A node inertia evaluation method in a high-proportion new energy power grid

Abstract

The application discloses a kind of node computing inertia evaluation methods in high proportion new energy power grid, comprising the following steps: S1, network reduction is carried out, and the synchronous torque coefficient between generator is determined;S2, based on the synchronous torque coefficient between generator, determine node computing inertia;S3, construct analytical calculation model for node computing inertia, complete analysis.The application can measure the local actual anti-disturbance capability of each electrical node in the network by establishing the sensitivity mapping relationship of active disturbance of load node to electromagnetic power transfer of all generators in the network, and fills the gap of dynamic frequency response spatial differentiation evaluation.

Description

Technical Field

[0001] This invention relates to the technical field, specifically to a method for evaluating the computational inertia of nodes in a high-proportion renewable energy power grid. Background Technology

[0002] With the advancement of new power system construction and the large-scale grid connection of new energy sources such as wind power and photovoltaics, the power grid exhibits a high proportion of power electronics. Traditional synchronous generator sets are gradually being replaced by new energy equipment connected to the grid via power electronic converter interfaces, resulting in a significant decrease in the overall rotational inertia of the system. Simultaneously, because new energy power plants are mostly distributed in remote areas, exhibiting a reverse distribution pattern, the originally concentrated inertia resources are severely fragmented, exhibiting significant spatiotemporal heterogeneity. Under the impedance of long-distance transmission lines, inertia voids with weak anti-interference capabilities are easily formed within the grid structure, drastically increasing the risk of local frequency exceedances. [1] Therefore, accurately assessing the local frequency support capability of each electrical node in a high-proportion renewable energy power grid and promptly identifying areas with weak inertia are crucial for ensuring the safe and stable operation of the power grid.

[0003] The existing assessment method treats the entire power grid as a highly synchronized single-point system, severely neglecting the attenuation and delay effects of electrical impedance on the frequency response signal during spatial propagation. This globally averaged assessment metric cannot accurately measure the actual local disturbance rejection capability of each electrical node within the network, making it difficult to accurately expose the true inertia weaknesses within the system. This can easily lead to the dispatch center underestimating the risk of local frequency overruns and instability.

[0004] The main drawbacks of existing technologies are as follows: Most existing evaluation methods focus on the overall equivalent inertia of the entire network, neglecting the fact that electrical impedance in renewable energy grids causes significant spatial attenuation and differences in inertia response. This inertia center model, which only focuses on the total inertia of the entire network, struggles to expose the true weaknesses in local grid structures, namely the inertia voids mentioned above. Furthermore, with the widespread integration of grid-connected converters, the power supply characteristics within the grid become more complex. Existing technologies struggle to incorporate the mechanical rotational physical inertia of traditional synchronous generators and the controlled virtual inertia of grid-connected converters into a unified quantitative mathematical model, and also cannot accurately assess the actual frequency support capabilities of each node by combining the underlying network electrical topology. [3] .

[0005] To achieve local or node-level inertia assessment, existing technology 2 introduces purely data-driven algorithms such as artificial intelligence and deep learning. These methods typically collect historical operating condition data, local frequency measurement information of nodes, and active power disturbances. By building a deep neural network, they mine and establish a nonlinear mapping relationship model between input measurement features and local inertia levels, thereby achieving local inertia prediction for specific regions or nodes.

[0006] Existing technology 2 heavily relies on the density of measurement devices and the quality of measured data. When faced with random fluctuations in new energy sources and extreme unseen operating conditions, the algorithm's generalization ability and robustness are poor. A more critical flaw is that the purely data-driven method lacks clear support from the physical mechanisms of the power system. It cannot incorporate the mechanical rotational inertia of traditional synchronous generators and the digitally controlled virtual inertia of grid-connected converters into a unified measurement scale, and cannot deeply reveal the mechanistic influence of the underlying network electrical topology on the spatiotemporal evolution distribution of inertia response. Summary of the Invention

[0007] To address the issues of low inertia and uneven spatiotemporal distribution of inertia caused by the grid connection of a large number of power electronic devices in high-proportion renewable energy power grids, this invention proposes a method for evaluating the node inertia in high-proportion renewable energy power grids.

[0008] The technical solution of this invention is: a method for evaluating the calculated inertia of nodes in a high-proportion renewable energy power grid, comprising the following steps:

[0009] S1. Perform network order reduction and determine the synchronous torque coefficient between generators;

[0010] S2. Determine the node inertia based on the synchronous torque coefficient between generators;

[0011] S3. Construct an analytical calculation model for the node inertia and complete the analysis.

[0012] Furthermore, S1 includes the following sub-steps:

[0013] S11. Divide all network nodes into a set of generator internal potential nodes and a set of load nodes, and construct the system block admittance matrix equation;

[0014] S12. Based on the system block admittance matrix equation, obtain the reduced-order network admittance matrix;

[0015] S13. Extract the elements of the reduced-order network admittance matrix to determine the electromagnetic power injected by the generator into the network;

[0016] S14. Based on the electromagnetic power injected into the network by the generator, determine the linearized state differential equation in matrix form;

[0017] S15. Determine the synchronous torque coefficient between generators based on the linearized state differential equation in matrix form.

[0018] Furthermore, in S11, the system block admittance matrix equation is:

[0019] ;

[0020] In the formula: for The vector of injected current into the internal potential node of the generator; for The voltage vector injected into the internal potential node of the generator; for The load node injected current vector; for The load node injected voltage vector; This is the block matrix of the first mutual admittance and self admittance between the corresponding node sets; This is the block matrix for the second mutual admittance and self admittance between corresponding node sets; This is the block matrix for the third mutual admittance and self-admittance between corresponding node sets; This is the fourth mutual admittance and self admittance block matrix between corresponding node sets;

[0021] In S13, the first Electromagnetic power injected into the network by the generator for:

[0022] ;

[0023] In the formula: For generator The amplitude of the internal potential during the transient process; For generator The amplitude of the internal potential during the transient process; These are the main diagonal conductance elements of the admittance matrix of the reduced-order network; For generator With generator Transfer conductance between; For generator With generator Transfer susceptance between; The phase difference is the electrical angle of the two rotors;

[0024] In S14, the linearized state differential equation in matrix form is:

[0025] ;

[0026] In the formula: This is the first diagonal matrix; It is the second diagonal matrix; For the rotor's angular velocity deviation, The rate of change of angular velocity; This refers to the deviation of the work angle; The deviation in mechanical power input to the prime mover; This is the electromagnetic power deviation vector output by the generator; This is the synchronous torque coefficient matrix;

[0027] In S15, the generator With generator Synchronous torque coefficient between for:

[0028] ;

[0029] In the formula: The initial power angle difference at the steady-state operating point. For generator The rotor power angle.

[0030] Furthermore, in S2, the nodal inertia H of node k is calculated. ck for:

[0031] ;

[0032] Where: H i For generator The rotor inertia time constant; When node k experiences an active power deficit, the generator... The initial electromagnetic power fraction extracted; This represents the number of generators.

[0033] Furthermore, S3 includes the following sub-steps:

[0034] S31. Construct the rotor motion equations of the synchronous machine in a per-unit system;

[0035] S32. Transform the rotor motion equation of the synchronous machine in the per-unit system into a complex frequency domain expression;

[0036] S33. Based on the complex frequency domain expression, determine the available energy reserves and the short-term maximum support power;

[0037] S34. Based on the available energy reserves and short-term maximum support power, construct an analytical calculation model for the node's computational inertia and complete the analysis.

[0038] Furthermore, in S31, the rotor motion equation of the synchronous machine in the per-unit system is:

[0039] ;

[0040] Where: H i For generator The rotor inertia time constant; This is the per-unit deviation of the rotor's electrical angular velocity; The deviation in mechanical power input to the prime mover; This refers to the deviation in electromagnetic power output by the generator. Damping coefficient provided for generator windings and load effects;

[0041] In S32, the rotor motion equation of the synchronous machine in the per-unit system is transformed into a complex frequency domain expression as follows:

[0042] ;

[0043] In the formula: s is the Laplace operator; This is the complex frequency domain form of the per-unit deviation of the rotor's electric angular velocity; The complex frequency domain form of the mechanical power deviation input to the prime mover; This is the complex frequency domain form of the electromagnetic power deviation of the generator output.

[0044] Furthermore, in S33, energy reserves can be utilized. With short-term maximum support power They are respectively:

[0045] ;

[0046] ;

[0047] Where: H i For generator The rotor inertia time constant; This refers to the rated base capacity of the unit; This is the maximum transient active overload factor set for the unit in the simulation model.

[0048] Furthermore, in S34, the analytical calculation model is as follows:

[0049] ;

[0050] In the formula: When node k experiences an active power deficit, the generator... The initial electromagnetic power fraction extracted; When node k experiences an active power deficit, it switches from the generator. The initial electromagnetic power fraction extracted; H i For generator The rotor inertia time constant; The equivalent virtual inertia of a grid-type wind turbine; S G It is a traditional synchronous generator; S W It is a grid-type wind turbine.

[0051] The beneficial effects of this invention are:

[0052] (1) This invention breaks through the limitations of traditional methods for evaluating the equivalent inertia of the entire network. Traditional methods usually treat the system as a single, highly synchronized whole, relying excessively on a uniform equivalent inertia index for the entire network, which can easily mask the real frequency exceedance risk in local weak network areas. This invention establishes a sensitivity mapping relationship between the active power disturbance of load nodes and the electromagnetic power transfer of generators across the entire network, which can accurately measure the actual local disturbance rejection capability of each electrical node in the network, filling the gap in the spatial differentiation evaluation of dynamic frequency response.

[0053] (2) This invention achieves unified quantification of the frequency support capability of heterogeneous power sources. Existing technologies struggle to incorporate the mechanical rotational physical inertia of traditional synchronous generators with the digitally controlled virtual inertia of new energy equipment into the same measurement scale. This invention, by extracting the control setting parameters of grid-connected converters and other equipment and deriving their equivalent virtual inertia time constants, successfully integrates traditional mechanical kinetic energy with the controlled virtual energy of power electronics, constructing a comprehensive inertia evaluation mechanism that spans different power source characteristics.

[0054] (3) This invention improves the identification and security defense capabilities of local high-risk instability. Because this method comprehensively considers the mechanism of the influence of the underlying network electrical topology on the spatial evolution distribution of inertia response, the scheduling center can accurately discover and identify hidden weak inertia areas and inertia voids in the system, thus providing crucial theoretical support and model foundation for forward-looking dynamic frequency characteristic early warning and the formulation of differentiated active defense strategies. Attached Figure Description

[0055] Figure 1 A flowchart illustrating the method for evaluating the inertia of nodes in a high-proportion renewable energy power grid;

[0056] Figure 2 A diagram illustrating the spatial coupling mechanism between physical inertia distribution and nodal frequency response;

[0057] Figure 3 A comparison diagram of the physical rotor and the virtual inertia control mechanism of synchronous machines / fans;

[0058] Figure 4 This is a topology diagram of the WSCC 3-machine 9-node system;

[0059] Figure 5 The graph shows a comparison between the theoretical and simulated values ​​of the frequency response of node 5 under disturbance.

[0060] Figure 6 A diagram showing the spatial distribution characteristics of the computed inertia of the nodes in the WSCC 9-node system;

[0061] Figure 7 The graph shows the sensitivity verification of the rate of change of frequency to the inertial constant H.

[0062] Figure 8 For inertia support capability with respect to electrical distance X line Sensitivity verification diagram;

[0063] Figure 9 Sensitivity analysis diagram of node frequency response characteristics to changes in virtual inertia parameters;

[0064] Figure 10 A topology diagram of an improved IEEE 39-node system for a grid-type wind farm;

[0065] Figure 11 The heat map of the electromechanical coupling strength matrix between units;

[0066] Figure 12 Calculate the inertia H for all 39 nodes in the network. ck Schematic diagram;

[0067] Figure 13 A comparison chart of the frequency response characteristics of node 39 and synchronous machine node 31. Detailed Implementation

[0068] The embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0069] like Figure 1 As shown, this invention provides a method for evaluating the computational inertia of nodes in a high-proportion renewable energy power grid, comprising the following steps:

[0070] S1. Perform network order reduction and determine the synchronous torque coefficient between generators;

[0071] S2. Determine the node inertia based on the synchronous torque coefficient between generators;

[0072] S3. Construct an analytical calculation model for the node inertia and complete the analysis.

[0073] The technical problems to be solved by this invention are mainly reflected in two aspects. First, traditional inertia assessment methods usually treat the system as a single, highly synchronized whole, relying excessively on the overall equivalent inertia index of the entire network. This conventional model ignores the spatial attenuation and spatiotemporal heterogeneity of inertia response caused by electrical impedance, and cannot accurately measure the local actual disturbance rejection capability of each electrical node. Second, existing technologies lack an effective mathematical model to uniformly quantify the mechanical physical inertia of traditional synchronous generators and the controlled virtual inertia of new energy equipment, resulting in the inability to accurately assess the actual local frequency support capability acquired by each node, and making it difficult to expose the real inertia voids within the system. To solve the above technical problems, the purpose of this invention is to provide a node inertia assessment method for high-proportion new energy power grids. This method constructs a node comprehensive computational inertia assessment model that takes into account network transmission impedance sensitivity and heterogeneous power supply characteristics, unifying physical mechanical inertia and controlled virtual inertia into a node-level quantification system, accurately identifying hidden weak areas of inertia within the system, and providing a scientific basis for frequency security defense of high-proportion new energy power grids.

[0074] In this embodiment of the invention, S1 includes the following sub-steps:

[0075] S11. Divide all network nodes into a set of generator internal potential nodes and a set of load nodes, and construct the system block admittance matrix equation;

[0076] S12. Based on the system block admittance matrix equation, obtain the reduced-order network admittance matrix;

[0077] S13. Extract the elements of the reduced-order network admittance matrix to determine the electromagnetic power injected by the generator into the network;

[0078] S14. Based on the electromagnetic power injected into the network by the generator, determine the linearized state differential equation in matrix form;

[0079] S15. Determine the synchronous torque coefficient between generators based on the linearized state differential equation in matrix form.

[0080] In this embodiment of the invention, in S11, the system block admittance matrix equation is:

[0081] ;

[0082] In the formula: for The vector of injected current into the internal potential node of the generator; for The voltage vector injected into the internal potential node of the generator; for The load node injected current vector; for The load node injected voltage vector; This is the block matrix of the first mutual admittance and self admittance between the corresponding node sets; This is the block matrix for the second mutual admittance and self admittance between corresponding node sets; This is the block matrix for the third mutual admittance and self-admittance between corresponding node sets; This is the fourth mutual admittance and self admittance block matrix between corresponding node sets;

[0083] In S13, the first Electromagnetic power injected into the network by the generator for:

[0084] ;

[0085] In the formula: For generator The amplitude of the internal potential during the transient process; For generator The amplitude of the internal potential during the transient process; These are the main diagonal conductance elements of the admittance matrix of the reduced-order network; For generator With generator Transfer conductance between; For generator With generator Transfer susceptance between; The phase difference of the electrical angles of the two rotors; satisfying .

[0086] In S14, the linearized state differential equation in matrix form is:

[0087] ;

[0088] In the formula: This is the first diagonal matrix; It is the second diagonal matrix; For the rotor's angular velocity deviation, The rate of change of angular velocity; This refers to the deviation of the work angle; The deviation in mechanical power input to the prime mover; This is the electromagnetic power deviation vector output by the generator; This is the synchronous torque coefficient matrix;

[0089] and All are diagonal matrices, with their diagonal elements being the inertial constants and damping coefficients of all generators in the network.

[0090] In S15, the generator With generator Synchronous torque coefficient between for:

[0091] ;

[0092] In the formula: The initial power angle difference at the steady-state operating point. For generator The rotor power angle.

[0093] In this embodiment of the invention, in S2, the node inertia H of node k is calculated. ck for:

[0094] ;

[0095] Where: H i For generator The rotor inertia time constant; When node k experiences an active power deficit, the generator... The initial electromagnetic power fraction extracted; This represents the number of generators.

[0096] In this embodiment of the invention, S3 includes the following sub-steps:

[0097] S31. Construct the rotor motion equations of the synchronous machine in a per-unit system;

[0098] S32. Transform the rotor motion equation of the synchronous machine in the per-unit system into a complex frequency domain expression;

[0099] S33. Based on the complex frequency domain expression, determine the available energy reserves and the short-term maximum support power;

[0100] S34. Based on the available energy reserves and short-term maximum support power, construct an analytical calculation model for the node's computational inertia and complete the analysis.

[0101] In this embodiment of the invention, in S31, the rotor motion equation of the synchronous machine in the per-unit system is:

[0102] ;

[0103] Where: H i For generator The rotor inertia time constant; This is the per-unit deviation of the rotor's electrical angular velocity; The deviation in mechanical power input to the prime mover; This refers to the deviation in electromagnetic power output by the generator. Damping coefficient provided for generator windings and load effects;

[0104] In S32, the rotor motion equation of the synchronous machine in the per-unit system is transformed into a complex frequency domain expression as follows:

[0105] ;

[0106] In the formula: s is the Laplace operator; This is the complex frequency domain form of the per-unit deviation of the rotor's electric angular velocity; The complex frequency domain form of the mechanical power deviation input to the prime mover; This is the complex frequency domain form of the electromagnetic power deviation of the generator output.

[0107] In this embodiment of the invention, in S33, the energy reserve can be called up. With short-term maximum support power They are respectively:

[0108] ;

[0109] ;

[0110] Where: H i For generator The rotor inertia time constant; This refers to the rated base capacity of the unit; This is the maximum transient active overload factor set for the unit in the simulation model.

[0111] In this embodiment of the invention, in S34, the analytical calculation model is:

[0112] ;

[0113] In the formula: When node k experiences an active power deficit, the generator... The initial electromagnetic power fraction extracted; When node k experiences an active power deficit, it switches from the generator. The initial electromagnetic power fraction extracted; H i For generator The rotor inertia time constant; The equivalent virtual inertia of a grid-type wind turbine; S G It is a traditional synchronous generator; S W It is a grid-type wind turbine.

[0114] In this embodiment of the invention, frequency fluctuations occur when an active power imbalance occurs in the power system. The essence of this phenomenon is that generator rotors within the grid compensate for the instantaneous power deficit by releasing or absorbing kinetic energy. To quantify and analyze the computational inertia of nodes, this invention derives the frequency interaction process within a multi-machine system. Specifically, firstly, a network order reduction method is used to eliminate passive load nodes that lack kinetic energy, retaining only generator nodes. Subsequently, a sensitivity analytical operator is introduced to quantitatively calculate the tightness of electrical connections between grid nodes using formulas, and the specific impact of this grid connection on the spatial transfer and distribution of generator kinetic energy across the entire network is analyzed.

[0115] Unlike the traditional simplified model of a single-unit infinitely large grid, in actual high-proportion renewable energy power grids, the dynamic response of each generating unit is constrained by the transmission network structure. Therefore, the frequency variation characteristics of any node in the network are not solely determined by the physical inertia of the local generating unit. Due to close electrical connections, when a local disturbance occurs, the distant generator group can transmit active power to the affected area through the transmission network and transient synchronous torque, thus jointly participating in the frequency support process of that node. The physical mechanism of this spatial transfer and support of generator rotor kinetic energy through the power grid is as follows: Figure 2 As shown.

[0116] Based on Kirchhoff's current law, the entire network nodes are divided into a set of generator internal potential nodes and a set of load nodes, and the system block admittance matrix equation is constructed.

[0117] Since the generator transient reactance is included in the network equivalent impedance, and the constant impedance load is incorporated into the diagonal elements of the self-admittance matrix, the equivalent injected current of the load node is a zero vector. By performing Gaussian elimination on the above matrix using the Kronoswiss simplification algorithm to eliminate passive load nodes, a reduced-order network admittance matrix retaining the generator core node can be derived. :

[0118] ;

[0119] In the formula: After removing load nodes Reduced-order complex admittance matrix of the generator node.

[0120] Extracting the matrix Internal elements , No. Electromagnetic power injected into the network by the generator It can be analytically expressed by the voltage amplitude and power angle of the network generator nodes.

[0121] When the power grid is operating under small disturbances, to facilitate solving the dynamic response of each generating unit, the swing equations of the multi-generator system are typically expanded using Taylor series near the steady-state equilibrium point. During the expansion, higher-order nonlinear terms are discarded, retaining only the first-order linear part. After this approximation, for an interconnected system containing n generating units, a linearized state differential equation in matrix form is derived.

[0122] In the differential equations constructed above, off-diagonal elements in a matrix This directly reflects the tightness of the electrical connection between any two generator nodes. To further quantify this spatial electrical coupling relationship, the analytical partial derivative of this element is calculated as follows:

[0123] ;

[0124] element The larger the absolute value, the shorter the electrical distance and the tighter the frequency coupling between the two units.

[0125] To define the disturbance rejection properties of any node in the network, a physical mapping relationship needs to be established between the local active power surge and the node's initial frequency response. At the transient instant of active power imbalance, the system frequency and rotor power angle have not yet undergone significant shifts, therefore... and At this time, the speed regulating valve of the prime mover has not yet been activated, and the mechanical power input remains constant. Sudden active power deficits in the network are entirely filled by the instantaneous release of rotor kinetic energy from all generator units in the network. To accurately quantify the specific power support share of each generator in the network when an active power disturbance occurs at load node k, this invention directly derives the equation from the block form of the system's steady-state Jacobian matrix:

[0126] ;

[0127] In the formula: and These are the active power injection deviation vectors for the generator and the load node, respectively. This is the generator's internal electromotive force angle deviation vector; This is the load node voltage phase angle deviation vector; , , , It is a sub-block in the steady-state Jacobian matrix consisting of the partial derivatives of active power with respect to phase angle.

[0128] exist Instantly, the generator rotor power angle cannot change abruptly due to inertia. Substituting into the above matrix equation eliminates the internal state variables. The mapping relationship between load disturbance and electromagnetic power transfer from generator set can be derived:

[0129] ;

[0130] Therefore, the inertia distribution coefficient matrix is ​​defined. Combining the initial motion equations of a single machine Generator i responds to step disturbance at node k The initial frequency change rate at time is:

[0131] ;

[0132] In the formula: elements The initial electromagnetic power share extracted from the i-th generator when node k experiences an active power deficit. The step active power disturbance experienced by node k; Let be the initial frequency change rate of generator i.

[0133] Based on the above sensitivity matrix derivation, the nodal inertia H is established. ck The mathematical definition of is: at the instant that the target node k encounters a step active disturbance, its overall ability to resist the decay of the local frequency change rate is equivalent to a virtual centralized inertial time constant. Its defining equation is:

[0134] ;

[0135] Where: H ck Calculate the inertia of node k; Let be the initial local frequency change rate measured at node k at the instant of the disturbance.

[0136] Combining the distribution coefficient matrix The equation for calculating node inertia based on analytical calculation of network parameters is derived:

[0137] ;

[0138] Where: H i Let be the rotor inertia time constant of the i-th generator set.

[0139] This formula shows that the equivalent computational inertia of a single node is essentially the weighted sum of the physical kinetic energy of all units in the network projected onto that node based on the electrical topology distance.

[0140] To explain the physical meaning of the above derivation and to explore the influence of system parameters on local frequency characteristics, this invention performs analytical derivation of the partial derivatives of key parameters for the rocking equation system. The initial frequency change rate of the system is defined. Dynamic sensitivity function relative to the target parameter x for:

[0141] ;

[0142] In the formula: The relative sensitivity index of the target parameter; The steady-state baseline value of the target parameter.

[0143] First, we analyze the degree to which the unit's own mass parameters affect the transient frequency. In the initial stage of the disturbance response, we approximate the expression for the rate of change of the node's initial frequency with respect to the generator's inertial constant. Performing first-order partial derivative operations, we obtain:

[0144] ;

[0145] The above equation shows that the rate of change of frequency affects the inertia of a single-machine system. The sensitivity follows the inertia calculated from the nodes. The relationship is an inverse square law with the base case. When the node is in a low inertia state, increasing physical kinetic energy has a more significant effect on improving the local frequency safety margin; however, when the node is in a high inertia state, the improvement effect of continuing to increase kinetic energy reserves will gradually weaken.

[0146] Secondly, the physical isolation effect of the transmission network impedance is analyzed. For any generator node in the network... and Interconnection transmission reactance Its off-diagonal elements of the synchronous torque coefficient matrix The partial derivative mapping satisfies:

[0147] ;

[0148] In the formula: This refers to the equivalent interconnection transmission reactance between generator nodes.

[0149] The fact that this partial derivative is always negative indicates that the conduction of supporting power is limited by electrical distance. With the reactance of the inter-regional tie line... As the voltage increases, the sensitivity of remote generators to disturbance power deficits decreases rapidly. Due to the weakening effect of network transmission impedance, high-proportion renewable energy power grids are prone to insufficient inertia support around renewable energy grid-connected areas. This theoretical derivation verifies the rationality of combining network topology impedance to measure node inertia calculation in this invention, providing a theoretical basis for subsequently introducing a reduced-order algorithm to implement dynamic grid zoning.

[0150] With the large-scale grid connection of new energy sources, the power structure of the power grid is gradually shifting from being dominated by traditional synchronous generators to a mixed structure of multiple power sources. Traditional synchronous generators rely on mechanical rotors to provide physical rotational inertia. Among various new energy devices, photovoltaic power generation systems are purely static components, lacking mechanical rotating parts. Without additional large-scale energy storage, they typically cannot provide substantial transient support power. Wind turbine generators, on the other hand, possess actual physical rotors and, through the control algorithms within the converter, can simulate and release support power to participate in the frequency response. Therefore, to accurately assess the frequency stability of a high-proportion new energy power grid, this invention uses grid-connected wind turbine generators as a typical representative of new energy sources providing virtual inertia. By eliminating the differences in the underlying response mechanisms between the physical rotor and the controlled converter, a unified and quantifiable comprehensive node inertia calculation model for local support power is established.

[0151] The physical response characteristics of synchronous generators are a traditional benchmark for defining the inertia of power systems. To achieve uniformity in dimensions and parameter systems with converter control models, this invention first standardizes the continuous-time rotor swing equations of the synchronous generator. For the first... The rotor motion equation of a synchronous machine can be described in per-unit terms as follows:

[0152] ;

[0153] In the formula: The inherent inertial time constant of the unit represents the ratio of the kinetic energy stored by the rotor at rated speed to its rated capacity. This is the per-unit deviation of the rotor's electrical angular velocity; and These are the deviations in mechanical power input to the prime mover and the deviations in electromagnetic power output to the generator, respectively. The damping coefficient provided for generator windings and load effects.

[0154] In practical digital control analysis or frequency domain stability assessment, the above differential equation is usually transformed into a complex frequency domain expression using a Laplace transform. The impedance of a synchronous generator to frequency changes can be equivalently represented as an integrator with low-pass filtering characteristics:

[0155] ;

[0156] In the formula: s is the Laplace operator; , , These are the complex frequency domain forms of the corresponding time-domain physical quantities.

[0157] From a physical perspective, the denominator contains This term constitutes a natural inertial delay element, which can effectively filter out high-frequency active power fluctuations in the initial stage of system disturbance, thereby maintaining a smooth transition between unit speed and grid frequency. This inherent physical response capability serves as a benchmark for deriving the equivalent parameters of virtual inertia of new energy sources.

[0158] Unlike traditional grid-based control, which treats the converter merely as a controlled current source, grid-based control simulates the operating characteristics of a synchronous generator, enabling the converter to autonomously establish voltage amplitude and phase. Based on the power supply characteristic analysis above, the rotor and impeller of a wind turbine contain real mechanical rotational energy. By adjusting the control strategy within the converter, this energy can be directly utilized to simulate inertial response, achieving a high degree of physical similarity to a traditional synchronous generator, thereby effectively improving the transient frequency characteristics of the local grid structure.

[0159] Taking a doubly-fed or direct-drive wind turbine generator employing grid-type control as an example, the core of its control strategy is to embed the rotor swing equation of a traditional synchronous machine into the active-frequency control loop of the converter. Therefore, the equivalent control equation for the unit during transient regulation can be expressed as:

[0160] ;

[0161] In the formula: The virtual moment of inertia set for the control algorithm; The rated electrical angular frequency of the power grid; Virtual angular velocity generated digitally inside the converter; and These are the active power reference value set for the converter and the actual port measurement value, respectively. The virtual damping coefficient set for the control loop.

[0162] To achieve dimensional unification with the physical inertia of the synchronous generator, a conversion relationship between digital control parameters and per-unit inertial time constants needs to be established. According to the classical definition of the inertial time constant, this constant is equal to the ratio of the kinetic energy stored by the rotor at synchronous speed to the rated apparent power of the unit. Referring to this physical definition, it is extended to the virtual synchronous control loop of the converter, where the equivalent virtual kinetic energy simulated by the converter... It can be represented as:

[0163] ;

[0164] Furthermore, the equivalent virtual inertial time constant of the grid-type converter is derived. :

[0165] ;

[0166] In the formula: The equivalent virtual inertial time constant for a grid-type wind turbine; This is the rated apparent power capacity of the converter equipment.

[0167] As the above derivation shows, the virtual inertia provided by a grid-connected converter is essentially an inertial response characteristic simulated by a digital control algorithm. The physical rotational inertia of a traditional synchronous generator is limited by the inherent mechanical structure of the unit and its value is a constant; while the equivalent virtual inertia of a wind turbine is controlled by an internal algorithm, possessing dynamic adjustability, and can adapt to the transient support requirements of the power grid under different operating conditions by changing control parameters. A comparison of the physical response mechanism of a synchronous generator and the control logic of a grid-connected converter is provided. Figure 3 As shown.

[0168] Combination Figure 3 As can be seen from the control response process, grid-connected converters can simulate the inertial response of synchronous generators through algorithms. However, this equivalent capability is limited by hardware conditions; it is only effective when the energy stored on the DC side is sufficient and the AC side current is within a safe operating range. Therefore, when a severe grid fault causes an excessive power deficit, the supporting power provided by grid-connected units will be limited by the maximum overcurrent limit of the equipment and the energy storage capacity, and cannot maintain unlimited compensation.

[0169] To enable subsequent quantitative assessment of the actual frequency support levels of various power supplies, the available energy reserves are provided here. With short-term maximum support power The general calculation formula.

[0170] For traditional synchronous generators, their physical construction allows for... Typically, it can reach over 2.5 pu; however, for grid-type converters, due to the thermal stability limits of power electronic devices, its... It is usually hard-limited to around 1.2 pu.

[0171] Based on the above modeling process, inertia sources in the new energy power grid can be divided into two categories: S G Traditional synchronous generators and S W In the context of grid-connected wind turbines, for any target node k in the renewable energy grid, the total inertia support at that location no longer depends solely on the local physical rotor, but rather on the spatial superposition of the entire network's physical inertia and the controlled virtual inertia.

[0172] Combined with the derived inertia distribution coefficient matrix based on network topology reduction This invention proposes a generalized nodal computation inertia. Analytical calculation model:

[0173] ;

[0174] In the formula: and Synchronous generators With grid-type wind turbine To the target node Inertia distribution weighting coefficient; The physical inertia of the synchronous generator; The equivalent virtual inertia of a grid-type wind turbine.

[0175] In summary, this evaluation model reveals the key factors for local frequency stability in high-proportion renewable energy power grids. The frequency performance of nodes is determined by three factors: first, the operating status and rotational inertia of synchronous generator units; second, the connection location of grid-connected converters and their virtual inertia control parameters; and third, the spatial electrical coupling strength determined by the grid topology. These three factors interact and jointly determine the frequency response process of local nodes. This model incorporates physical and virtual inertia into a unified mathematical model, eliminating the differences in evaluation attributes between different types of power sources, and providing a reliable physical basis for subsequent chapters on dynamic grid zoning and frequency characteristic prediction.

[0176] A standard electromechanical transient simulation model of a WSCC 3-machine 9-node system was built in PSS / E simulation software. The system's base power was selected as 100 MVA, and the base frequency as 50 Hz. The key electrical and transient parameters of the generator sets in the network are configured as shown in Table 1.

[0177] Table 1

[0178] Unit number Access Node <![CDATA[Rated apparent power S base (MVA)]]> Inertial time constant H(s) <![CDATA[Transient reactance X ’ d (p.u.)]]> G1 Node1 247.5 5.148 0.0608 G2 Node2 192.0 4.000 0.1198 G3 Node3 128.0 3.010 0.1813

[0179] After the model is built based on the above parameter configuration, the network topology of the simulation system is as follows: Figure 4 As shown.

[0180] Small-scale test system model verification and sensitivity analysis: To verify the actual evaluation error of the node's calculated inertia, the simulation was set at t=1.0s, introducing an active power step disturbance with an amplitude of 0.1pu for load node 5. First, the theoretical value of the node's calculated inertia was calculated, and the theoretical initial frequency change rate of the target node was determined. Subsequently, the theoretical trajectory calculated analytically was compared with the dynamic frequency response waveform output by the actual PSS / E simulation. The comparison curves are shown below. Figure 5 As shown.

[0181] like Figure 5It can be seen that the calculated theoretical initial slope basically matches the actual simulated frequency drop trajectory. Taking the load node 5 at the center of the system as an example, the theoretical value of the node's calculated inertia is 2.10s, while the result derived from the actual PSS / E simulation waveform is 2.06s. The relative error between the two is small, which preliminarily verifies the accuracy of the proposed evaluation method and shows that the model can reflect the frequency response characteristics of the node at the instant of disturbance.

[0182] To clarify the spatial distribution pattern of disturbance rejection capability, the computational inertia of each node under steady-state conditions was calculated using the WSCC 3-machine 9-node system. The spatial distribution characteristics are shown in the heatmap below. Figure 6 As shown.

[0183] from Figure 6 The heatmap showing the distribution characteristics reveals significant differences in disturbance immunity levels at different locations within the same power grid. Near generator node 1, a typical high-inertia support zone is formed due to ample physical rotational inertia, with the node's calculated inertia reaching 8.5s. However, with the increase in spatial electrical distance and the obstruction of line impedance, the power supported at the far end is difficult to conduct without loss. This results in a localized inertia void with extremely poor disturbance immunity around the central load node 5, far from the power source, with the value decaying all the way down to 2.1s. This obvious spatial decay phenomenon directly confirms that if the traditional average parameters of the entire network are continued to be used to evaluate the system, it will be impossible to detect hidden frequency security risks in high-proportion renewable energy power grids.

[0184] Based on the derived partial derivatives, to verify the tracking capability of the node inertia calculation model for changes in physical parameters, this invention verifies the sensitivity of the initial frequency change rate relative to the unit's inertial time constant. Under the condition of constant system topology and disturbance power, the initial frequency change rate response value of the target node is obtained by scanning the inherent inertial time constant of the synchronous generator unit. The theoretical curve calculated by the analytical model is compared with the measured scatter points extracted from the actual simulation, and the results are as follows. Figure 7 As shown.

[0185] analyze Figure 7 It can be seen that the theoretical calculation curve and the scatter plot of the simulation example agree well, verifying the physical law that the local initial frequency change rate is inversely proportional to the inertia level. The trend of the curve shows that when the inertia level decreases to a certain range, the frequency drop rate increases sharply, indicating that the system's disturbance rejection capability has a clear safety critical region.

[0186] Besides the physical parameters of the generator itself, the grid topology is also a key factor affecting the local inertia support capability. Based on the analytical conclusions of the partial derivatives of the synchronous torque coefficient, to verify the weakening effect of network impedance on the spatial transmission of power to distant supports, the change process of electrical coupling strength between nodes was analyzed by gradually increasing the transmission reactance of critical tie lines. Specific results are as follows: Figure 8 As shown.

[0187] like Figure 8 As shown, with the continuous increase of the equivalent transmission reactance between the two nodes, the system operating state clearly transitions from a strongly coupled support region to a local inertia void region. This change process indicates that the increase in network impedance will severely weaken the spatial transmission capability of the far-end physical kinetic energy, causing the disturbed node to be unable to obtain sufficient active power support in a timely manner, ultimately leading to a significant increase in its initial frequency change rate.

[0188] To verify the correctness of the virtual inertia equivalent model of the grid-type converter derived above, this invention built an electromechanical transient simulation model of the IEEE 39-bus system in PSS / E simulation software. The equivalent capacity of the traditional synchronous generator at node 34 was replaced with a grid-type wind farm. The core control parameters of this grid-type converter are configured as shown in Table 2.

[0189] Table 2

[0190] Parameter name Symbolic representation Numerical settings Rated apparent power <![CDATA[S N,vir ]]> 1000.0MVA Rated phase angle frequency <![CDATA[ω0]]> 314.15 rad / s Virtual damping coefficient <![CDATA[K d,vir ]]> 1.0-2.0pu Virtual inertial time constant <![CDATA[H vir ]]> 2.0-8.0s

[0191] Based on the core parameter configuration of the above model, an active step disturbance with an amplitude of 0.1 pu is applied to system node 39. This is achieved by adjusting the virtual inertial time constant of the grid-type converter. Record the frequency response trajectory corresponding to node 34, specifically as follows: Figure 9 As shown.

[0192] Depend on Figure 9 It can be seen that when the virtual inertia control parameter set by the grid-type converter increases from 2s to 8s, the absolute value of the initial frequency change rate of local nodes during the initial stage of active step disturbance shows a gradual decreasing trend, while the extreme value of transient frequency drop also increases accordingly. This phenomenon indicates that the transient energy released by the grid-type wind farm through virtual control effectively improves the dynamic frequency response characteristics of the locally vulnerable grid structure. This demonstrates that the heterogeneous power source unified quantization model constructed in this invention can accurately map the digital control logic inside the converter into equivalent physical parameters that can be uniformly evaluated across the entire network.

[0193] After verifying the equivalent model of a single converter, further analysis was conducted under a complex grid structure with multiple grid-connected wind farms to construct a typical high-proportion renewable energy grid simulation scenario. This invention replaces the traditional synchronous generators of the core power nodes G8 (node ​​37), G9 (node ​​38), and G10 (node ​​39) that play a major supporting role in the original system with grid-connected wind farms of equal capacity. After the replacement, the total installed capacity of these three renewable energy units reaches 3240 MVA, increasing the renewable energy capacity penetration rate of the entire grid to over 50%. Simultaneously, the high proportion of renewable energy capacity is highly concentrated in the edge region of the grid, resulting in a highly uneven spatial distribution of system inertia and forming significant local weak inertia characteristics. The modified high-proportion renewable energy grid topology is as follows: Figure 10 As shown.

[0194] To assess the coupling strength between units in the network, this invention extracts the potential amplitude and initial power angle difference within the units under steady state, and combines this with the inter-node transfer conductance and transfer susceptance obtained after network order reduction. Based on the derived partial derivatives of the synchronous torque coefficient, the coupling strength of all 10 units in the network is calculated. The quantized matrix data is shown in Table 3.

[0195] Table 3

[0196] Unit G1(39W) G8(37W) G9 (38W) G2(31) G3(32) G4(33) G5(34) G6(35) G7(36) G10(30) G1(39W) 1.000 0.124 0.089 0.045 0.032 0.021 0.028 0.019 0.015 0.008 G8(37W) 0.124 1.000 0.356 0.089 0.067 0.054 0.061 0.042 0.038 0.012 G9(38W) 0.089 0.356 1.000 0.076 0.058 0.049 0.055 0.039 0.031 0.010 G2(31) 0.045 0.089 0.076 1.000 0.654 0.432 0.389 0.210 0.156 0.145 G3(32) 0.032 0.067 0.058 0.654 1.000 0.512 0.410 0.234 0.167 0.156 G4(33) 0.021 0.054 0.049 0.432 0.512 1.000 0.567 0.345 0.210 0.123 G5(34) 0.028 0.061 0550 0.389 0.410 0.567 1.000 0.456 0.289 0.110 G6(35) 0.019 0.042 0.039 0.210 0.234 0.345 0.456 1.000 0.589 0.098 G7(36) 0.015 0.038 0.031 0.156 0.167 0.210 0.289 0.589 1.000 0.087 G10(30) 0.008 0.012 0.010 0.145 0.156 0.123 0.110 0.098 0.087 1.000

[0197] According to the quantitative data in Table 3, the coupling coefficient between nodes G1 and G10 is only 0.008. This value indicates that the corresponding wind power cluster is relatively isolated in terms of electrical topology, with weak mutual influence with nodes in other areas. Conversely, the coupling coefficients within conventional synchronous generator clusters such as G2 and G3 are generally higher, indicating close electrical connections between the nodes. To visually represent the spatial distribution characteristics of the electrical coupling strength among all network nodes, the coupling matrix calculated above is mapped to a visual heatmap, as shown below. Figure 11 As shown.

[0198] Depend on Figure 11 The visualized heatmap reveals that the electromechanical coupling matrix of the entire grid exhibits a block-like spatial distribution. The strongly coupled regions corresponding to traditional synchronous generators generally have higher values; for example, the coupling coefficient between adjacent G2 and G3 reaches 0.654, and between G6 and G7 it reaches 0.589. This indicates that the physical units within the backbone grid have close electrical connections and strong transient energy exchange capabilities. Conversely, the values ​​of the grid-connected wind power nodes located at the matrix edges are generally lower, with their coupling coefficients with the core synchronous generators only around 0.05, and the coupling between some distant nodes even below 0.01. This numerical difference reflects the significant reduction in the ability of edge wind power nodes to obtain kinetic energy support from the backbone grid in the early stages of disturbance due to the influence of actual electrical distance and the impedance of long-distance transmission networks.

[0199] With a high proportion of renewable energy access, accurately quantifying the dynamic support capabilities of each node in the entire network is a prerequisite for conducting regional studies. Considering that the IEEE 39-node system is a reduced-order equivalent of a complex actual network structure, this paper adopts a large-capacity aggregation cluster model to restore the actual callable energy reserves of the regional coherent cluster, thus aligning with the engineering realities of centralized grid connection in my country's large-scale renewable energy bases. The specific electrical and inertia parameter configurations of each aggregation cluster within the system are shown in Table 4.

[0200] Table 4

[0201] Unit number Access Node Unit type <![CDATA[rated reference capacity S base (MVA)]]> Inertial time constant H(s) Damping coefficient D(pu) G1 30 Synchronous generator 2500.0 2.0 2.0 G2 31 Synchronous generator 2500.0 2.05 2.0 G3 32 Synchronous generator 2500.0 1.94 2.0 G4 33 Synchronous generator 2500.0 1.85 2.0 G5 34 Synchronous generator 2500.0 1.82 2.0 G6 35 Synchronous generator 2500.0 1.91 2.0 G7 36 Synchronous generator 2500.0 1.96 2.0 G8 37 Grid-type wind power 960.0 0.79 1.0 G9 38 Grid-type wind power 1080.0 0.69 1.0 G10 39 Grid-type wind power 1200.0 0.76 1.0

[0202] Based on the aforementioned unit fundamental parameters and the electromechanical coupling matrix determined earlier, the initial disturbance immunity attributes of all 39 nodes in the network were calculated. This process maps the physical rotational inertia discrete at each generator node and the controlled virtual inertia of the grid-type converter to the actual support capacity of each electrical node in the network against active power disturbances, according to the network topology. Detailed data on the calculated inertia of all nodes in the improved IEEE 39-node system are shown in Table 5.

[0203] Table 5

[0204] Node number type <![CDATA[H ck (s)]]> Node number type <![CDATA[H ck (s)]]> Node number type <![CDATA[H ck (s)]]> 1 Load 0.952 14 Load 2.560 27 Load 2.120 2 Load 1.120 15 Load 2.780 28 Load 4.100 3 Load 1.450 16 Load 2.450 29 Load 1.890 4 Load 2.120 17 Load 2.330 30 Load 4.200 5 Load 3.210 18 Load 2.890 31 Gen 5.120 6 Load 4.120 19 Load 3.120 32 Gen 4.850 7 Load 3.890 20 Load 4.010 33 Gen 4.620 8 Load 3.560 21 Load 3.450 34 Gen 4.550 9 Load 1.050 22 Load 4.560 35 Gen 4.780 10 Load 2.450 23 Load 4.230 36 Gen 4.910 11 Load 2.105 24 Load 2.670 37 Wind 0.754 12 Load 2.010 25 Load 1.560 38 Wind 0.825 13 Load 1.980 26 Load 2.890 39 Wind 0.912

[0205] Analysis of Table 5 shows that the spatial distribution of computational inertia of nodes in high-proportion renewable energy power grids is extremely uneven. The inertia of nodes 37, 38, and 39 are only 0.754s, 0.825s, and 0.912s, respectively, far lower than that of traditional synchronous machines. Due to network impedance constraints, their neighboring load nodes are also only in the 1.0~2.0s range, and after disturbances, they will face a large initial frequency change rate due to a lack of kinetic energy support. In contrast, the synchronous machine group centered on nodes 31 to 36, thanks to strong electrical coupling, maintains computational inertia above 4.5s. The simultaneous existence of a weak support region of less than 1.0s and a strong support region of more than 5.0s within the network objectively reflects the significant spatial differentiation in the system's disturbance rejection capability caused by the evolution of the power supply structure. For details on the spatial distribution characteristics of the computational inertia of all 39 nodes in the network, please refer to [link to relevant documentation]. Figure 12 .

[0206] like Figure 12 As shown in the bar chart, this visually illustrates the spatial distribution characteristics of the calculated inertia of all network nodes. High-inertia nodes are concentrated in the synchronous turbine cluster area, while low-inertia nodes are mainly located in the wind turbine area, reflecting the formation pattern of local grid inertia voids. To further compare the impact of these inertia differences on actual disturbance rejection capability, this invention selects weakly supported node 39 and strongly supported node 31 for PSS / E electromechanical transient simulation. The dynamic frequency response curves of the two nodes after disturbance are shown in the figure. Figure 13 As shown.

[0207] like Figure 13 The dynamic frequency response curves show that the low-inertia wind power access node 39 experiences a larger frequency drop after being disturbed, with its transient frequency reaching a minimum of 49.608 Hz. In contrast, the frequency decrease trend of the high-inertia synchronous turbine node 31 is relatively gentle, with the frequency extreme value remaining at 49.733 Hz. Further observation of the local magnified curves at the initial stage of the disturbance reveals that the wind power access node exhibits a steeper frequency drop slope at the moment of the fault, showing a larger initial frequency change rate; while the initial frequency drop of the synchronous turbine node is relatively gentle.

[0208] The above comparison of dynamic waveforms based on electromechanical transient simulation verifies the quantitative results regarding the magnitude of node calculated inertia mentioned earlier. Regions with smaller node calculated inertia experience more severe frequency drops in the initial stages of disturbance. This demonstrates that the evaluation model constructed in this invention can accurately and objectively reflect the true disturbance resistance characteristics of the local grid structure.

[0209] To further quantitatively verify the accuracy of the evaluation method of this invention, this invention introduces a relative error index to compare the calculated values ​​of the method with the actual simulation measurement values. The calculation formula is as follows:

[0210] ;

[0211] In the formula: To theoretically assess the relative error; The initial frequency change rate measured in the simulation; This is the calculated initial frequency change rate.

[0212] Based on the above error definition, some representative nodes were selected, and the initial frequency change rate and electromechanical transient simulation values ​​calculated by the method of this invention were extracted respectively. The absolute error and relative error between the two were compared, and the data are shown in Table 6.

[0213] Table 6

[0214] Simulation Nodes <![CDATA[The method of the present invention calculates the inertia H ck (s)]]> The method of this invention measures RoCoF (Hz / s). Actual simulation of RoCoF (Hz / s) Absolute error (Hz / s) relative error (%) Node39(Wind) 0.912 -1.644 -1.615 0.029 1.76% Node38(Wind) 0.825 -1.818 -1.780 0.038 2.09% Node31(Sync) 5.120 -0.293 -0.298 0.005 1.67% Node25(Load) 1.560 -0.961 -0.942 0.019 1.97%

[0215] Analysis of the data in Table 6 shows that, under complex operating conditions considering network spatial coupling, the relative error between the theoretical calculation values ​​and the time-domain simulation measurements remains within 2.5%. This calculation accuracy verifies that the constructed analytical model can accurately quantify the actual disturbance resistance characteristics of the local network structure, thus providing reliable data support for the subsequent implementation of dynamic topology partitioning of high-proportion renewable energy power grids.

[0216] Through the above numerical examples, the node inertia assessment method proposed in this invention has been effectively verified. It should be noted that when a real system encounters complex faults such as low-voltage ride-through, the dynamics of the phase-locked loop (PLL) and the internal nonlinear limiting circuitry of the grid-type converter will be activated. This multi-timescale coupling of electromagnetic and electromechanical transients objectively causes the actual inertia of the nodes to exhibit certain time-varying and strongly nonlinear characteristics. However, this invention is based on the analysis of electromechanical transient processes in large-scale interconnected power grids. Directly introducing the underlying electromagnetic nonlinear dynamics into the entire cascaded network simulation would not only impose a computational burden but also easily lead to the macroscopic spatial distribution of inertia being disturbed by local high-frequency microscopic processes. Therefore, in modeling and simulation, this invention utilizes the dominant supporting role of grid-type equipment at the electromechanical scale to appropriately reduce the order of the underlying nonlinear dynamics through engineering. This boundary treatment, while considering the solution efficiency of multi-machine systems, leverages the main contradiction of each node providing effective inertia support, laying a reasonable model foundation for subsequent dynamic partitioning and frequency prediction.

[0217] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of this invention.

Claims

1. A method for evaluating the computational inertia of nodes in a high-proportion renewable energy power grid, characterized in that, Includes the following steps: S1. Perform network order reduction and determine the synchronous torque coefficient between generators; S2. Determine the node inertia based on the synchronous torque coefficient between generators; S3. Construct an analytical calculation model for the node inertia and complete the analysis.

2. The method for evaluating the computational inertia of nodes in a high-proportion renewable energy power grid according to claim 1, characterized in that, S1 includes the following sub-steps: S11. Divide all network nodes into a set of generator internal potential nodes and a set of load nodes, and construct the system block admittance matrix equation; S12. Based on the system block admittance matrix equation, obtain the reduced-order network admittance matrix; S13. Extract the elements of the reduced-order network admittance matrix to determine the electromagnetic power injected by the generator into the network; S14. Based on the electromagnetic power injected into the network by the generator, determine the linearized state differential equation in matrix form; S15. Determine the synchronous torque coefficient between generators based on the linearized state differential equation in matrix form.

3. The method for evaluating the computational inertia of nodes in a high-proportion renewable energy power grid according to claim 2, characterized in that, In S11, the system block admittance matrix equation is: ； In the formula: for The vector of injected current into the internal potential node of the generator; for The voltage vector injected into the internal potential node of the generator; for The load node injected current vector; for The load node injected voltage vector; This is the block matrix of the first mutual admittance and self admittance between the corresponding node sets; This is the block matrix for the second mutual admittance and self admittance between corresponding node sets; This is the block matrix for the third mutual admittance and self-admittance between corresponding node sets; This is the fourth mutual admittance and self admittance block matrix between corresponding node sets; In S13, the first Electromagnetic power injected into the network by the generator for: ； In the formula: For generator The amplitude of the internal potential during the transient process; For generator The amplitude of the internal potential during the transient process; These are the main diagonal conductance elements of the admittance matrix of the reduced-order network; For generator With generator Transfer conductance between; For generator With generator Transfer susceptance between; The phase difference is the electrical angle of the two rotors; In S14, the linearized state differential equation in matrix form is: ； In the formula: This is the first diagonal matrix; It is the second diagonal matrix; For the rotor's angular velocity deviation, The rate of change of angular velocity; This refers to the deviation of the work angle; The deviation in mechanical power input to the prime mover; This is the electromagnetic power deviation vector output by the generator; This is the synchronous torque coefficient matrix; In S15, the generator With generator Synchronous torque coefficient between for: ； In the formula: The initial power angle difference at the steady-state operating point. For generator The rotor power angle.

4. The method for evaluating the computational inertia of nodes in a high-proportion renewable energy power grid according to claim 1, characterized in that, In S2, the node computational inertia H of node k ck for: ； Where: H i For generator The rotor inertia time constant; When node k experiences an active power deficit, the generator... The initial electromagnetic power fraction extracted; This represents the number of generators.

5. The method for evaluating the computational inertia of nodes in a high-proportion renewable energy power grid according to claim 1, characterized in that, S3 includes the following sub-steps: S31. Construct the rotor motion equations of the synchronous machine in a per-unit system; S32. Transform the rotor motion equation of the synchronous machine in the per-unit system into a complex frequency domain expression; S33. Based on the complex frequency domain expression, determine the available energy reserves and the short-term maximum support power; S34. Based on the available energy reserves and short-term maximum support power, construct an analytical calculation model for the node's computational inertia and complete the analysis.

6. The method for evaluating the computational inertia of nodes in a high-proportion renewable energy power grid according to claim 5, characterized in that, In S31, the rotor motion equation of the synchronous machine in the per-unit system is: ； In the formula: H i For generator The rotor inertia time constant; This is the per-unit deviation of the rotor's electrical angular velocity; The deviation in mechanical power input to the prime mover; This refers to the deviation in electromagnetic power output by the generator. Damping coefficient provided for generator windings and load effects; In step S32, the rotor motion equation of the synchronous machine in the per-unit system is transformed into a complex frequency domain expression as follows: ； In the formula: s is the Laplace operator; This is the complex frequency domain form of the per-unit deviation of the rotor's electric angular velocity; The complex frequency domain form of the mechanical power deviation input to the prime mover; This is the complex frequency domain form of the electromagnetic power deviation of the generator output.

7. The method for evaluating the computational inertia of nodes in a high-proportion renewable energy power grid according to claim 5, characterized in that, In S33, the energy reserve can be called up. With short-term maximum support power They are respectively: ； ； In the formula: H i For generator The rotor inertia time constant; This refers to the rated base capacity of the unit; This is the maximum transient active overload factor set for the unit in the simulation model.

8. The method for evaluating the computational inertia of nodes in a high-proportion renewable energy power grid according to claim 5, characterized in that, In S34, the analytical calculation model is as follows: ； In the formula: When node k experiences an active power deficit, the generator... The initial electromagnetic power fraction extracted; When node k experiences an active power deficit, it switches from the generator. The initial electromagnetic power fraction extracted; H i For generator The rotor inertia time constant; The equivalent virtual inertia of a grid-type wind turbine; S G It is a traditional synchronous generator; S W It is a grid-type wind turbine.

Images