An emergency coordinated support system for power grid frequency and voltage adapting to new energy fluctuation

Abstract

The application discloses an emergency coordinated support system for power grid frequency and voltage adapting to new energy fluctuation, and relates to the technical field of new energy power systems. After full-dimensional data are collected by a perception module, a pretreatment module obtains frequency / voltage fluctuation and power grid operation disturbance basic modes through mode decomposition and clustering. A demand quantification module builds a coupled nonlinear second-order model based on the basic modes, estimates unknown model deviation through a compensation function observer, selects effective support demand modes, and outputs node power angle support demand. A coordinated control module matches mode characteristics, dynamically allocates support resources according to numerical demand, and outputs output reference values. An execution module responds to frequency and voltage regulation demand through a model compensation control law, and outputs support power. The application realizes emergency coordinated support for frequency and voltage, improves the adaptability and anti-interference ability of the power grid to new energy fluctuation, and guarantees the safe and stable operation of the power grid under high new energy penetration.

Description

Technical Field

[0001] This invention relates to the field of new energy power system technology, and more specifically to an emergency coordinated support system for grid frequency and voltage that adapts to new energy fluctuations. Background Technology

[0002] As the global energy structure transitions towards clean and low-carbon energy, new energy power generation technologies such as wind power and photovoltaics have been widely adopted and applied, with their installed capacity accounting for a continuously increasing proportion of the power system, providing important support for achieving the "dual carbon" goal. However, new energy power generation has inherent intermittency, volatility, and uncertainty. Its output is easily affected by natural environmental factors such as wind speed, sunlight, and temperature, resulting in drastic fluctuations. This leads to a significant reduction in grid inertia and a weakening of frequency and voltage regulation capabilities, posing a severe challenge to the safe and stable operation of the power system.

[0003] In power systems with high penetration of new energy sources, the proportion of traditional synchronous generators is gradually decreasing, resulting in insufficient rotational inertia of the power grid. When faced with disturbances such as sudden changes in new energy output and load shocks, the frequency deviation and frequency change rate increase significantly, which can easily lead to frequency over-limit problems. At the same time, new energy units are usually connected to the grid through power electronic converters. Their weak damping and low inertia characteristics make the voltage at the grid connection point extremely sensitive to power fluctuations, and problems such as voltage drops and oscillations occur frequently. In severe cases, this may lead to large-scale disconnection of new energy units from the grid, thereby inducing cascading failures and threatening the safety of the power grid.

[0004] Therefore, how to achieve emergency coordinated support of grid frequency and voltage during new energy fluctuations is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] In view of this, the present invention provides an emergency coordinated support system for grid frequency and voltage to adapt to new energy fluctuations, thereby realizing emergency coordinated support for grid frequency and voltage during new energy fluctuations.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: An emergency coordinated support system for grid frequency and voltage that adapts to fluctuations in new energy sources includes: a sensing module, a preprocessing module, a demand quantification module, a coordinated control module, and an execution module; The sensing module collects comprehensive data on power grid operation, new energy output, and the environment, and outputs the comprehensive data to the preprocessing module. The preprocessing module performs modal decomposition and clustering on the full-dimensional data to obtain the fundamental modes of frequency fluctuation, voltage fluctuation, and power grid operation disturbance, and outputs them to the demand quantification module. The demand quantification module builds a grid frequency-voltage coupled nonlinear second-order model based on the basic modes, estimates the unknown model deviation of the system through the compensation function observer, and selects effective demand support modes by combining the correlation model. It outputs the total active power and total reactive power of the nodes to the collaborative control module to support the demand. The collaborative control module makes collaborative decisions based on the effective support demand mode, the total active power support demand of the nodes and the total reactive power support demand. It dynamically allocates support resources by formulating strategies through modal feature matching and determining the scale of numerical demand, and outputs the active power output reference value and reactive power output reference value of each device to the execution module. Based on the active and reactive power output reference values ​​of each device, the execution module adopts a model compensation control law to respond to frequency and voltage regulation needs, and outputs supporting power according to preset thresholds and priority scheduling rules to smooth out grid frequency and voltage fluctuations.

[0007] Preferably, the full-dimensional data includes: New energy data: real-time power output data, power output fluctuation amplitude, power output fluctuation frequency, and power output prediction time window data; Power grid operation data: Real-time power grid frequency Real-time voltage at each node Transmission line resistance Transmission line reactance With transmission line susceptance Real-time active power of load at each node and reactive power System state variables , ; Equipment parameter data: Converter station rated capacity Adjustment coefficient Action time constant and the inertial time constant of the new energy converter station ; Disturbance characteristic data: power grid frequency fluctuation deviation, voltage fluctuation deviation, transmission line parameter drift, deviation between renewable energy output and planned value, and power grid power flow distribution offset data.

[0008] Preferably, the preprocessing module specifically includes: The power grid operation data and new energy output data are initially decomposed using a variational mode decomposition algorithm to obtain several intrinsic mode functions (IMFs). The number of extreme points and standard deviation of each IMF are extracted, and characteristic trajectory curves are plotted. Several monotonic intervals are obtained by using the point where the derivative of the characteristic trajectory curve is 0 as the dividing point. Clustering is performed on the monotonic intervals, and the extreme points of each monotonic interval are extracted as the initial centers of the spherical clusters. The feature points within the monotonic intervals are spherically clustered using a preset monitoring radius, and abnormal feature points exceeding the monitoring radius are screened out. The abnormal feature points screened out by the spherical clusters are rectangularly clustered using a preset x-axis interval with fluctuation frequency as the dimension and a y-axis interval with fluctuation amplitude as the dimension. If an abnormal feature point falls within a rectangular interval, it is divided into several frequency fluctuation fundamental modes and voltage fluctuation fundamental modes. Those that do not fall into any rectangular interval are marked as unclassified modes. The disturbance feature data is subjected to nonlinear mapping transformation to obtain the basic value of power grid disturbance; the basic value of power grid disturbance is compared with the preset disturbance threshold range to divide it into the basic mode of weak power grid operation disturbance, the basic mode of medium power grid operation disturbance, and the basic mode of strong power grid operation disturbance.

[0009] Preferably, the construction of the power grid frequency-voltage coupled nonlinear second-order model based on the fundamental modes specifically includes: ; in, Given a known model function for the power grid, The unknown model bias is represented by d(t), and the unknown nonlinear disturbance of the power grid is represented by d(t). g ( x , t ) is an unknown nonlinear function of the power grid frequency-voltage coupling, u is the control quantity, and b is an empirical parameter; The above second-order model is rewritten as a state-space expression, and the unknown model deviation part is included. Expand to a new state The state-space form of the power grid frequency-voltage coupled system is obtained as follows:

[0010] In the formula, Let y represent the state of the power grid system, and y represent the output of the power grid.

[0011] Preferably, the compensation function observer specifically includes: Design a compensation function observer to estimate unknown model bias. It can be in one of the following two forms: Form 1, estimating only the unknown bias: ; Form 2, estimating the overall model: ; in, The gain matrix of the observer is the compensation function. The first gain coefficient of the observer. The second gain coefficient of the observer. Let be the error vector between the actual state of the power grid and the state of the observer. For the compensation term corresponding to the power grid operation disturbance, in the state-space model match; This is an estimate of the unknown model bias. Here, λ is the overall estimated value, and λ is the compensation coefficient. For the compensation function observer, the first Several observed state variables are used to track system state variables. ; Pole placement of the compensation function observer ensures the observation stability of the power grid system; based on the fundamental modes of the preprocessing module, an XGBoost correlation model is constructed, which cross-combines the fundamental modes of fluctuation with response characteristics, and selects effective demand support modes through recursive feature elimination and variance inflation factor; the load demand and power flow loss corresponding to the effective demand support modes are superimposed, combined with the estimated values ​​of unknown model biases. Correction: Output nodes always have active power to support demand. Total reactive power support requirements .

[0012] Preferably, the logical dynamic allocation of support resources through modal feature matching to formulate strategies and numerical requirements to determine scale specifically includes: obtaining effective support requirement modalities and the total active power support requirements of each node. Total reactive power support demand and the rated capacity of the converter station Frequency intensity parameters of each node in the power grid Voltage strength parameters Initial active power of nodes Initial reactive power of nodes This generates a complete set of input data for collaborative decision-making; and extracts wave feature vectors from the effective wave modes. Extracting perturbation feature vectors from the effective modes of operation perturbation Calculate the wave eigenvector With perturbation eigenvectors Based on the correlation, a modal similarity matrix is ​​constructed. The KM algorithm is used to perform modal matching with the highest modal similarity and the most urgent node support needs. Priority is given to matching high-demand nodes corresponding to the basic modes of strong power grid operation disturbances, and finally the optimal support mode pair is obtained.

[0013] Preferably, the output of active power output reference values ​​and reactive power output reference values ​​for each device specifically includes: summarizing the active and reactive power support requirements of all grid nodes to obtain the total active power support requirement of the system. Total reactive power support requirements of the system The frequency and voltage regulation coordination coefficient is calculated based on the power grid frequency and voltage intensity parameters, using the following formula: ; The active and reactive power allocation intervals of the converter station's rated capacity are divided according to the coordination coefficient, and the active power allocation capacity is: The reactive power distribution capacity is ;like and If the capacity of the converter station is sufficient, it is considered to have sufficient capacity; otherwise, it is considered to have insufficient capacity. In scenarios with sufficient capacity, power output reference values ​​are directly generated based on the actual needs of each node, including active power output reference values. Reference value for reactive power output ; In scenarios with insufficient capacity, resources are dynamically allocated based on the matching priority of the optimal support mode pair and the frequency and voltage regulation coordination coefficient, with the support requirements of each node as the base. The formula for calculating the reference values ​​of active and reactive power output is as follows: ; .

[0014] Preferably, the model compensation control law specifically includes: ; in, Given the desired frequency / voltage value, Given the desired value The second derivative with respect to time, , To adjust for errors, For virtual control variables, , >0 indicates that the controller adjusts the gain.

[0015] As can be seen from the above technical solution, compared with the prior art, the present invention discloses an emergency coordinated support system for grid frequency and voltage that adapts to the fluctuations of new energy sources. Through the hierarchical collaborative design of sensing, preprocessing, demand quantification, coordinated control, and execution modules, it first performs variational mode decomposition and clustering on all-dimensional data such as grid operation and new energy output to accurately extract the basic modes of frequency / voltage fluctuations and grid operation disturbances. Then, relying on the grid frequency-voltage coupled nonlinear second-order model and compensation function observer, it realizes real-time high-precision estimation of unknown model deviations. Combined with the XGBoost correlation model, it selects effective support demand modes. At the same time, it completes the dynamic allocation of support resources by using the logic of mode feature matching to determine the strategy and numerical demand to determine the scale. Finally, it realizes the precise response of frequency and voltage regulation through model compensation control law. It effectively solves the problems of reduced grid inertia and weakened frequency and voltage regulation capabilities caused by the intermittency and fluctuation of new energy output, greatly improves the grid's adaptability and anti-disturbance capability to new energy fluctuations, realizes emergency coordinated support for frequency and voltage, and effectively ensures the safe and stable operation of the power system in the scenario of high penetration of new energy. Attached Figure Description

[0016] 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.

[0017] Figure 1 The structural flowchart provided for this invention; Detailed Implementation

[0018] 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.

[0019] This invention discloses an emergency coordinated support system for power grid frequency and voltage to adapt to fluctuations in new energy sources, such as... Figure 1 As shown, it includes: a perception module, a preprocessing module, a demand quantification module, a collaborative control module, and an execution module; The sensing module collects comprehensive data on power grid operation, new energy output, and the environment, and outputs the comprehensive data to the preprocessing module. The preprocessing module performs modal decomposition and clustering on the full-dimensional data to obtain the fundamental modes of frequency fluctuation, voltage fluctuation, and power grid operation disturbance, and outputs them to the demand quantification module. The demand quantification module builds a grid frequency-voltage coupled nonlinear second-order model based on the basic modes, estimates the unknown model deviation of the system through the compensation function observer, and selects effective demand support modes by combining the correlation model. It outputs the total active power and total reactive power of the nodes to the cooperative control module to support the demand. The collaborative control module makes collaborative decisions based on the effective support demand mode, the total active power support demand of the nodes, and the total reactive power support demand. It dynamically allocates support resources by formulating strategies through modal feature matching and setting scales based on numerical requirements, and outputs the active power output reference values ​​and reactive power output reference values ​​of each device to the execution module. Based on the active and reactive power output reference values ​​of each device, the execution module adopts a model compensation control law to respond to frequency and voltage regulation needs, and outputs supporting power according to preset thresholds and priority scheduling rules to smooth out grid frequency and voltage fluctuations.

[0020] In one specific embodiment, the full-dimensional data includes: New energy data: real-time power output data, power output fluctuation amplitude, power output fluctuation frequency, and power output prediction time window data; Power grid operation data: Real-time power grid frequency Real-time voltage at each node Transmission line resistance Transmission line reactance With transmission line susceptance Real-time active power of load at each node and reactive power System state variables (Frequency / Voltage Response) (Frequency / Voltage Change Rate); Equipment parameter data: Converter station rated capacity Adjustment coefficient Action time constant and the inertial time constant of the new energy converter station ; Disturbance characteristic data: power grid frequency fluctuation deviation, voltage fluctuation deviation, transmission line parameter drift, deviation between renewable energy output and planned value, and power grid power flow distribution offset data.

[0021] In one specific embodiment, the preprocessing module specifically includes: The power grid operation data and new energy output data are initially decomposed using the variational mode decomposition algorithm to obtain several intrinsic mode functions. The number of extreme points and standard deviation of each IMF are extracted, and characteristic trajectory curves are plotted. The points where the derivative of the characteristic trajectory curve is 0 are used as the dividing points to obtain several monotonic intervals. Clustering is performed on monotonic intervals, and the extreme points of each monotonic interval are extracted as the initial centers of spherical clusters. A preset monitoring radius is used to perform spherical clustering on feature points within the monotonic intervals, and abnormal feature points exceeding the monitoring radius are screened out. A preset x-axis interval with fluctuation frequency as the dimension and y-axis interval with fluctuation amplitude as the dimension are used to perform rectangular clustering on the abnormal feature points screened out by spherical clustering. If an abnormal feature point falls within a rectangular interval, it is divided into several frequency fluctuation fundamental modes and voltage fluctuation fundamental modes. Those that do not fall into any rectangular interval are marked as unclassified modes. The disturbance characteristic data are subjected to nonlinear mapping transformation to obtain the basic value of power grid disturbance; the basic value of power grid disturbance is compared with the preset disturbance threshold range to divide it into the basic mode of weak power grid operation disturbance, the basic mode of medium power grid operation disturbance, and the basic mode of strong power grid operation disturbance.

[0022] In one specific embodiment, building a second-order nonlinear power grid frequency-voltage coupling model based on fundamental modes specifically includes: ; in, The known model function of the power grid (including power grid inertia, line parameters, and known load characteristics). The unknown model bias is represented by d(t), which represents the unknown nonlinear disturbances of the power grid (including sudden changes in new energy output, unknown load fluctuations, and random fault disturbances in transmission lines). g ( x , t ) is an unknown nonlinear function of the power grid frequency-voltage coupling (reflecting the nonlinear correlation characteristics of the power grid power angle, power flow and frequency / voltage), u is the control quantity (i.e. the power grid active / reactive power regulation quantity), and b is an empirical parameter; The above second-order model is rewritten as a state-space expression, and the unknown model deviation part is included. Expand to a new state The state-space form of the power grid frequency-voltage coupled system is obtained as follows:

[0023] In the formula, Let y represent the state of the power grid system, and y represent the output of the power grid.

[0024] In one specific embodiment, the compensation function observer specifically includes: Design a compensation function observer to estimate unknown model bias. It can be in one of the following two forms: Form 1, estimating only the unknown bias: ; Form 2, estimating the overall model: ; in, The gain matrix of the observer is the compensation function. The first gain coefficient of the observer. The second gain coefficient of the observer. Let be the error vector between the actual state of the power grid and the state of the observer. For the compensation term corresponding to the power grid operation disturbance, in the state-space model match; This is an estimate of the unknown model bias. Here, λ is the overall estimated value, and λ is the compensation coefficient. For the compensation function observer, the first Several observed state variables are used to track system state variables. ; Pole placement of the compensation function observer ensures the observation stability of the power grid system. Based on the fundamental modes of the preprocessing module, an XGBoost correlation model is constructed, which cross-combines the fluctuating fundamental modes with the power grid frequency / voltage response characteristics. Effective demand-supporting modes are selected through recursive feature elimination and variance inflation factor filtering. Load demand and power flow losses corresponding to the effective demand-supporting modes are then superimposed, combined with estimates of unknown model bias. Correction: Output nodes always have active power to support demand. Total reactive power support requirements .

[0025] In one specific embodiment, the logical dynamic allocation of support resources through modal feature matching to formulate strategies and numerical requirements to determine scale specifically includes: obtaining effective support requirement modes and the total active power support requirements of each node. Total reactive power support demand and the rated capacity of the converter station Frequency intensity parameters of each node in the power grid Voltage strength parameters Initial active power of nodes Initial reactive power of nodes This generates a complete set of input data for collaborative decision-making; and extracts wave feature vectors from the effective wave modes. Extracting perturbation feature vectors from the effective modes of operation perturbation Calculate the wave eigenvector With perturbation eigenvectors Based on the correlation, a modal similarity matrix is ​​constructed. The KM algorithm is used to perform modal matching with the highest modal similarity and the most urgent node support needs. Priority is given to matching high-demand nodes corresponding to the basic modes of strong power grid operation disturbances, and finally the optimal support mode pair is obtained.

[0026] In one specific embodiment, outputting the active power output reference values ​​and reactive power output reference values ​​for each device specifically includes: summarizing the active power and reactive power support requirements of all grid nodes to obtain the total active power support requirement of the system. Total reactive power support requirements of the system The frequency and voltage regulation coordination coefficient is calculated based on the power grid frequency and voltage intensity parameters, using the following formula: ; The active and reactive power allocation intervals of the converter station's rated capacity are divided according to the coordination coefficient, and the active power allocation capacity is: The reactive power distribution capacity is ;like and If the capacity of the converter station is sufficient, it is considered to have sufficient capacity; otherwise, it is considered to have insufficient capacity. In scenarios with sufficient capacity, power output reference values ​​are directly generated based on the actual needs of each node, including active power output reference values. Reference value for reactive power output ; In scenarios with insufficient capacity, resources are dynamically allocated based on the matching priority of the optimal support mode pair and the frequency and voltage regulation coordination coefficient, with the support requirements of each node as the base. The formula for calculating the reference values ​​of active and reactive power output is as follows: ; .

[0027] In one specific embodiment, the model compensation control law specifically includes: ; in, Given the desired frequency / voltage value, Given the desired value The second derivative with respect to time, , To adjust for errors, For virtual control variables, , >0 indicates that the controller adjusts the gain.

[0028] 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. The methods disclosed in the embodiments are described simply because they correspond to the methods disclosed in the embodiments; relevant parts can be found in the method section.

[0029] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An emergency coordinated support system for power grid frequency and voltage to adapt to fluctuations in new energy sources, characterized in that, include: The module comprises a perception module, a preprocessing module, a demand quantification module, a collaborative control module, and an execution module. The sensing module collects comprehensive data on power grid operation, new energy output, and the environment, and outputs the comprehensive data to the preprocessing module. The preprocessing module performs modal decomposition and clustering on the full-dimensional data to obtain the fundamental modes of frequency fluctuation, voltage fluctuation, and power grid operation disturbance, and outputs them to the demand quantification module. The demand quantification module builds a grid frequency-voltage coupled nonlinear second-order model based on the basic modes, estimates the unknown model deviation of the system through the compensation function observer, and selects effective demand support modes by combining the correlation model. It outputs the total active power and total reactive power of the nodes to the collaborative control module to support the demand. The collaborative control module makes collaborative decisions based on the effective support demand mode, the total active power support demand of the nodes and the total reactive power support demand. It dynamically allocates support resources by formulating strategies through modal feature matching and determining the scale of numerical demand, and outputs the active power output reference value and reactive power output reference value of each device to the execution module. Based on the active and reactive power output reference values ​​of each device, the execution module adopts a model compensation control law to respond to frequency and voltage regulation needs, and outputs supporting power according to preset thresholds and priority scheduling rules to smooth out grid frequency and voltage fluctuations.

2. The emergency coordinated support system for power grid frequency and voltage to adapt to new energy fluctuations as described in claim 1, characterized in that, The full-dimensional data includes: New energy data: real-time power output data, power output fluctuation amplitude, power output fluctuation frequency, and power output prediction time window data; Power grid operation data: Real-time power grid frequency Real-time voltage at each node Transmission line resistance Transmission line reactance With transmission line susceptance Real-time active power of load at each node and reactive power System state variables , ; Equipment parameter data: Converter station rated capacity Adjustment coefficient Action time constant and the inertial time constant of the new energy converter station ; Disturbance characteristic data: power grid frequency fluctuation deviation, voltage fluctuation deviation, transmission line parameter drift, deviation between renewable energy output and planned value, and power grid power flow distribution offset data.

3. The emergency coordinated support system for power grid frequency and voltage to adapt to new energy fluctuations according to claim 1, characterized in that, The preprocessing module specifically includes: The power grid operation data and new energy output data are initially decomposed using a variational mode decomposition algorithm to obtain several intrinsic mode functions (IMFs). The number of extreme points and standard deviation of each IMF are extracted, and characteristic trajectory curves are plotted. Several monotonic intervals are obtained by using the point where the derivative of the characteristic trajectory curve is 0 as the dividing point. Clustering is performed on the monotonic intervals, and the extreme points of each monotonic interval are extracted as the initial centers of the spherical clusters. The feature points within the monotonic intervals are spherically clustered using a preset monitoring radius, and abnormal feature points exceeding the monitoring radius are screened out. The abnormal feature points screened out by the spherical clusters are rectangularly clustered using a preset x-axis interval with fluctuation frequency as the dimension and a y-axis interval with fluctuation amplitude as the dimension. If an abnormal feature point falls within a rectangular interval, it is divided into several frequency fluctuation fundamental modes and voltage fluctuation fundamental modes. Those that do not fall into any rectangular interval are marked as unclassified modes. The disturbance feature data is subjected to nonlinear mapping transformation to obtain the basic value of power grid disturbance; the basic value of power grid disturbance is compared with the preset disturbance threshold range to divide it into the basic mode of weak power grid operation disturbance, the basic mode of medium power grid operation disturbance, and the basic mode of strong power grid operation disturbance.

4. The emergency coordinated support system for power grid frequency and voltage to adapt to new energy fluctuations according to claim 2, characterized in that, The construction of the power grid frequency-voltage coupled nonlinear second-order model based on the fundamental modes specifically includes: ； in, Given a known model function for the power grid, The unknown model bias is represented by d(t), and the unknown nonlinear disturbance of the power grid is represented by d(t). g ( x , t ) is an unknown nonlinear function of the power grid frequency-voltage coupling, u is the control quantity, and b is an empirical parameter; The above second-order model is rewritten as a state-space expression, and the unknown model deviation part is included. Expand to a new state The state-space form of the power grid frequency-voltage coupled system is obtained as follows: In the formula, Let y represent the state of the power grid system, and y represent the output of the power grid.

5. The emergency coordinated support system for power grid frequency and voltage to adapt to new energy fluctuations according to claim 4, characterized in that, The compensation function observer specifically includes: Design a compensation function observer to estimate unknown model bias. It can be in one of the following two forms: Form 1, estimating only the unknown bias: ； Form 2, estimating the overall model: ； in, The gain matrix of the observer is the compensation function. The first gain coefficient of the observer. The second gain coefficient of the observer. Let be the error vector between the actual state of the power grid and the state of the observer. For the compensation term corresponding to the power grid operation disturbance, in the state-space model match; This is an estimate of the unknown model bias. Here, λ is the overall estimated value, and λ is the compensation coefficient. For the compensation function observer, the first Several observed state variables are used to track system state variables. ; Pole placement of the compensation function observer ensures the observation stability of the power grid system; based on the fundamental modes of the preprocessing module, an XGBoost correlation model is constructed, which cross-combines the fundamental modes of fluctuation with response characteristics, and selects effective demand support modes through recursive feature elimination and variance inflation factor; the load demand and power flow loss corresponding to the effective demand support modes are superimposed, combined with the estimated values ​​of unknown model biases. Correction: Output nodes always have active power to support demand. Total reactive power support requirements .

6. The emergency coordinated support system for power grid frequency and voltage to adapt to new energy fluctuations according to claim 5, characterized in that, The logical dynamic allocation of support resources through modal feature matching to formulate strategies and numerical requirements to determine scale specifically includes: obtaining effective support requirement modalities and the total active power support requirements of each node. Total reactive power support demand and the rated capacity of the converter station Frequency intensity parameters of each node in the power grid Voltage strength parameters Initial active power of nodes Initial reactive power of nodes This generates a complete set of input data for collaborative decision-making; and extracts wave feature vectors from the effective wave modes. Extracting perturbation feature vectors from the effective modes of operation perturbation Calculate the wave eigenvector With perturbation eigenvectors Based on the correlation, a modal similarity matrix is ​​constructed. The KM algorithm is used to perform modal matching with the highest modal similarity and the most urgent node support needs. Priority is given to matching high-demand nodes corresponding to the basic modes of strong power grid operation disturbances, and finally the optimal support mode pair is obtained.

7. The emergency coordinated support system for power grid frequency and voltage to adapt to new energy fluctuations according to claim 6, characterized in that, The output of active power output reference values ​​and reactive power output reference values ​​for each device specifically includes: summarizing the active power and reactive power support requirements of all grid nodes to obtain the total active power support requirement of the system. Total reactive power support requirements of the system The frequency and voltage regulation coordination coefficient is calculated based on the power grid frequency and voltage intensity parameters, using the following formula: ; The active and reactive power allocation intervals of the converter station's rated capacity are divided according to the coordination coefficient, and the active power allocation capacity is: The reactive power distribution capacity is ;like and If the capacity of the converter station is sufficient, it is considered to have sufficient capacity; otherwise, it is considered to have insufficient capacity. In scenarios with sufficient capacity, power output reference values ​​are directly generated based on the actual needs of each node, including active power output reference values. Reference value for reactive power output ; In scenarios with insufficient capacity, resources are dynamically allocated based on the matching priority of the optimal support mode pair and the frequency and voltage regulation coordination coefficient, with the support requirements of each node as the base. The formula for calculating the reference values ​​of active and reactive power output is as follows: ； 。 8. The emergency coordinated support system for power grid frequency and voltage to adapt to new energy fluctuations according to claim 7, characterized in that, The model compensation control law specifically includes: ； in, Given the desired frequency / voltage value, Given the desired value The second derivative with respect to time, , To adjust for error, For virtual control variables, , >0 indicates that the controller adjusts the gain.

Images