A method, device, equipment and medium for identifying regional inertia of a power system

Abstract

The application discloses a regional inertia identification method, device, equipment and medium of a power system, and relates to the technical field of power systems, which comprises the following steps: dividing the power system into target regions, constructing a regional center frequency, and constructing a regional net accelerating power based on the power variation of the power components in the target region; preprocessing the regional center frequency and the regional net accelerating power respectively, determining the regional frequency deviation and the regional frequency variation rate based on the processed frequency; constructing the regional equivalent frequency response relationship of the target region; constructing a candidate function library, screening the leading term by using a sparse dynamics identification method taking the regional frequency variation rate as the identification target, constructing a regional frequency dynamic model and determining the model coefficients; and mapping the model coefficients to the equivalent inertia of the target region and the equivalent damping of the target region, and determining the regional inertia distribution result of the power system. The application can meet the accurate identification demand of dynamic and regional equivalent inertia under the multi-regional high-new energy power grid.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a method, apparatus, equipment and medium for identifying the regional inertia of a power system. Background Technology

[0002] With the continuous expansion of grid-connected power electronic equipment such as wind power, photovoltaics, energy storage, and virtual synchronous generators, the proportion of synchronous generator units in the grid is continuously decreasing. The system inertia has transformed from the mechanical inertia of a single synchronous generator to a hybrid inertia formed by the coupling of the actual inertia of the synchronous generator, the virtual inertia of new energy sources, and the frequency regulation response of energy storage. Synchronous generator units rely on the natural kinetic energy of the rotor to support the frequency, while power electronic equipment relies on control algorithms to simulate inertial damping. Multiple types of equipment jointly dominate the dynamics of system frequency. Significant differences exist in the power supply structure and network topology of different regions, resulting in regionally differentiated and time-varying dynamic characteristics of the system's equivalent inertia. Accurate identification of regional equivalent inertia has become a key requirement for the stable operation of the power grid.

[0003] Current mainstream inertia identification methods suffer from several inherent flaws. First, traditional mechanistic inertia assessment methods rely on fixed inertia constants on the nameplates of synchronous generator units, making them only suitable for traditional power grids dominated by synchronous power sources. They are unsuitable for converter equipment such as wind power, photovoltaics, energy storage, and virtual synchronous machines. The equivalent inertia of these devices is constrained by multiple factors including control parameters, real-time output, power margin, and grid connection conditions, resulting in the absence of constant mechanical inertia parameters and significant identification errors with traditional methods. Second, while existing data-driven identification algorithms can solve parameters using frequency and power time-series data under disturbances, most models do not incorporate the physical constraints corresponding to the power system rotor motion equations. The model coefficients lack clear physical meaning and cannot be directly mapped to regional equivalent inertia or equivalent damping, leading to poor interpretability of the identification results. Third, conventional identification algorithms require pre-setting of fixed model order and structure, making it difficult to adapt to dynamic frequency changes caused by fluctuations in new energy output, grid operation mode switching, and adjustments to virtual inertia control parameters. Inappropriate model structure selection can easily lead to large inertia identification biases and weak model generalization capabilities.

[0004] Therefore, how to meet the need for accurate identification of dynamic and regional equivalent inertia under multi-regional high-energy power grids is a technical problem that urgently needs to be solved. Summary of the Invention

[0005] In view of this, the purpose of this invention is to provide a method, apparatus, device, and storage medium for regional inertia identification in power systems, capable of meeting the requirements for accurate identification of dynamic, regional equivalent inertia under multi-regional high-energy power grids. The specific solution is as follows: In a first aspect, this application provides a method for regional inertia identification of a power system, including: The power system is divided into target regions, and corresponding regional center frequencies are constructed. The regional net acceleration power is constructed based on the power change of the power components within the target regions. The regional center frequency is an equivalent frequency variable that characterizes the overall frequency dynamics within the target regions. The regional net acceleration power is an equivalent power imbalance that causes frequency acceleration or deceleration within the target regions. The center frequency and net acceleration power of the region are preprocessed to obtain the processed frequency and processed power. Based on the processed frequency, the regional frequency deviation and the corresponding regional frequency change rate of the target region are determined. The equivalent frequency response relationship of the target region is constructed based on the processed power, the regional frequency deviation, the regional frequency change rate, the regional equivalent inertia and the regional equivalent damping of the target region. A candidate function library containing various candidate terms affecting the power system is constructed. A sparse dynamics identification method with regional frequency change rate as the identification target is used to screen dominant terms from the candidate function library. A regional frequency dynamic model is constructed based on the dominant terms, and the model coefficients of the regional frequency dynamic model are determined. The dominant terms are terms that play a predetermined decisive role in regional frequency dynamics. Based on the regional equivalent frequency response relationship, the model coefficients are mapped to the target region equivalent inertia and target region equivalent damping. Based on the target region equivalent inertia and target region equivalent damping of each target region, the regional inertia distribution result of the power system is determined, so as to trigger the regional frequency security analysis operation and regional inertia control operation of the power system based on the regional inertia distribution result.

[0006] Optionally, the preprocessing of the regional center frequency and the regional net acceleration power to obtain the processed frequency and processed power includes: The center frequency and net acceleration power of the region are subjected to outlier removal, time alignment, filtering, normalization and smoothing differentiation to obtain the processed frequency and processed power.

[0007] Optionally, constructing the regional net acceleration power based on the power change of power components within the target region includes: The active power change of synchronous generators, the output change of new energy equipment, the power change of energy storage equipment, and the active power change of virtual synchronous generators or grid-type equipment within the target area are obtained to obtain the active power change on the power supply side. Obtain the load power change and tie line switching power change of the target area; The net acceleration power of the region is constructed based on the changes in active power on the power supply side, the changes in load power, and the changes in tie-line switching power.

[0008] Optionally, the candidate function library includes regional frequency deviation terms, regional net acceleration power terms, power response terms for various devices, tie-line power variation terms, and nonlinear correction terms.

[0009] Optionally, the method of using sparse dynamics identification with the regional frequency change rate as the identification target to filter dominant terms from the candidate function library includes: The regional frequency change rate is determined as the identification target, and a sparse dynamic identification model is constructed based on the identification target and the sparse coefficients to be solved. An output vector is constructed based on the target region frequency change rate at sampling times within a preset disturbance time window, and a candidate function matrix is ​​constructed based on the candidate function values ​​corresponding to the sampling times; the candidate function values ​​are determined based on the candidate functions and the sampling data of the target region. The sparse dynamics identification problem is constructed based on the output vector, the candidate function matrix, the sparse coefficients, and the model residuals of the sparse dynamics identification model. Based on the output vector, the candidate function matrix, the sparse coefficients, and the preset ridge regression regularization parameters, a target optimization function is constructed. The target optimization function is minimized using the ridge regression least squares method to solve the sparse dynamics identification problem and obtain the target sparse coefficients. The coefficients in the target sparse coefficients whose absolute values ​​are less than a preset threshold are set to zero to obtain processed coefficients that only contain non-zero coefficients. Obtain the target candidate corresponding to the processed coefficient from the candidate function matrix, construct a new candidate function matrix based on the target candidate, and determine the dominant term in the candidate function library based on the new candidate function matrix and the ridge regression least squares method.

[0010] Optionally, the model coefficients include the coefficients of the regional net acceleration power term and the coefficients of the regional frequency deviation term in the dominant term.

[0011] Optionally, after determining the regional inertia distribution results of the power system, the method further includes: Target nodes within the target area are identified, and a node dynamic contribution index for the target nodes is constructed based on the change in active power and the sampling time interval at each preset sampling time. The nodes include load nodes and frequency measurement nodes. Determine the node frequency sensitivity, which characterizes the degree of influence of node power changes on the center frequency of the region; The node inertia contribution weight of the target node is determined based on the node dynamic contribution index and the node frequency sensitivity. The equivalent inertia of the target region is decomposed based on the node inertia contribution weight to obtain the node equivalent inertia contribution of the target node; the node equivalent inertia contribution characterizes the degree of contribution of the node and its associated devices to the regional inertia support.

[0012] Secondly, this application provides a regional inertia identification device for a power system, comprising: The region construction module is used to divide the power system into target regions, construct corresponding regional center frequencies, and construct regional net acceleration power based on the power change of power components within the target regions. The regional center frequency is an equivalent frequency variable characterizing the overall frequency dynamics within the target regions. The regional net acceleration power is an equivalent power imbalance that causes frequency acceleration or deceleration within the target regions. The preprocessing module is used to preprocess the center frequency of the region and the net acceleration power of the region respectively to obtain the processed frequency and the processed power, and to determine the regional frequency deviation and the corresponding regional frequency change rate of the target region based on the processed frequency. The relationship construction module is used to construct the regional equivalent frequency response relationship of the target region based on the processed power, the regional frequency deviation, the regional frequency change rate, the regional equivalent inertia and the regional equivalent damping of the target region; The dominant term identification module is used to construct a candidate function library containing various candidate terms that affect the power system, and to use a sparse dynamics identification method with the regional frequency change rate as the identification target to screen dominant terms from the candidate function library, and to construct a regional frequency dynamic model based on the dominant terms, and to determine the model coefficients of the regional frequency dynamic model; the dominant term is a term that plays a predetermined decisive role in the regional frequency dynamics. An inertia determination module is used to map the model coefficients to target region equivalent inertia and target region equivalent damping based on the regional equivalent frequency response relationship, and to determine the regional inertia distribution result of the power system based on the target region equivalent inertia and target region equivalent damping of each target region, so as to trigger regional frequency security analysis and regional inertia control operations of the power system based on the regional inertia distribution result. Thirdly, this application provides an electronic device, including: Memory, used to store computer programs; A processor is used to execute the computer program to implement the aforementioned method for regional inertia identification of power systems.

[0013] Fourthly, this application provides a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements the aforementioned method for regional inertia identification of a power system.

[0014] In this application, the power system is divided into target regions, and corresponding regional center frequencies are constructed. A regional net acceleration power is constructed based on the power changes of power components within the target regions. The regional center frequency is an equivalent frequency variable characterizing the overall frequency dynamics within the target regions. The regional net acceleration power is an equivalent power imbalance that causes frequency acceleration or deceleration within the target regions. The regional center frequencies and the regional net acceleration power are preprocessed to obtain processed frequencies and processed power. Based on the processed frequencies, the regional frequency deviation and corresponding regional frequency change rate of the target regions are determined. A regional equivalent of the target regions is constructed based on the processed power, the regional frequency deviation, the regional frequency change rate, the regional equivalent inertia of the target regions, and the regional equivalent damping. The system establishes an effective frequency response relationship; constructs a candidate function library containing various candidate factors affecting the power system; uses a sparse dynamics identification method with the regional frequency change rate as the identification target to screen dominant terms from the candidate function library; constructs a regional frequency dynamic model based on the dominant terms; and determines the model coefficients of the regional frequency dynamic model. The dominant terms are those that play a predetermined decisive role in the regional frequency dynamics. Based on the regional equivalent frequency response relationship, the model coefficients are mapped to the target regional equivalent inertia and target regional equivalent damping. Based on the target regional equivalent inertia and target regional equivalent damping of each target region, the regional inertia distribution results of the power system are determined. Based on the regional inertia distribution results, regional frequency security analysis and regional inertia control operations of the power system are triggered. As can be seen from the above, this application first divides the power system into target regions, constructs a regional center frequency representing the overall frequency dynamics of the region, and constructs the regional net acceleration power based on the power changes of power components within the target region. Then, it preprocesses the regional center frequency and regional net acceleration power, determines the regional frequency deviation and regional frequency change rate based on the processed frequencies, and constructs a regional equivalent frequency response relationship with clear physical meaning. On this basis, it constructs a candidate function library containing various candidate options, and uses a sparse dynamics identification method that identifies targets based on the regional frequency change rate to adaptively select dominant terms and determine model coefficients. Finally, based on the regional equivalent frequency response relationship, it maps the model coefficients to the target region's equivalent inertia and target region's equivalent damping, obtaining the regional inertia distribution results of the power system. In this way, this application can meet the accurate identification requirements of dynamic, regional equivalent inertia under multi-regional high-energy power grids, providing a reliable basis for regional frequency security analysis and regional inertia control of the power system. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0016] Figure 1 This is a flowchart of a method for identifying the regional inertia of a power system disclosed in this application; Figure 2 This is a flowchart of a specific method for regional inertia identification of a power system disclosed in this application; Figure 3 (a) is a schematic diagram of a region 1 time-domain simulation disclosed in this application; Figure 3 (b) is a schematic diagram of a region 1 derivative fitting disclosed in this application; Figure 3 (c) is a schematic diagram of a time-domain simulation error in region 1 disclosed in this application; Figure 3 (d) is a schematic diagram of a region 2 time-domain simulation disclosed in this application; Figure 3 (e) is a schematic diagram of a region 2 derivative fitting method disclosed in this application. Figure 3 (f) is a schematic diagram of a region 2 time-domain simulation error disclosed in this application; Figure 4 This is a schematic diagram of the structure of a regional inertia identification device for a power system disclosed in this application; Figure 5 This is a structural diagram of an electronic device disclosed in this application. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Currently, mainstream inertia identification methods suffer from several inherent flaws. First, traditional mechanistic inertia assessment methods rely on fixed inertia constants on the nameplates of synchronous generator units for calculation. This only applies to traditional power grids dominated by synchronous power sources and cannot be adapted to converter equipment such as wind power, photovoltaics, energy storage, and virtual synchronous machines. The equivalent inertia of such equipment is constrained by multiple factors including control parameters, real-time output, power margin, and grid connection conditions, and there are no constant mechanical inertia parameters, leading to significant identification errors using traditional methods. Second, while existing data-driven identification algorithms can solve parameters based on frequency and power time-series data under disturbances, most models do not incorporate the physical constraints corresponding to the power system rotor motion equations. The model coefficients lack clear physical meaning and cannot be directly mapped to regional equivalent inertia or equivalent damping, resulting in poor interpretability of the identification results. Third, conventional identification algorithms require pre-setting of fixed model order and structure, making it difficult to adapt to dynamic frequency changes caused by fluctuations in new energy output, switching of grid operation modes, and adjustments to virtual inertia control parameters. Inappropriate model structure selection can easily lead to large inertia identification biases and weak model generalization capabilities. Therefore, this application provides a method, device, equipment and medium for regional inertia identification of power systems, which can meet the needs of accurate identification of dynamic and regional equivalent inertia under multi-regional high-energy power grids.

[0019] See Figure 1 As shown, this embodiment of the invention discloses a method for regional inertia identification in a power system, including: Step S11: Divide the power system into target regions, construct corresponding regional center frequencies, and construct regional net acceleration power based on the power change of power components within the target regions; the regional center frequency is an equivalent frequency variable characterizing the overall frequency dynamics within the target regions; the regional net acceleration power is an equivalent power imbalance that causes frequency acceleration or deceleration within the target regions.

[0020] In this embodiment, the power system can be divided into several target regions based on the electrical, geographical, or dispatching zones of the power grid. For each target region, a regional center frequency can be constructed based on the frequency measurement data of each frequency measurement node within the region to characterize the overall frequency dynamics within that target region.

[0021] Simultaneously, the changes in active power of synchronous generators, output of new energy equipment, power of energy storage equipment, and active power of virtual synchronous generators or grid-type equipment within the target area are obtained to obtain the change in active power on the power supply side; the changes in load power and tie-line switching power of the target area are obtained; and the net acceleration power of the area is constructed based on the changes in active power on the power supply side, the changes in load power, and the changes in tie-line switching power.

[0022] Step S12: Preprocess the center frequency of the region and the net acceleration power of the region to obtain the processed frequency and processed power. Based on the processed frequency, determine the regional frequency deviation and the corresponding regional frequency change rate of the target region.

[0023] In this embodiment, considering that there may be bad data, inconsistent time scales, and measurement noise in the collected frequency and power data, outlier removal, time alignment, filtering, normalization, and smoothing differentiation can be performed on the center frequency of the region and the net acceleration power of the region to obtain the processed frequency and processed power.

[0024] Then, the regional frequency deviation is determined based on the difference between the processed frequency and the system rated frequency, and the regional frequency deviation is smoothed and differentiated to obtain the regional frequency change rate.

[0025] Step S13: Construct the regional equivalent frequency response relationship of the target region based on the processed power, the regional frequency deviation, the regional frequency change rate, the regional equivalent inertia and the regional equivalent damping of the target region.

[0026] In this embodiment, the regional equivalent frequency response relationship of the target region can be constructed based on the processed power, the regional frequency deviation, the regional frequency change rate, the regional equivalent inertia, and the regional equivalent damping of the target region. This regional equivalent frequency response relationship allows the subsequently identified model coefficients to be directly mapped to physically meaningful regional equivalent inertia and regional equivalent damping.

[0027] Step S14: Construct a candidate function library containing various candidate options that affect the power system; use a sparse dynamics identification method with the regional frequency change rate as the identification target to screen dominant terms from the candidate function library; construct a regional frequency dynamic model based on the dominant terms; and determine the model coefficients of the regional frequency dynamic model; the dominant terms are terms that play a predetermined decisive role in the regional frequency dynamics.

[0028] In this embodiment, a candidate function library is constructed, containing various candidate terms that affect the power system. The candidate function library includes regional frequency deviation terms, regional net acceleration power terms, power response terms for various equipment, tie-line power variation terms, and nonlinear correction terms.

[0029] Furthermore, the regional frequency change rate is determined as the identification target, and a sparse dynamics identification model is constructed based on the identification target and the sparse coefficients to be solved. An output vector is constructed based on the target region's frequency change rate at sampling times within a preset perturbation time window, and a candidate function matrix is ​​constructed based on the candidate function values ​​corresponding to the sampling times. The candidate function values ​​are determined based on the candidate values ​​and the sampling data of the target region. A sparse dynamics identification problem is constructed based on the output vector, the candidate function matrix, the sparse coefficients, and the model residuals of the sparse dynamics identification model. Based on the output vector and the candidate function matrix, a sparse dynamics identification problem is constructed. A target optimization function is constructed using a function matrix, the sparse coefficients, and a preset ridge regression regularization parameter. The target optimization function is minimized using ridge regression least squares to solve the sparse dynamics identification problem and obtain the target sparse coefficients. Coefficients in the target sparse coefficients whose absolute values ​​are less than a preset threshold are set to zero to obtain processed coefficients containing only non-zero coefficients. Target candidates corresponding to the processed coefficients are obtained from the candidate function matrix. A new candidate function matrix is ​​constructed based on the target candidate candidates, and the dominant term in the candidate function library is determined based on the new candidate function matrix and the ridge regression least squares method.

[0030] Next, a regional frequency dynamic model is constructed based on the dominant term, and its model coefficients are determined. The model coefficients include the coefficients of the regional net acceleration power term and the coefficients of the regional frequency deviation term in the dominant term.

[0031] Step S15: Based on the regional equivalent frequency response relationship, map the model coefficients to the target region equivalent inertia and target region equivalent damping, and determine the regional inertia distribution result of the power system based on the target region equivalent inertia and target region equivalent damping of each target region, so as to trigger the regional frequency security analysis operation and regional inertia control operation of the power system based on the regional inertia distribution result.

[0032] In this embodiment, by comparing the model coefficients with the regional equivalent frequency response, the coefficients of the regional net acceleration power term and the regional frequency deviation term in the dominant terms can be mapped to the target regional equivalent inertia and target regional equivalent damping. By summarizing the target regional equivalent inertia and target regional equivalent damping for each target region, the regional inertia distribution result of the power system can be obtained. Based on this regional inertia distribution result, regional frequency security analysis and regional inertia control operations of the power system can be triggered. After determining the regional inertia distribution of the power system, in order to identify the specific contribution of each node and its associated equipment to the regional inertia support, and thus provide a more specific decision-making basis for regional frequency security analysis and inertia regulation, it is necessary to further decompose the regional equivalent inertia to the node level.

[0033] To this end, target nodes within the target area are identified, and a node dynamic contribution index for the target nodes is constructed based on the active power changes and sampling time intervals of the target nodes at each preset sampling time. Nodes include load nodes and frequency measurement nodes. A node frequency sensitivity, characterizing the impact of node power changes on the center frequency of the region, is determined. Based on the node dynamic contribution index and the node frequency sensitivity, a node inertia contribution weight for the target nodes is determined. Based on the node inertia contribution weight, the equivalent inertia of the target region is decomposed to obtain the node equivalent inertia contribution of the target nodes. The node equivalent inertia contribution characterizes the contribution of the node and its associated equipment to the regional inertia support. Based on the node equivalent inertia contribution, the support capability of each node in the regional frequency response can be quantified, thereby guiding the coordinated configuration of distributed resources in regional inertia control operations and improving the accuracy of identifying weak links in regional frequency security analysis.

[0034] As can be seen from the above, this application first divides the power system into target regions, constructs a regional center frequency representing the overall frequency dynamics of the region, and constructs the regional net acceleration power based on the power changes of power components within the target region. Then, it preprocesses the regional center frequency and regional net acceleration power, determines the regional frequency deviation and regional frequency change rate based on the processed frequencies, and constructs a regional equivalent frequency response relationship with clear physical meaning. On this basis, it constructs a candidate function library containing various candidate options, and uses a sparse dynamics identification method that identifies targets based on the regional frequency change rate to adaptively select dominant terms and determine model coefficients. Finally, based on the regional equivalent frequency response relationship, it maps the model coefficients to the target region's equivalent inertia and target region's equivalent damping, obtaining the regional inertia distribution results of the power system. In this way, this application can meet the accurate identification requirements of dynamic, regional equivalent inertia under multi-regional high-energy power grids, providing a reliable basis for regional frequency security analysis and regional inertia control of the power system.

[0035] The following is combined Figure 2 The schematic diagram shown illustrates the technical solution of the embodiments of this application in detail.

[0036] First, some of the concepts in this application will be explained in detail.

[0037] Regional inertia refers to the equivalent inertia support capacity of a specific power grid region to frequency changes. Due to differences in synchronous generator capacity, renewable energy ratio, load levels, grid structure, and disturbance propagation characteristics across different regions, the inertia levels may vary significantly. Therefore, a single system-wide inertia value cannot be used to describe the frequency support capacity of the entire power grid; rather, inertia assessments and identifications for different regions are necessary.

[0038] The regional center frequency is an equivalent frequency variable used to represent the overall frequency dynamics of a region. In multi-regional power systems, there may be multiple frequency measurement points within the same region, such as synchronous generator terminal buses, renewable energy plant grid connection points, energy storage grid connection points, virtual synchronous generator grid connection points, or critical load buses. Since the frequency response of different measurement points may be affected by local disturbances, measurement point locations, and measurement noise, it is usually necessary to select a frequency signal that can represent the overall dynamic characteristics of the region as the state variable for subsequent regional inertia identification.

[0039] Regional net acceleration power refers to the equivalent power imbalance that causes frequency acceleration or deceleration changes within a certain region. It includes not only sudden load changes, but also factors such as fluctuations in new energy output, changes in electromagnetic power of synchronous generators, active power response of virtual synchronous generators, energy storage power response, changes in power exchanged by regional tie lines, and unmodeled power disturbances.

[0040] The rate of change of frequency refers to how quickly the frequency changes over time, and is usually used to reflect how fast the frequency rises or falls after a disturbance. The smaller the inertia, the greater the rate of change of frequency is usually under the same power disturbance; the greater the inertia, the smaller the rate of change of frequency is usually.

[0041] The regional equivalent oscillation equation is the fundamental equation describing the frequency dynamics of a synchronous generator or equivalent power system. In regional inertia identification, a region can be equivalently represented as a aggregated frequency response object, and the change in the region's center frequency is jointly determined by the region's net acceleration power, regional equivalent inertia, and regional equivalent damping.

[0042] Sparse dynamics identification first constructs a candidate function library based on possible dynamic relationships, and then uses sparse regression methods to filter out the few terms that truly play a dominant role from a large number of candidate options. For regional inertia identification in power systems, regional frequency response, rate of change of frequency, and regional net acceleration power data can be used to identify the main physical terms in the regional frequency dynamic equations.

[0043] The physical constraint candidate function library is a set of dynamic terms to be screened. When performing data-driven modeling, the actual dynamic equations of the system are usually unknown. Therefore, it is necessary to first list a set of candidate terms that may affect the system dynamics based on existing physical laws and variable relationships, and then use sparse regression to screen out the terms that truly play a major role.

[0044] Specifically, the implementation process of this scheme involves dividing the power system into R regions based on electrical connections, operating topology, dispatching zones, or frequency response similarity. The r-th region contains several synchronous generators, new energy equipment, energy storage devices, virtual synchronous generators, load nodes, and frequency measurement nodes, where r = 1, 2, ... ,R.

[0045] For the r-th region, a regional center frequency that can characterize the overall frequency dynamics of the region is selected, denoted as . The regional center frequency can be obtained from representative measurement points within the region. The regional center frequency is the equivalent frequency variable used to represent the overall frequency dynamics of a region. In multi-regional power systems, multiple frequency measurement points may exist within the same region, such as synchronous generator terminal buses, renewable energy plant grid connection points, energy storage grid connection points, virtual synchronous generator grid connection points, or critical load buses. Since the frequency response of different measurement points may be affected by local disturbances, measurement point locations, and measurement noise, it is usually necessary to select a frequency signal that can represent the overall dynamic characteristics of the region as the state variable for subsequent regional inertia identification.

[0046] Define the frequency deviation of the r-th region as: ; in, f 0 represents the system's rated frequency.

[0047] Further data collection is needed on the active power changes of synchronous generators, new energy equipment, energy storage equipment, virtual synchronous generators, loads, and tie lines within the r-th region to construct the region's multi-source net acceleration power: ; in, For the first The regional multi-source net acceleration power of each region This represents the change in active power on the power supply side within the region. This represents the change in regional load power. This refers to the change in power exchanged on the regional tie line.

[0048] The change in active power on the regional power supply side can be further expressed as: ; in, This represents the change in active power of the synchronous generator. The change in the amount of energy contributed to new energy sources This represents the change in energy storage power. This refers to the change in active power of virtual synchronous generators or grid-type equipment.

[0049] For the regional center frequency and regional multi-source net acceleration power Outlier removal, time alignment, filtering, standardization, and smoothing differentiation are performed to obtain the regional frequency deviation. and its rate of change : ; According to the regional frequency dynamic mechanism, the first The equivalent frequency response relationship of each region can be expressed as: ; in, For the first The equivalent inertia of each region For the first Equivalent damping in each region.

[0050] The above formula can be rearranged as follows: ; Therefore, it can be seen that the coefficient of the regional multi-source net acceleration power term is related to the regional equivalent inertia, and the coefficient of the regional frequency deviation term is related to the regional equivalent damping.

[0051] To identify the dominant regional frequency dynamics from perturbation data, a library of candidate physical constraint functions is constructed: ; Candidate terms include regional frequency deviation, regional multi-source net acceleration power, power response of various equipment, tie-line power variation, and necessary nonlinear correction terms.

[0052] With regional frequency change rate As the identification target, a sparse dynamics identification model is established: ; in, For the first A sparse coefficient vector of regions to be identified is obtained. Regional frequency dynamics dominant terms are screened from a candidate function library using sparse regression to obtain the regional frequency dynamic model.

[0053] Based on discrete sampled data, let the first... The region has a total of [number] regions within the disturbance time window. Each sampling point is used. The regional frequency change rate at each sampling time is used to form an output vector: ; The candidate function values ​​corresponding to each sampling time are used to form a candidate function matrix: ; in, This represents the candidate function matrix composed of the candidate function values ​​at each sampling time. It is a set of candidates. It is a data matrix obtained by substituting the sampled data into the candidate options.

[0054] The sparse dynamics identification problem can then be written as: ; in, Let be the sparse coefficient vector to be identified. This represents the model residuals.

[0055] To improve the stability of coefficient calculation under noisy conditions, this invention employs the ridge regression least squares method to solve for the coefficient vector. Its optimization objective is: ; in, This is the ridge regression regularization parameter, used to suppress coefficient fluctuations caused by ill-conditioned candidate function matrices or measurement noise.

[0056] The corresponding closed-form solution is: .

[0057] in, It is an identity matrix.

[0058] To maintain the sparsity of the model, after obtaining the initial coefficients through ridge regression, a coefficient threshold can be set, and coefficients with absolute values ​​less than the threshold can be set to zero. ; in, for Middle The coefficients corresponding to each candidate. Use the threshold to filter parameters.

[0059] Then, retaining the candidate options corresponding to the non-zero coefficients, the candidate function matrix is ​​reconstructed. And then the ridge regression least squares method was used again to solve the problem: ; The final coefficient vector is obtained through a process of "ridge regression solution—threshold screening—re-estimation of retained terms." The regional multi-source net acceleration power term is among the final retained dominant candidates. The coefficient is Regional frequency deviation term The coefficient is The regional multi-source net acceleration power term and the regional frequency deviation term are the main retained terms, which can be simplified to: ; in, and These are the model coefficients obtained through identification.

[0060] By mapping the identification model to the regional equivalent frequency response relationship, we can obtain: , ; Therefore, the first The equivalent inertia and equivalent damping of each region can be calculated as follows: , ; This completes the first... Identification of the equivalent inertia and equivalent damping of each region. Repeating the above process for all regions yields the regional inertia distribution of the entire system. ; After obtaining the equivalent inertia of the region, the relationship between the regional inertia and the contribution of the nodal equivalent inertia is further established. For the th The first region For each node, a dynamic contribution index is constructed based on the node's active power response during the disturbance process: ; in, For the first The node at the th The change in active power at each sampling time. This represents the sampling time interval.

[0061] Further introduce node frequency sensitivity or node influence factor This is used to characterize the impact of node power variations on the regional center frequency. Based on the node dynamic contribution index and node frequency sensitivity, the node inertia contribution weight is constructed as follows: ; in, For the first A set of nodes within a region. For the first The inertia contribution weights of each node, satisfying: ; equivalent inertia of the region The equivalent inertia contribution of each node is obtained by decomposing it according to its contribution weight: ; in, For the first The equivalent inertia contribution of each node. The equivalent inertia contribution of a node represents the degree to which the node and its associated equipment contribute to the regional inertia support.

[0062] Furthermore, based on this scheme, simulation verification was conducted on a four-machine, two-zone system. Synchronous generator G4 (Generator No. 4) was replaced with a Virtual Synchronous Generator (VSG), and the system was divided into two regions for regional inertia assessment. The time-domain and derivative-domain simulations are as follows: Figure 3As shown in Table 1, the goodness of fit of the time-domain simulations for both regions is above 99%, and the goodness of fit in the derivative domain is also close to 1, thus verifying the accuracy of the evaluation and the effectiveness of the proposed method. The regional inertia identification results are shown in Table 1 below.

[0063] Table 1. Regional Inertia Identification Results

[0064] In summary, this application establishes a physically constrained sparse dynamic identification model in multi-regional power systems with a high proportion of renewable energy access by utilizing regional frequency response, regional net acceleration power, and equipment power change data during disturbance processes. It automatically extracts the dominant terms in regional frequency dynamics and maps the identified sparse model coefficients to regional equivalent inertia and regional equivalent damping. This enables accurate, interpretable, and adaptable identification of regional inertia to changes in operating modes, providing a basis for low inertia risk assessment, regional frequency stability analysis, and subsequent inertia-supported control.

[0065] Accordingly, see Figure 4 As shown in the figure, this application provides a regional inertia identification device for a power system, including: The region construction module 11 is used to divide the power system into target regions, construct corresponding regional center frequencies, and construct regional net acceleration power based on the power change of power components within the target regions. The regional center frequency is an equivalent frequency variable characterizing the overall frequency dynamics within the target regions. The regional net acceleration power is an equivalent power imbalance that causes frequency acceleration or deceleration within the target regions. Preprocessing module 12 is used to preprocess the center frequency of the region and the net acceleration power of the region respectively to obtain the processed frequency and the processed power, and to determine the regional frequency deviation and the corresponding regional frequency change rate of the target region based on the processed frequency. The relationship construction module 13 is used to construct the regional equivalent frequency response relationship of the target region based on the processed power, the regional frequency deviation, the regional frequency change rate, the regional equivalent inertia and the regional equivalent damping of the target region. The dominant term identification module 14 is used to construct a candidate function library containing various candidate terms that affect the power system, use a sparse dynamics identification method with the regional frequency change rate as the identification target to screen dominant terms from the candidate function library, construct a regional frequency dynamic model based on the dominant terms, and determine the model coefficients of the regional frequency dynamic model; the dominant term is a term that plays a predetermined decisive role in the regional frequency dynamics. The inertia determination module 15 is used to map the model coefficients to target area equivalent inertia and target area equivalent damping based on the regional equivalent frequency response relationship, and to determine the regional inertia distribution result of the power system based on the target area equivalent inertia and target area equivalent damping of each target area, so as to trigger the regional frequency security analysis operation and regional inertia control operation of the power system based on the regional inertia distribution result.

[0066] In some specific embodiments, the region construction module 11 specifically includes: The first change determination unit is used to obtain the change in active power of synchronous generators, the change in output of new energy equipment, the change in power of energy storage equipment, and the change in active power of virtual synchronous generators or grid-type equipment in the target area, so as to obtain the change in active power on the power supply side. The second change determination unit is used to obtain the load power change and tie line switching power change of the target area; A power construction unit is used to construct the net acceleration power of the region based on the change in active power on the power supply side, the change in load power, and the change in tie-line switching power.

[0067] In some specific embodiments, the preprocessing module 12 specifically includes: The information processing unit is used to perform outlier removal, time alignment, filtering, standardization, and smoothing differentiation on the center frequency and net acceleration power of the region, respectively, to obtain the processed frequency and processed power.

[0068] In some specific implementations, the candidate function library includes regional frequency deviation terms, regional net acceleration power terms, power response terms for various devices, tie-line power variation terms, and nonlinear correction terms.

[0069] In some specific embodiments, the dominant item identification module 14 specifically includes: The model building unit is used to determine the regional frequency change rate as the identification target and to build a sparse dynamic identification model based on the identification target and the sparse coefficients to be solved. A matrix construction unit is used to construct an output vector based on the target region frequency change rate at sampling times within a preset disturbance time window, and to construct a candidate function matrix based on the candidate function values ​​corresponding to the sampling times; the candidate function values ​​are determined based on the candidate functions and the sampling data of the target region. The problem construction unit is used to construct a sparse dynamics identification problem based on the output vector, the candidate function matrix, the sparse coefficients, and the model residual of the sparse dynamics identification model. The problem-solving unit is used to construct a target optimization function based on the output vector, the candidate function matrix, the sparse coefficients, and a preset ridge regression regularization parameter, and to minimize the target optimization function using the ridge regression least squares method to solve the sparse dynamics identification problem and obtain the target sparse coefficients. The coefficient processing unit is used to set the coefficients whose absolute values ​​are less than a preset threshold in the target sparse coefficients to zero, so as to obtain processed coefficients that only contain non-zero coefficients. The dominant term determination unit is used to obtain the target candidate terms corresponding to the processed coefficients in the candidate function matrix, construct a new candidate function matrix based on the target candidate terms, and determine the dominant terms in the candidate function library based on the new candidate function matrix and the ridge regression least squares method.

[0070] In some specific implementations, the model coefficients include the coefficients of the regional net acceleration power term and the coefficients of the regional frequency deviation term in the dominant terms.

[0071] In some specific embodiments, the inertia determination module 15 further includes: An index construction unit is used to determine target nodes within the target area and construct a node dynamic contribution index for the target nodes based on the change in active power and the sampling time interval at each preset sampling time. The nodes include load nodes and frequency measurement nodes. A sensitivity determination unit is used to determine the node frequency sensitivity, which characterizes the degree of influence of node power changes on the center frequency of the region. The weight determination unit is used to determine the node inertia contribution weight of the target node based on the node dynamic contribution index and the node frequency sensitivity. The contribution determination unit is used to decompose the equivalent inertia of the target region based on the node inertia contribution weight to obtain the node equivalent inertia contribution of the target node; the node equivalent inertia contribution characterizes the degree of contribution of the node and its associated devices to the regional inertia support.

[0072] Furthermore, embodiments of this application also disclose an electronic device, Figure 5This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content of the diagram should not be construed as limiting the scope of this application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 stores a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the regional inertia identification method for power systems disclosed in any of the foregoing embodiments. Furthermore, the electronic device 20 in this embodiment may specifically be an electronic computer.

[0073] In this embodiment, the power supply 23 is used to provide operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and external devices, and the communication protocol it follows can be any communication protocol applicable to the technical solution of this application, and is not specifically limited here; the input / output interface 25 is used to acquire external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs, and is not specifically limited here.

[0074] In addition, the memory 22, as a carrier for resource storage, can be a read-only memory, random access memory, disk or optical disk, etc. The resources stored thereon can include operating system 221, computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0075] The operating system 221 is used to manage and control the various hardware devices on the electronic device 20 and the computer program 222, which may be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program capable of performing the regional inertia identification method of the power system executed by the electronic device 20 as disclosed in any of the foregoing embodiments, the computer program 222 may further include computer programs capable of performing other specific tasks.

[0076] Furthermore, this application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements the aforementioned method for regional inertia identification of a power system. Specific steps of this method can be found in the corresponding content disclosed in the foregoing embodiments, and will not be repeated here.

[0077] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0078] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0079] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0080] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0081] The technical solutions provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for identifying the regional inertia of a power system, characterized in that, include: The power system is divided into target regions, and corresponding regional center frequencies are constructed. The regional net acceleration power is constructed based on the power change of the power components within the target regions. The regional center frequency is an equivalent frequency variable that characterizes the overall frequency dynamics within the target regions. The regional net acceleration power is an equivalent power imbalance that causes frequency acceleration or deceleration within the target regions. The center frequency and net acceleration power of the region are preprocessed to obtain the processed frequency and processed power. Based on the processed frequency, the regional frequency deviation and the corresponding regional frequency change rate of the target region are determined. The equivalent frequency response relationship of the target region is constructed based on the processed power, the regional frequency deviation, the regional frequency change rate, the regional equivalent inertia and the regional equivalent damping of the target region. A candidate function library containing various candidate terms affecting the power system is constructed. A sparse dynamics identification method with regional frequency change rate as the identification target is used to screen dominant terms from the candidate function library. A regional frequency dynamic model is constructed based on the dominant terms, and the model coefficients of the regional frequency dynamic model are determined. The dominant terms are terms that play a predetermined decisive role in regional frequency dynamics. Based on the regional equivalent frequency response relationship, the model coefficients are mapped to the target region equivalent inertia and target region equivalent damping. Based on the target region equivalent inertia and target region equivalent damping of each target region, the regional inertia distribution result of the power system is determined, so as to trigger the regional frequency security analysis operation and regional inertia control operation of the power system based on the regional inertia distribution result.

2. The method for regional inertia identification of a power system according to claim 1, characterized in that, The preprocessing of the region's center frequency and the region's net acceleration power to obtain processed frequency and processed power includes: The center frequency and net acceleration power of the region are subjected to outlier removal, time alignment, filtering, normalization and smoothing differentiation to obtain the processed frequency and processed power.

3. The method for identifying the regional inertia of a power system according to claim 1, characterized in that, The construction of the regional net acceleration power based on the power change of the power components within the target region includes: The active power change of synchronous generators, the output change of new energy equipment, the power change of energy storage equipment, and the active power change of virtual synchronous generators or grid-type equipment within the target area are obtained to obtain the active power change on the power supply side. Obtain the load power change and tie line switching power change of the target area; The net acceleration power of the region is constructed based on the changes in active power on the power supply side, the changes in load power, and the changes in tie-line switching power.

4. The method for identifying the regional inertia of a power system according to claim 1, characterized in that, The candidate function library includes regional frequency deviation terms, regional net acceleration power terms, power response terms for various devices, tie-line power variation terms, and nonlinear correction terms.

5. The method for identifying the regional inertia of a power system according to claim 1, characterized in that, The method of sparse dynamics identification, which uses the rate of change of regional frequency as the identification target, to screen dominant terms from the candidate function library includes: The regional frequency change rate is determined as the identification target, and a sparse dynamic identification model is constructed based on the identification target and the sparse coefficients to be solved. An output vector is constructed based on the target region frequency change rate at sampling times within a preset disturbance time window, and a candidate function matrix is ​​constructed based on the candidate function values ​​corresponding to the sampling times; the candidate function values ​​are determined based on the candidate functions and the sampling data of the target region. The sparse dynamics identification problem is constructed based on the output vector, the candidate function matrix, the sparse coefficients, and the model residuals of the sparse dynamics identification model. Based on the output vector, the candidate function matrix, the sparse coefficients, and the preset ridge regression regularization parameters, a target optimization function is constructed. The target optimization function is minimized using the ridge regression least squares method to solve the sparse dynamics identification problem and obtain the target sparse coefficients. The coefficients in the target sparse coefficients whose absolute values ​​are less than a preset threshold are set to zero to obtain processed coefficients that only contain non-zero coefficients. Obtain the target candidate corresponding to the processed coefficient from the candidate function matrix, construct a new candidate function matrix based on the target candidate, and determine the dominant term in the candidate function library based on the new candidate function matrix and the ridge regression least squares method.

6. The method for regional inertia identification of a power system according to claim 1, characterized in that, The model coefficients include the coefficients of the regional net acceleration power term and the coefficients of the regional frequency deviation term in the dominant terms.

7. The method for regional inertia identification of a power system according to any one of claims 1 to 6, characterized in that, After determining the regional inertia distribution of the power system, the method further includes: Target nodes within the target area are identified, and a node dynamic contribution index for the target nodes is constructed based on the change in active power and the sampling time interval at each preset sampling time. The nodes include load nodes and frequency measurement nodes. Determine the node frequency sensitivity, which characterizes the degree of influence of node power changes on the center frequency of the region; The node inertia contribution weight of the target node is determined based on the node dynamic contribution index and the node frequency sensitivity. The equivalent inertia of the target region is decomposed based on the node inertia contribution weight to obtain the node equivalent inertia contribution of the target node; the node equivalent inertia contribution characterizes the degree of contribution of the node and its associated devices to the regional inertia support.

8. A regional inertia identification device for a power system, characterized in that, include: The region construction module is used to divide the power system into target regions, construct corresponding regional center frequencies, and construct regional net acceleration power based on the power change of power components within the target regions. The regional center frequency is an equivalent frequency variable characterizing the overall frequency dynamics within the target regions. The regional net acceleration power is an equivalent power imbalance that causes frequency acceleration or deceleration within the target regions. The preprocessing module is used to preprocess the center frequency of the region and the net acceleration power of the region respectively to obtain the processed frequency and the processed power, and to determine the regional frequency deviation and the corresponding regional frequency change rate of the target region based on the processed frequency. The relationship construction module is used to construct the regional equivalent frequency response relationship of the target region based on the processed power, the regional frequency deviation, the regional frequency change rate, the regional equivalent inertia and the regional equivalent damping of the target region; The dominant term identification module is used to construct a candidate function library containing various candidate terms that affect the power system, and to use a sparse dynamics identification method with the regional frequency change rate as the identification target to screen dominant terms from the candidate function library, and to construct a regional frequency dynamic model based on the dominant terms, and to determine the model coefficients of the regional frequency dynamic model; the dominant term is a term that plays a predetermined decisive role in the regional frequency dynamics. The inertia determination module is used to map the model coefficients to target area equivalent inertia and target area equivalent damping based on the regional equivalent frequency response relationship, and to determine the regional inertia distribution result of the power system based on the target area equivalent inertia and target area equivalent damping of each target area, so as to trigger the regional frequency security analysis operation and regional inertia control operation of the power system based on the regional inertia distribution result.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program to implement the regional inertia identification method for a power system as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, Used to store a computer program; wherein, when the computer program is executed by a processor, it implements the regional inertia identification method for a power system as described in any one of claims 1 to 7.

Images