii are the negative of the conductances of the elements directly connected between nodes i, j; in the branch-node incidence matrix (as shown in Fig. 2), the elements corresponding to the nodes connected by single-port elements are set to the identity matrix I2, the elements corresponding to the two nodes connected by an AC two-port element are set to I2and -I2respectively, the elements corresponding to the two nodes connected by a DC two-port element are set to I2, and the elements in the rest of the matrix are set to 0. ij Figure 2 t .

[0116] (4)

[0117] (5)

[0118] Step 3: Transform the discrete state variables of the system into the discrete state space of the system represented by modal quantities through linear transformation.

[0119] The calculation expression of the discrete state matrix A D of the AC system is:

[0120] (6)

[0121] In eigenvalue analysis, in order to characterize the influence of the state variables of the system on the characteristic modes, modal decomposition is generally performed on the state matrix. In essence, it is diagonalization decomposition on the state matrix:

[0122] (7)

[0123] where Λ is a diagonal matrix composed of all eigenvalues of the discrete state matrix A D , V and W are respectively the left and right eigenvector matrices corresponding to the discrete eigenvalues, and they are orthogonal to each other, satisfying V T W = I, where I is the unit matrix with the same dimension as the state matrix.

[0124] After the same linear transformation is performed on the state variables, a new set of system state variables Z(t) is obtained, which is called modal quantities

[0125] (8)

[0126] By combining equation (32) and equation (33), the discrete state space equation of the system represented by modal quantities is obtained:

[0127] (9)​​

[0128] Step 4: Simultaneously solve the discrete state-space coefficient matrix obtained in Step 1, the system admittance matrix G obtained in Step 2, and the branch-node correlation matrix L. t Based on the relationship between the state variables and modal quantities obtained in step three, the relationship between node voltage and modal quantities is obtained, and thus the response of the i-th mode in each node voltage is obtained. The voltage x-axis component and y-axis component are superimposed to obtain the voltage oscillation component of the i-th mode in node p. Based on the voltage oscillation component Obtain the node distribution coefficient Based on node distribution coefficient Locate key areas and routes, and deploy relevant measuring devices based on the located key areas and routes to obtain the operating status of the key areas. When the operating status is at the threshold edge, issue an early warning and take corresponding measures.

[0129] Specifically, by simultaneously solving the equations in steps two and three, under zero-input conditions, the node injection current in the discrete circuit network consists of the historical current source (HCS) of the components connected to that node. Since the historical current source is a function of the component's discrete state variable h(t), the relationship between the node injection current and the component's discrete state variable can be established as follows:

[0130] (10)

[0131] Furthermore, the relationship between node voltage and state variables can be obtained:

[0132] (11)

[0133] Furthermore, the relationship between system node voltages and modal quantities is obtained:

[0134] (12)

[0135] Since the modal quantities represented by Z(t) are decoupled from each other, the i-th column of matrix S characterizes the response of the i-th mode in each node voltage. Because the node voltages and branch currents are converted to a unified xy rotating coordinate system for the entire network during state-space modeling, for an n-node system, the node voltage vector in equation (37) should be of order 2n×1 (containing the x / y axis components of the voltage). Therefore, to obtain the total response of the oscillation components in each node voltage, the x-axis components and y-axis components need to be superimposed. Specifically, the voltage oscillation component of the i-th mode in node p... The calculation is as follows:

[0136] (13)

[0137] The node voltage distribution coefficient is further defined as follows:

[0138] (14)

[0139] According to the calculation results of the node voltage distribution coefficients, the power grid section formed by the nodes with relatively large amplitudes can be delimited as the main oscillation risk area of the system. Further, the key area and line are located, helping the system operator to strengthen the monitoring of the local power grid and take relevant measures at the appropriate position, so as to realize the power grid oscillation risk early warning.

[0140] Method effectiveness verification:

[0141] Based on the actual large system such as Figure 3 , a discretized state space model is constructed, the common mode and the different mode modalities contained between the new energy are analyzed by the eigenvalue and participation factor, the node modal voltage distribution coefficients of each node are calculated as shown in Figure 4-7 , the system stability is tested under different operating modes, and the effectiveness of the proposed method is verified by eigenvalue analysis and time domain simulation.

[0142] The light and photovoltaic parameters are kept unchanged, and the scenes are set reasonably as follows:

[0143] Case1: Set N-1 operating mode, disconnect the large ring in the new energy near area (i.e. Figure 1 C220-T220 bus in

[0144] Case2: Set N-1 operating mode, disconnect the small ring in the new energy near area (i.e. Figure 1 D220-H220 bus in

[0145] Case3: Set N-1 operating mode, disconnect the large ring in the new energy far area (i.e. Figure 1 Q220-R220 bus in

[0146] Case4: Set N-2 operating mode, disconnect the small ring and the large ring at the same time (i.e. Figure 1 C220-T220 bus and D220-H220 bus in

[0147] The eigenvalue calculation results are as follows:

[0148] Table 1 Eigenvalues of common mode and different mode under different scenes

[0149] Scenario Common mode modal λ 384,385 ]]> Isomodal mode λ 386,387 ]]> Full wiring -2.32±j39.32×2π -6.17±j40.36×2π Case 1 0.68±j38.65×2π -2.85±j39.06×2π Case 2 -1.15±j38.86×2π -6.10±j40.34×2π Case 3 -2.31±j39.09×2π -6.11±j40.34×2π Case 4 3.53±j38.33×2π 0.79±j39.97×2π

[0150] The stability of the system will change with the change of power flow characteristics and network structure in different operation scenarios. In the operation mode of near-zone solution large ring, both modes are greatly affected, and the common mode mode appears instability phenomenon, while the near-zone solution small ring and the far-zone solution large ring have little effect on the stability of both. In the N-2 operation mode of Case4, the system appears the phenomenon of superposition of common mode and different mode modes. Correspondingly, from the node modal voltage distribution, when C220-T220 is disconnected, the influence on the common mode and the different mode is the largest, and the influence of D220-H220 disconnection on the common mode is greater than that on the different mode, and Q220-R220 disconnection has little effect on both, which is consistent with the calculation result in Table 1. The simulation result is shown in Figure 8 .

[0151] The embodiment of the application also provides a key area identification and power grid oscillation risk early warning device based on node modal voltage distribution, comprising:

[0152] The first matrix acquisition module is used for constructing a discrete state space model of the whole system to obtain a discrete state space coefficient matrix A d , B d , C d , D d composed of coefficients in historical current sources of each element, wherein A d is a state matrix, B d is an input matrix, C d is an output matrix, and D d is a direct transfer matrix.

[0153] The second matrix acquisition module is used for constructing a system admittance matrix G and a branch-node association matrix L t of the whole system.

[0154] The mode relationship acquisition module is used for simultaneously obtaining the discrete state space coefficient matrix A d , B d , C d , D d , obtaining a discrete state space state matrix A D , and diagonalizing and decomposing the same to obtain the relationship between the state variable and the modal quantity through linear change of the corresponding state variable.

[0155] The early warning module is used for simultaneously obtaining the discrete state space coefficient matrix, the system admittance matrix G and the branch-node association matrix L t , and the relationship between the state variable and the modal quantity to obtain a relationship between the node voltage and the modal quantity, and then obtain the response of the i-th mode in each node voltage, superimpose the voltage x-axis component and the y-axis component to obtain the voltage oscillation component of the i-th mode in the node p , and the voltage oscillation component is used for early warning. obtaining a node distribution coefficient based on the node distribution coefficient locating the critical area and the line, arranging the related measuring device based on the located critical area and the line to obtain the operation condition of the critical area, when the operation condition is running at the threshold edge, giving an early warning and taking corresponding measures.

[0156] The present application is based on the discrete state space modeling of complex system, combining the system node voltage equation and the linear transformation of the discrete state space state variable to solve the modal quantity response of each node of the system, and defines the node modal voltage distribution coefficient as the index, so as to locate the critical area and the line, help the system operator to strengthen the monitoring of the local power grid and take the relevant measures at the appropriate position. The method is verified by combining the actual multi-loop network wind and light distributed access system, and the effectiveness of the method is verified by time domain simulation and eigenvalue analysis.

[0157] Finally, it should be noted that: the above examples are only used to illustrate the technical solutions of the present application, but not to limit it, although the present application has been described in detail with reference to the above examples, those skilled in the art should understand that: the specific embodiments of the present application can still be modified or replaced by the equivalent, without departing from the spirit and scope of the present application, any modification or equivalent replacement, which should be covered within the protection scope of the claims of the present application.

Claims

1. A method for identifying key areas and providing early warning of power grid oscillation risks based on nodal modal voltage distribution, characterized in that, Includes the following steps: Step 1: Construct a discrete state-space model of the entire system to obtain the discrete state-space coefficient matrix A, which is composed of the coefficient matrices of the historical current sources of each component. d B d C d D d A d Let B be the state matrix. d Given the input matrix, C d For the output matrix, D d For direct matrix transmission; Step 2: Construct the system admittance matrix G and branch-node correlation matrix L for the entire system. t ; Step 3: Simultaneously solve the discrete state-space coefficient matrix A obtained in Step 1. d B d C d D d Obtain the discrete state space state matrix A D Then, by diagonalizing and decomposing the state variables, the relationship between the state variables and the modal quantities is obtained by expanding the corresponding state variables into linear transformations. Step 4: Simultaneously solve the discrete state-space coefficient matrix obtained in Step 1, the system admittance matrix G obtained in Step 2, and the branch-node correlation matrix L. t Based on the relationship between the state variables and modal quantities obtained in step three, the relationship between node voltage and modal quantities is obtained, and thus the response of the i-th mode in each node voltage is obtained. The voltage x-axis component and y-axis component are superimposed to obtain the voltage oscillation component of the i-th mode in node p. Based on the voltage oscillation component Obtain the node distribution coefficient Based on node distribution coefficient Locate key areas and routes, and deploy relevant measuring devices based on the located key areas and routes to obtain the operating status of the key areas. When the operating status is at the threshold edge, issue an early warning and take corresponding measures.

2. The method for identifying key areas and providing early warning of power grid oscillation risks based on nodal mode voltage distribution according to claim 1, characterized in that, Step one includes: Based on the system under study, a discrete state-space model of the entire system is constructed: First, the relevant parameters and operating conditions of the system components are input to establish the discrete state-space model of each component, thereby obtaining the discrete state-space coefficient matrix A of the system. d B d C d D d The discretized state-space model of the individual element includes a single-port model, an AC two-port model, and a DC two-port model. in: The iterative expression for the historical current term in the single-port model and its relationship with the unit output current are as follows: （1）； The iterative expression for the historical current term in the AC two-port model and its relationship with the unit output current are as follows: （2）； The iterative expression for the historical current term in the DC two-port model and its relationship with the unit output current are as follows: （3）； In the formula, h(t) represents the state variable, and A is the discrete state-space coefficient matrix of the system. d B d C d D d A diagonal block matrix composed of the coefficient matrices from the historical current sources of each element, where D d The coefficient matrix D of the historical current sources of each component d-s D d-L D d-h1 D d-h2 The diagonal block matrix is ​​formed, with subscripts ds, dL, d-h1, d-h2, d-h3, and d-h4 representing the coefficient matrices for different components; U xy and i xy The voltage and current at the model port are indicated by the subscripts sxy, Lxy, 1xy, and 2xy to distinguish different components.

3. The method for identifying key areas and providing early warning of power grid oscillation risks based on nodal mode voltage distribution according to claim 1, characterized in that, Step two specifically includes: Based on the discrete circuit network, the corresponding system admittance matrix G and branch-node correlation matrix Lt are obtained: （4）； （5）； The diagonal elements of the system admittance G matrix are G... ii Equal to the sum of the equivalent conductances of all elements connected to node i, and the off-diagonal element G ij It is equal to the negative of the equivalent conductance of the element directly connected between nodes i and j; in the branch-node correlation matrix L t In the matrix, the elements at the corresponding positions of the nodes connected to the single-port element are set to the identity matrix I2, the elements at the corresponding positions of the two nodes connected to the AC two-port element are set to I2 and -I2 respectively, the elements at the corresponding positions of the two nodes connected to the DC two-port element are both set to I2, and the elements at the remaining positions are all set to 0.

4. The method for identifying key areas and providing early warning of power grid oscillation risks based on nodal mode voltage distribution according to claim 1, characterized in that, Step three specifically includes: Discrete state matrix A of the AC system D The calculation expression is: （6）； For discrete state matrix A D Perform diagonalization decomposition: （7）； In the formula, Λ is the discrete state matrix A D A is a diagonal matrix consisting of all its eigenvalues, where A D =A d +B d L t G -1 (-L t ) T D d； V and W are the left and right eigenvector matrices corresponding to the discrete eigenvalues, respectively, satisfying V T W=I, where I is the identity matrix with the same dimension as the state matrix; By performing equivalent linear transformations on the state variables, we obtain the modal quantities: （8）； Combining equations (6) and (7), we obtain the discrete state-space equations of the system expressed in terms of modal quantities: （9）。 5. The method for identifying key areas and providing early warning of power grid oscillation risks based on nodal mode voltage distribution according to claim 3, characterized in that, Step four specifically includes: In the discrete state space, the historical current source is a function of the discrete state variable h(t) of the component, and the relationship between the node injection current and the discrete state variable of the component is established accordingly: （10）； In the formula, the superscript 'T' denotes the transpose of the matrix, and D d h(t) is a diagonal block matrix composed of the coefficient matrices of the historical current sources of each element, and h(t) is the state vector of the whole system containing the discrete state variables of all elements. Simultaneously establish the nodal conductance matrix G and the branch-node correlation matrix L from step two. t The relationship between system node voltages and discrete state variables is obtained: （11）； By combining the mathematical relationships between the modal quantities and discrete state variables from step three, we can obtain the relationship between the system node voltages and the modal quantities: （12）； In the formula, S is defined as the node voltage observability matrix of the system; The i-th column of matrix S represents the response of the i-th mode in the voltage at each node. Since the node voltages and branch currents are converted to a unified xy rotating coordinate system for the entire network when modeling the state space of the whole system, the x-axis components are superimposed with the y-axis components to obtain the voltage oscillation component of the i-th mode at node p. : （13）； In the formula, the letters in the parentheses represent the index numbers of the matrix elements; The node voltage distribution coefficient is defined as follows: （14）； Based on the calculation results of the voltage distribution coefficient of each node, the power grid section composed of nodes with relatively large amplitudes is designated as the main oscillation risk area of ​​the system. This helps to locate key areas and lines, enabling system operators to strengthen the monitoring of the local power grid and take relevant countermeasures at appropriate locations, thereby achieving early warning of power grid oscillation risks.

6. A device for identifying key areas and providing early warning of power grid oscillation risks based on nodal mode voltage distribution, characterized in that, include: The first matrix acquisition module is used to construct the discrete state-space model of the entire system and obtain the discrete state-space coefficient matrix A composed of the coefficient matrices of the historical current sources of each component. d B d C d D d A d Let B be the state matrix. d Given the input matrix, C d For the output matrix, D d For direct matrix transmission; The second matrix acquisition module is used to construct the system admittance matrix G and the branch-node correlation matrix L of the entire system. t ; The module for obtaining state-mode relations is used to simultaneously solve the obtained discrete state-space coefficient matrix A. d B d C d D d Obtain the discrete state space state matrix A D Then, by diagonalizing and decomposing the state variables, the relationship between the state variables and the modal quantities is obtained by expanding the corresponding state variables into linear transformations. The early warning module is used to simultaneously establish the discrete state-space coefficient matrix, the system admittance matrix G, and the branch-node correlation matrix L. t By analyzing the relationship between state variables and modal quantities, the relationship between node voltages and modal quantities is obtained, leading to the response of the i-th mode in each node voltage. The x-axis and y-axis voltage components are then superimposed to obtain the voltage oscillation component of the i-th mode at node p. Based on the voltage oscillation component Obtain the node distribution coefficient Based on node distribution coefficient Locate key areas and routes, and deploy relevant measuring devices based on the located key areas and routes to obtain the operating status of the key areas. When the operating status is at the threshold edge, issue an early warning and take corresponding measures.

7. The device for identifying key areas and providing early warning of power grid oscillation risks based on nodal mode voltage distribution as described in claim 6, characterized in that, The first matrix acquisition module is specifically used for: constructing a discrete state-space model of the entire system based on the system under study; firstly, inputting the relevant parameters and operating conditions of the system components to establish a discrete state-space model of a single component, and obtaining the discrete state-space coefficient matrix A of the system. d B d C d D d The discretized state-space model of the individual element includes a single-port model, an AC two-port model, and a DC two-port model. in: The iterative expression for the historical current term in the single-port model and its relationship with the unit output current are as follows: （1）； The iterative expression for the historical current term in the AC two-port model and its relationship with the unit output current are as follows: （2）； The iterative expression for the historical current term in the DC two-port model and its relationship with the unit output current are as follows: （3）； In the formula, h(t) represents the state variable, and A is the discrete state-space coefficient matrix of the system. d B d C d D d A diagonal block matrix composed of the coefficient matrices from the historical current sources of each element, where D d The coefficient matrix D of the historical current sources of each component d-s D d-L D d-h1 D d-h2 The diagonal block matrix is ​​formed, with subscripts ds, dL, d-h1, d-h2, d-h3, and d-h4 representing the coefficient matrices for different components; U xy and i xy The voltage and current at the model port are indicated by the subscripts sxy, Lxy, 1xy, and 2xy to distinguish different components.

8. The key area identification and power grid oscillation risk early warning device based on nodal mode voltage distribution as described in claim 6, characterized in that, The second matrix acquisition module is specifically used to: obtain the corresponding system admittance matrix G and branch-node correlation matrix Lt based on the discrete circuit network. （4）； （5）； The diagonal elements of the system admittance G matrix are G... ii Equal to the sum of the equivalent conductances of all elements connected to node i, and the off-diagonal element G ij It is equal to the negative of the equivalent conductance of the element directly connected between nodes i and j; in the branch-node correlation matrix L t In the matrix, the elements at the corresponding positions of the nodes connected to the single-port element are set to the identity matrix I2, the elements at the corresponding positions of the two nodes connected to the AC two-port element are set to I2 and -I2 respectively, the elements at the corresponding positions of the two nodes connected to the DC two-port element are both set to I2, and the elements at the remaining positions are all set to 0.

9. The key area identification and power grid oscillation risk early warning device based on nodal mode voltage distribution as described in claim 6, characterized in that, The state-mode relationship acquisition module is specifically used for: the discrete state matrix A of an AC system. D The calculation expression is: （6）； For discrete state matrix A D Perform diagonalization decomposition: （7）； In the formula, Λ is the discrete state matrix A D A is a diagonal matrix consisting of all its eigenvalues, where A D =A d +B d L t G -1 (-L t ) T D d ; V and W are the left and right eigenvector matrices corresponding to the discrete eigenvalues, respectively, satisfying V T W=I, where I is the identity matrix with the same dimension as the state matrix; By performing equivalent linear transformations on the state variables, we obtain the modal quantities: （8）； Combining equations (6) and (7), we obtain the discrete state-space equations of the system expressed in terms of modal quantities: （9）。 10. The key area identification and power grid oscillation risk early warning device based on nodal mode voltage distribution as described in claim 6, characterized in that, The early warning module is specifically used to: establish the relationship between node injection current and component discrete state variables h(t) in the discrete state space, where historical current sources are functions of component discrete state variables h(t). （10）； In the formula, the superscript 'T' denotes the transpose of the matrix, and D d h(t) is a diagonal block matrix composed of the coefficient matrices of the historical current sources of each element, and h(t) is the state vector of the whole system containing the discrete state variables of all elements. Establish the joint node conductance matrix G and the branch-node correlation matrix L t The relationship between system node voltages and discrete state variables is obtained: （11）； By combining the mathematical relationships between modal quantities and discrete state variables, we can obtain the relationship between system node voltages and modal quantities: （12）； In the formula, S is defined as the node voltage observability matrix of the system; The i-th column of matrix S represents the response of the i-th mode in the voltage at each node. Since the node voltages and branch currents are converted to a unified xy rotating coordinate system for the entire network when modeling the state space of the whole system, the x-axis components are superimposed with the y-axis components to obtain the voltage oscillation component of the i-th mode at node p. : （13）； In the formula, the letters in the parentheses represent the index numbers of the matrix elements; The node voltage distribution coefficient is defined as follows: （14）； Based on the calculation results of the voltage distribution coefficient of each node, the power grid section composed of nodes with relatively large amplitudes is designated as the main oscillation risk area of ​​the system. This helps to locate key areas and lines, enabling system operators to strengthen the monitoring of the local power grid and take relevant countermeasures at appropriate locations, thereby achieving early warning of power grid oscillation risks.

Images