Frequency instability risk quantitative evaluation method of wind power grid-connected power system in steady state considering active disturbance probability

Abstract

The invention discloses a steady-state wind power grid-connected power system frequency instability risk quantitative evaluation method considering an active disturbance probability, and belongs to the field of power system stability analysis. The method comprises the following steps: inputting steady-state power flow information of each grid-connected power generation cluster into an XGBoost-based speed regulation system frequency modulation parameter identification model to obtain frequency modulation parameters, and obtaining a primary frequency modulation polynomial equivalent analysis model of each cluster; simplifying a calculation formula of an inertia center frequency deviation extreme value based on an active-frequency open-loop decoupling model; setting a multi-level inertia safety boundary by referring to various active disturbance occurrence probabilities; and establishing a piecewise linear thought-based steady-state frequency stability margin quantitative evaluation index for online evaluation of the inertia level of the system, and analyzing the steady-state system frequency stability margin and the frequency instability risk in real time. According to the method, the influence of wind power integration on the frequency response process of the power system is considered, online real-time accurate evaluation of the system frequency stability margin in a steady state is realized, and a reference basis is provided for formulating frequency stability control measures.

Description

Technical Field

[0001] This invention belongs to the field of power system stability analysis, and more specifically, relates to a method for quantitatively assessing the frequency instability risk of wind-powered grid-connected power systems under steady-state conditions, taking into account the probability of active power disturbances. Background Technology

[0002] With the rapid advancement of wind power technology, traditional power systems dominated by synchronous generators are accelerating their transformation into new power systems with a high proportion of wind power connected to the grid.

[0003] However, wind power equipment differs significantly from traditional synchronous turbines in terms of frequency regulation characteristics. Specifically, wind power cannot actively provide inertia support for the system, which profoundly alters the dynamic frequency response of the power grid. The high proportion of wind power integration has led to a series of new characteristics in the system's frequency response, including an overall reduction in system inertia, more diverse frequency regulation methods, more complex response mechanisms, and increased spatial variability in frequency distribution. These changes pose entirely new challenges to the frequency stability control of the power system.

[0004] Meanwhile, wide-area measurement technologies, represented by phasor measurement units (PMUs), are widely used in power grids, forming a key infrastructure for the underlying sensing of new power systems. PMUs rely on high-speed communication networks to interact with the dispatch master station and, together with the existing supervisory control and data acquisition (SCADA) system, construct a hybrid measurement system covering the entire network. This system provides a rich data foundation for real-time assessment and intelligent analysis of system frequency stability in wind power grid-connected environments, and also opens up new paths for building data-driven stability assessment methods.

[0005] For frequency stability assessment of systems operating stably under undisturbed or minimally disturbed conditions, current research often employs the following approach: first, a series of anticipated disturbances are set up; then, the frequency response characteristics or variation curves of the steady-state system under each anticipated disturbance are predicted; finally, evaluation indices are calculated to analyze the potential frequency instability risk of the system. Regarding the establishment of a system of frequency stability assessment indices under steady-state conditions, many recent studies have used key frequency response characteristics such as the maximum rate of frequency change, the maximum frequency offset, and the steady-state frequency deviation as basic frequency stability assessment indices. However, current research does not comprehensively consider all frequency instability safety constraints, nor does it analyze the frequency instability risk of the system under all possible active power disturbances during steady-state operation. Therefore, it is necessary to further optimize existing methods for quantitatively assessing the frequency instability risk of wind-powered grid-connected power systems under steady-state conditions and improve the performance of these methods in practical large-scale wind-powered grid-connected power systems. Summary of the Invention

[0006] To address the aforementioned deficiencies or improvement needs of existing technologies, this invention provides a quantitative assessment method for the frequency instability risk of wind-powered grid-connected power systems under steady-state conditions, taking into account the probability of active power disturbances. This method enables accurate online assessment of the frequency stability margin of wind-powered grid-connected power systems, providing an important reference for the formulation of subsequent frequency stability control measures.

[0007] To achieve the above objectives, according to a first aspect of the present invention, a method for quantitatively assessing the frequency instability risk of a wind-powered grid-connected power system under steady-state conditions, taking into account the probability of active power disturbances, is provided, comprising: S1, after the disturbance occurs, the steady-state power flow information of each power generation cluster in the system is input into the pre-trained speed regulation system frequency regulation parameter identification model based on XGBoost to obtain the speed regulation system frequency regulation parameters of each power generation cluster, so as to construct the equivalent analysis model transfer function of each power generation cluster. S2, combining the transfer function of the equivalent analysis model of each power generation cluster with the active-frequency open-loop decoupling model based on the inertia center, the relationship between the extreme value of the frequency deviation of the inertia center and the magnitude of the active disturbance and the equivalent inertia level is simplified in order to calculate the extreme value of the frequency deviation of the inertia center. S3, obtain the probability curve of active power disturbance that the system may be subjected to under the current stable operating state, calculate the first multi-level inertia safety boundary in combination with the extreme value safety constraint of the inertia center frequency deviation, and calculate the second multi-level inertia safety boundary in combination with the safety constraint of the maximum rate of change of the inertia center frequency. S4 calculates the frequency instability risk quantification index corresponding to the real-time equivalent inertia level of the system under the first and second multi-level inertia safety boundaries, respectively. S5, take the minimum value of the frequency instability risk quantification assessment index calculated in S4 as the final frequency instability risk quantification assessment result of the system.

[0008] According to a second aspect of the present invention, an electronic device is provided, comprising: a computer-readable storage medium and a processor; The computer-readable storage medium is used to store executable instructions; The processor is configured to read executable instructions stored in the computer-readable storage medium and execute the method as described in the first aspect.

[0009] According to a third aspect of the invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to perform the method as described in the first aspect.

[0010] According to a fourth aspect of the invention, a computer program product is provided, comprising a computer program or instructions that, when executed by a processor, implement the method described in the first aspect.

[0011] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects: 1. Traditional steady-state frequency stability assessment models are mostly based on a single preset active power disturbance scenario. In order to comprehensively analyze the frequency instability risk under various anticipated active power disturbances, a large number of model sets need to be established, thus limiting their application in complex power systems with wind power grid connection. The method provided by this invention analyzes the occurrence probability of various active power disturbances and sets multi-level inertia safety boundaries, realizing a comprehensive assessment of the frequency stability of steady-state systems that may have various active power disturbances using a single analysis model, which is more valuable for practical application.

[0012] 2. The method provided by this invention analyzes the impact of large-scale grid connection of wind power generation equipment on the system frequency response process and the resulting frequency response characteristics. Considering the frequency regulation control measures of wind power generation equipment, a simplified analysis model for wind power frequency regulation is established. This model can be applied to the online quantitative assessment of frequency instability risk in the current wind power grid-connected power system, making it more practical.

[0013] 3. The method provided by this invention is based on the center of inertia theory. It deconstructs the traditional active-frequency closed-loop analysis model and establishes an active-frequency open-loop decoupled model, which greatly reduces the dimensionality of the calculation and analysis. While ensuring a certain prediction accuracy, it significantly improves the calculation speed and can meet the analysis speed requirements of online evaluation. Attached Figure Description

[0014] Figure 1 A flowchart illustrating a method for quantitatively assessing the steady-state frequency instability risk taking into account the probability of active power disturbances, provided in an embodiment of the present invention.

[0015] Figure 2 To improve the ASF model and the active-frequency open-loop decoupling analysis model diagram.

[0016] Figure 3 This is a diagram of the IEEE 39-node system including wind power.

[0017] Figure 4 A comparison chart of prediction errors for frequency response characteristic quantities.

[0018] Figure 5 This is a comparison chart of the comprehensive prediction errors of the extreme values ​​of the inertia center frequency deviation.

[0019] Figure 6 This is a comparison chart of the comprehensive prediction errors for the maximum rate of change of the inertia center frequency.

[0020] Figure 7 The figure shows the evaluation results of the FSAI-I1 index under different inertia levels.

[0021] Figure 8The figure shows the evaluation results of the FSAI-I2 index under different inertia levels. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0023] This invention provides a method for quantitatively assessing the frequency instability risk of a wind-powered grid-connected power system under steady-state conditions, taking into account the probability of active power disturbances. The method includes: S1. The steady-state power flow information of each power generation cluster after the disturbance occurs is input into the pre-trained speed regulation system frequency regulation parameter identification model based on XGBoost to obtain the speed regulation system frequency regulation parameters of each power generation cluster, so as to construct the equivalent analysis model transfer function of each power generation cluster.

[0024] Specifically, in step S1, a frequency regulation parameter identification model for the speed regulation system based on the extreme gradient boosting (XGBoost) algorithm is trained offline. Then, the identified parameters are used to construct the transfer function of the polynomial equivalent analysis model for each power generation cluster, including: S11, Parameter identification and sample set generation are performed offline. The transfer function of the equivalent analysis model of the frequency response process of the speed regulation system of each power generation cluster in each sample is fitted using the least squares method to obtain the sample labels of the frequency regulation parameters of the speed regulation system, i.e., the output features of the XGBoost model. The input features of the XGBoost model under different samples are a mixture of steady-state power flow information of each power generation cluster in the power system, including the output active power P of each power generation cluster before and after the disturbance. G Generator output reactive power Q G Active power of load P L Reactive power of load Q L Voltage amplitude V and voltage phase angle δ.

[0025] S12: Based on the input features and labels of the sample set generated in step S11, train the XGBoost parameter identification model offline. Test the model performance, and save the model with the best test performance after multiple parameter tunings.

[0026] S13. Using the steady-state power flow information of each power generation cluster after the disturbance occurs, the frequency regulation parameters of the speed regulation system of each power generation cluster are obtained based on the trained XGBoost parameter identification model, and the transfer function of the polynomial equivalent analysis model of each power generation cluster is constructed.

[0027] For each synchronous power generation cluster, the equivalent analysis model of its speed regulation system is as follows: .in, G i,gen (s) is i The transfer function of the equivalent analysis model of the frequency response process of the speed regulation system of the No. 1 synchronous power generation cluster. k i,0 , k i,1 and k i,2 They are respectively i The transfer function parameters of the equivalent analysis model of the synchronous generator cluster are the frequency regulation parameters of the speed regulation system of the synchronous generator cluster; for each grid-connected wind turbine generator cluster, the equivalent analysis model of its speed regulation system is as follows: .in, G i,wind (s) is i The transfer function of the equivalent analysis model of the active frequency support process of the No. 1 wind power cluster. J i,NE , D i,NE and K i,NE They are respectively i The relevant parameters of the transfer function of the equivalent analysis model of the No. 1 wind power cluster are the frequency regulation parameters of the wind power cluster speed regulation system.

[0028] S2, combining the transfer function of the equivalent analysis model of each power generation cluster with the active-frequency open-loop decoupling model based on the inertia center, the relationship between the extreme value of the frequency deviation of the inertia center and the magnitude of the active disturbance and the equivalent inertia level is simplified in order to calculate the extreme value of the frequency deviation of the inertia center.

[0029] Specifically, in step S2, an open-loop decoupling model of the active power-frequency system based on the inertia center is established, and a simplified relationship between the extreme value of the frequency deviation of the inertia center and the magnitude of the active power disturbance and the equivalent inertia level is derived, including: S21, compared with the improved ASF closed-loop frequency analysis model, the active power-frequency open-loop decoupling analysis model is as follows: Figure 2 As shown. Among them, This refers to the total power deficit suffered by the system. This refers to the transient active power compensation for all grid-connected power generation equipment. For the transient mechanical power response changes of all synchronous power generation equipment; They are 1~ m The conversion factor between the base capacity of an individual synchronous generator cluster and the total base capacity of the system. They are 1~ m Transfer function of the equivalent model of the speed regulation system of a synchronous generator clusterm This refers to the total number of grid-connected synchronous generator sets; for m +1~ m + n The conversion factor between the baseline capacity of the No. 1 wind power cluster and the total baseline capacity of the system. They are respectively m +1~ m + n Transfer function of the frequency regulation equivalent model of the No. 1 wind power cluster n This refers to the total number of grid-connected wind power generation clusters; H sys For the system's equivalent inertia, This is the frequency response curve of the system's center of inertia.

[0030] S22, the time from the instant the system is subjected to an active disturbance to the point where the frequency deviation reaches its extreme value, is considered to be represented by a linear function with a constant slope. A simplified modeling analysis is performed on the changes in the total active power compensation for primary frequency regulation of the system (including the total mechanical power compensation of the synchronous generator cluster governor and the total active power frequency regulation compensation of the wind power cluster). Among these, This represents the total active power change due to primary frequency regulation compensation in the system. This refers to the time it takes for the center frequency deviation of the inertia to reach its extreme value after an active disturbance occurs.

[0031] S23, using a step function Approximate fitting of the total electromagnetic power deficit in wind power grid-connected power systems According to the rotor motion equation The expression for the center frequency deviation of inertia can be derived. .in, It is the initial moment when the active disturbance occurs. It is an unknown parameter.

[0032] S24, by applying a Laplace transform to the quadratic function approximation of the inertia center frequency deviation, we obtain... By feeding it into the equivalent analysis model of the first-order frequency modulation polynomial of the system, the mechanical power output in the frequency domain can be obtained as follows: Wherein, in the formula, G sys ( s The sum of the transfer functions for the primary frequency regulation polynomial equivalent analysis model of all grid-connected power generation equipment is given by: .

[0033] S25, when the inertia center frequency deviation reaches its extreme value, the total transient active power compensation of each grid-connected power generation cluster in the system is equal to the transient active power deficit of the system. This can be determined by solving... The equation was derived, and the extreme value of the inertia center frequency deviation was obtained. , where Δ f max The extreme value of the frequency deviation of the system's inertia center. k sys,1 and k sys,2 The equivalent model parameters of the primary frequency modulation of the system are respectively.

[0034] S3, obtain the probability curve of possible active disturbances that the system may be subjected to under the current stable operating state, calculate the first multi-level inertia safety boundary by combining the extreme value safety constraint of the inertia center frequency deviation, and calculate the second multi-level inertia safety boundary by combining the safety constraint of the maximum rate of change of the inertia center frequency.

[0035] Specifically, in step S3, the various active power disturbance faults that may occur in the steady-state operating system in the future and their probabilities are analyzed. Based on the disturbance probability curve, combined with the simplified relationship of the extreme value of the inertia center frequency deviation and the frequency safety constraint, a multi-level inertia safety boundary based on the frequency deviation extreme value safety constraint is calculated. The same method is used to calculate the multi-level inertia safety boundary based on the active power disturbance probability curve for the maximum rate of frequency change. H rocof,limit1 , H rocof,limit2 , H rocof,limit3 These indicate the critical equivalent inertia levels required to ensure the maximum rate of change of the center frequency of inertia meets safety constraints under 20%, 70%, and 100% of the anticipated active power disturbances, respectively, including: S31, by analyzing the operating status of the wind power grid-connected power system, predicts the probability of various active power disturbances that the system may suffer under the current stable operating state, and summarizes the disturbance probability curve.

[0036] S32, select the magnitude cutoff points for active disturbances with occurrence probabilities of 20%, 70%, and 100%, respectively. P 1. P 2. P 3. That is, by analyzing the real-time operating conditions of the current steady-state system, the magnitude of the active power disturbance that the stable operating system will suffer in the future is less than or equal to... P The probability of 1 is 20%, and the magnitude of the active disturbance is less than or equal to P The probability of 2 is 70%, and the magnitude of the active disturbance is less than or equal to P The probability of 3 is 100%.

[0037] If a system in steady-state operation can maintain frequency stability under as many anticipated active power disturbances as possible, it indicates a higher frequency stability margin. Based on this idea, combined with... P 1. P 2. P3. These three active disturbance probability dividing points are used to further set the inertia safety boundary of the frequency stability margin assessment index.

[0038] S33, based on P 1. P 2. P 3. Based on the probability thresholds of three active disturbances and the established frequency deviation extreme value safety constraints, the critical active disturbance magnitudes that the system can withstand under the fixed inertia center frequency deviation extreme value constraints are calculated using the simplified relationship of the inertia center frequency deviation extreme value derived above. H fmax,limit1 , H fmax,limit2 , H fmax,limit3 This serves as a multi-level inertia safety boundary (i.e., the first multi-level inertia safety boundary) based on the extreme value safety constraint of the inertia center frequency deviation for a stable operating system.

[0039] S34, According to the rotor motion equation, the initial change rate of the center frequency of the inertia at the instant the active disturbance occurs is... Under the premise that the total active power disturbance of the system changes in an approximate step response form, the initial rate of change of the inertia center frequency is the maximum rate of change of frequency during the process from the moment the disturbance occurs until the frequency reaches its extreme value.

[0040] S35, using the same method described above, calculate the multi-level inertia safety boundary based on the active disturbance probability curve for the maximum rate of frequency change. H rocof,limit1 , H rocof,limit2 , H rocof,limit3 (i.e., the second multi-level inertia safety boundary), which respectively indicate the critical equivalent inertia level at which the maximum rate of change of the inertia center frequency satisfies the safety constraints under 20%, 70% and 100% of the expected active disturbance.

[0041] S4 calculates the frequency instability risk quantification index corresponding to the real-time equivalent inertia level of the system under the first and second multi-level inertia safety boundaries.

[0042] Specifically, in step S4, based on the piecewise linear approach, considering the safety constraints of frequency deviation extrema and maximum rate of change of frequency, the frequency stability assessment index based on inertia level (FSAI-I) corresponding to the real-time equivalent inertia level of the system under multi-level inertia safety boundaries is calculated, including: S41, firstly, based on the piecewise linear approach and considering the frequency deviation extreme value safety constraint, calculates the quantitative assessment index of frequency instability risk corresponding to the real-time equivalent inertia level of the system under multi-level inertia safety boundaries.

[0043] S42, also based on the piecewise linear idea and considering the constraint of the maximum rate of change of frequency, calculates the quantitative assessment index of frequency instability risk corresponding to the real-time equivalent inertia level of the system under multi-level inertia safety boundaries.

[0044] S5, take the minimum value of the frequency instability risk quantification assessment index calculated in S4 as the final frequency instability risk quantification assessment result of the system.

[0045] Specifically, during online frequency stability assessment, the inertia level of the wind power grid-connected power system is evaluated and updated in real time. Finally, the FSAI-I index is calculated online, and the quantitative assessment results of system frequency instability risk under steady state are analyzed. The minimum value between frequency instability risk quantitative assessment index 1 and index 2 is taken as the multi-dimensional fusion frequency instability risk quantitative assessment index. .

[0046] The higher the equivalent inertia level of a stable operating system, the larger the FSAI-I1 and FSAI-I2, the larger the FSAI-I multi-dimensional comprehensive evaluation index, the greater the frequency stability margin of the stable operating system, and the lower the risk of frequency instability.

[0047] When the system's equivalent inertia is less than H fmax,limit1 (or H rocof,limit1 When the system's effective inertia is less than 20% in the face of various possible active power disturbances, it indicates weak disturbance rejection capability and a high risk of frequency instability. Therefore, the FSAI-I is 0%, and the system's steady state is "unsafe." When the system's equivalent inertia is greater than... H fmax,limit1 (or H rocof,limit1 (less than) H fmax,limit2 (or H rocof,limit2 When faced with active disturbances with a probability of 20% to 80%, the system can maintain frequency stability, and the larger the equivalent inertia, the stronger the system's disturbance rejection capability. Therefore, adopting... H fmax,limit1 (or H rocof,limit1 )and H fmax,limit2 (or H rocof,limit2The two safety boundaries linearly normalize the system's equivalent inertia, quantifying it into a unified FSAI-I evaluation index within the range of 0% to 100%. At this point, the stable operating system is considered "relatively safe." When the system's equivalent inertia is greater than... H fmax,limit2 (or H rocof,limit2 (less than) H fmax,limit3 (or H rocof,limit3 When faced with active power disturbances with a probability of 80%–100%, the system can maintain frequency stability. Using the same linearization method, the system's equivalent inertia level is uniformly quantified into an FSAI-I evaluation index within the range of 100%–200%. At this point, the stable operating state of the system is considered "safe." Finally, when the system's equivalent inertia is greater than... H fmax,limit3 (or H rocof,limit3 When the frequency is stable, the system can maintain frequency stability under all possible active power fault scenarios, so the frequency stability state is "absolutely safe" and FSAI-I is 200%.

[0048] Unlike traditional synchronous generators, wind power equipment exhibits different frequency regulation characteristics. Its large-scale grid connection brings new changes to the frequency response characteristics of the power system. A detailed analysis of the impact of wind power equipment on the frequency response process is conducted to ensure that the calculation results are applicable to engineering practice.

[0049] The method provided by this invention is applied to an improved IEEE 39-node system incorporating wind power: the system's rated frequency is 50Hz. Wind farms 1 through 5 are connected to nodes 3, 4, 16, 17, and 18 of the original IEEE 39-node system, respectively. Each wind farm consists of multiple doubly-fed induction generators (DFIGs) with a rated capacity of 2MW. The DFIGs employ a third-order model that ignores stator transients. Each DFIG participates in the frequency regulation process using droop control and virtual inertia control measures. The input to each turbine's control measures is the transient frequency of the grid connection point of the synchronous generator (No. 1). PMU devices are deployed at the grid connection nodes of the synchronous generators and wind farms, with the PMU sampling frequency set to 100Hz. The total initial load of the system is 6254.23MW. The improved IEEE 39-node system incorporating wind power is as follows: Figure 3 As shown. The present invention will now be used to evaluate the frequency stability margin of the wind power grid-connected power system under steady-state conditions.

[0050] The present invention provides a method for quantitatively assessing the frequency instability risk of wind-powered grid-connected power systems under steady-state conditions, taking into account the probability of active power disturbances, as follows: Figure 1 As shown, it includes the following steps: (1) The frequency regulation parameter identification model of the speed regulation system based on the extreme gradient boosting (XGBoost) algorithm is obtained by training offline. Then, the polynomial equivalent analysis model transfer function of each power generation cluster is constructed using the identified parameters: (11) Parameter identification and sample set generation are performed offline. The transfer function of the equivalent analysis model of the frequency response process of the speed regulation system of each power generation cluster in each sample is fitted by least squares method to obtain the sample label of the frequency regulation parameter of the speed regulation system, that is, the output feature of the XGBoost model. The input feature of the XGBoost model under different samples is a mixture of steady-state power flow information of each power generation cluster in the power system, including the output active power P of each power generation cluster before and after the disturbance. G Generator output reactive power Q G Active power of load P L Reactive power of load Q L Voltage amplitude V and voltage phase angle δ.

[0051] (12) Based on the input features and labels of the sample set generated in step (11), train the XGBoost parameter identification model offline. Test the model performance, and save the model with the best test performance after multiple parameter tunings.

[0052] (13) Using the steady-state power flow information of each power generation cluster after the disturbance occurs, the frequency regulation parameters of the speed regulation system of each power generation cluster are obtained based on the trained XGBoost parameter identification model, and the transfer function of the polynomial equivalent analysis model of each power generation cluster is constructed.

[0053] For each synchronous power generation cluster, the equivalent analysis model of its speed regulation system is as follows: .in, G i,gen (s) is i The transfer function of the equivalent analysis model of the frequency response process of the speed regulation system of the No. 1 synchronous power generation cluster. k i,0 , k i,1 and k i,2 They are respectively i The transfer function parameters of the equivalent analysis model of the synchronous generator cluster are as follows: For each grid-connected wind turbine generator cluster, the equivalent analysis model of its speed regulation system is as follows: .in, G i,wind (s) is i The transfer function of the equivalent analysis model of the active frequency support process of the No. 1 wind power cluster. J i,NE , Di,NE and K i,NE They are respectively i The relevant parameters of the transfer function of the equivalent analysis model of the No. 1 wind power cluster.

[0054] Taking synchronous generator G1 and doubly-fed wind turbine WF1 as examples, their equivalent analysis models have the following transfer functions: and The identification error of the equivalent analysis model is within 5%.

[0055] (2) Establish an open-loop decoupling model of active power and frequency based on the center of inertia for the whole system, and derive a simplified relationship between the extreme value of frequency deviation of the center of inertia and the magnitude of active power disturbance and the equivalent inertia level.

[0056] (21) Compared with the improved ASF closed-loop frequency analysis model, the active power-frequency open-loop decoupling analysis model is as follows: Figure 2 As shown. Among them, This refers to the total power deficit suffered by the system. This refers to the transient active power compensation for all grid-connected power generation equipment. For the transient mechanical power response changes of all synchronous power generation equipment; They are 1~ m The conversion factor between the base capacity of an individual synchronous generator cluster and the total base capacity of the system. They are 1~ m Transfer function of the equivalent model of the speed regulation system of a synchronous generator cluster m This refers to the total number of grid-connected synchronous generator sets; for m +1~ m + n The conversion factor between the baseline capacity of the No. 1 wind power cluster and the total baseline capacity of the system. They are respectively m +1~ m + n Transfer function of the frequency regulation equivalent model of the No. 1 wind power cluster n This refers to the total number of grid-connected wind power generation clusters; H sys For the system's equivalent inertia, This is the frequency response curve of the system's center of inertia.

[0057] (22) Consider using a linear function with a constant slope for the time from the moment the system is subjected to an active disturbance to the moment the frequency deviation reaches its extreme value. A simplified modeling analysis is performed on the changes in the total active power compensation for primary frequency regulation of the system (including the total mechanical power compensation of the synchronous generator cluster governor and the total active power frequency regulation compensation of the wind power cluster). Among these, This represents the total active power change due to primary frequency regulation compensation in the system. This refers to the time it takes for the center frequency deviation of the inertia to reach its extreme value after an active disturbance occurs.

[0058] (23) Using step function Approximate fitting of the total electromagnetic power deficit in wind power grid-connected power systems According to the rotor motion equation The expression for the center frequency deviation of inertia can be derived. .in, It is the initial moment when the active disturbance occurs. It is an unknown parameter.

[0059] (24) The approximate quadratic function expression of the inertia center frequency deviation can be obtained by performing a Laplace transform. By feeding it into the equivalent analysis model of the first-order frequency modulation polynomial of the system, the mechanical power output in the frequency domain can be obtained as follows: Wherein, in the formula, G sys ( s The sum of the transfer functions for the primary frequency regulation polynomial equivalent analysis model of all grid-connected power generation equipment is given by: .

[0060] (25) When the inertia center frequency deviation reaches its extreme value, the total transient active power compensation of each grid-connected power generation cluster in the system is equal to the transient active power deficit of the system. This can be achieved by solving... The equation was derived, and the extreme value of the inertia center frequency deviation was obtained. , where Δ f max The extreme value of the frequency deviation of the system's inertia center. k sys,1 and k sys,2 The equivalent model parameters of the primary frequency modulation of the system are respectively.

[0061] Table 1 Comparison of prediction errors for multi-sample frequency response features

[0062] The output and inertia time constant of the grid-connected power generation cluster were adjusted, and 210 steady-state operating conditions were set with wind power penetration rates ranging from 0% to 30% (incrementing by 5%) and system equivalent inertia levels ranging from 3.37s to 30.35s (incrementing by 0.67s). Active power disturbances ranging from 100MW to 1000MW were randomly applied to the stable operating system, increasing by 100MW each time. The extreme values ​​of the inertia center frequency deviation and the maximum rate of change for the sample operating conditions were predicted using both the active power-frequency open-loop model and the improved SFR model, and the prediction errors were recorded. Figure 4Table 1 records the prediction errors of the improved SFR model method and the active-frequency open-loop model method for the extreme values ​​of the inertia center frequency deviation and the maximum frequency change rate under various sample operating conditions.

[0063] As shown in the charts, the prediction accuracy of the active-frequency open-loop model method is higher than that of the improved SFR model method. Specifically, compared with the improved SFR model method, the open-loop model method reduces the median and maximum values ​​of the sample prediction errors for the extreme values ​​of the inertia center frequency deviation by 67.3% and 70.6%, respectively.

[0064] The following three types of indicators are used as error assessment indicators: mean absolute error Average relative error Root mean square error A comprehensive analysis is performed on the prediction results under all sample operating conditions. Where, in the formula, i Indicates the sample number. T ={( x 1, y 1),( x 2, y 2),…( x m , y m )} represents the dataset, x i and y i These represent the sample predicted value (the extreme value predicted by the frequency deviation of the inertia center or the predicted value of the maximum rate of change of frequency) and the actual frequency response characteristic quantity, respectively. M Indicates the number of samples.

[0065] The prediction accuracy of the improved SFR model and the active-frequency open-loop model under all operating conditions of the sample set is as follows: Figure 5 and Figure 6 As shown in the figure, in predicting the extreme values ​​of the inertia center frequency deviation, compared with the improved SFR model method, the active power-frequency open-loop model prediction method reduces the comprehensive prediction error indices MAPE, MAE, and RMSE of the sample operating condition by 66.42%, 68.24%, and 66.92%, respectively; in predicting the maximum rate of change of the inertia center frequency, compared with the improved SFR model method, the active power-frequency open-loop model prediction method reduces the comprehensive prediction error indices MAPE, MAE, and RMSE of the sample operating condition by 3.64%, 3.59%, and 4.34%, respectively. In summary, the proposed active power-frequency open-loop model method meets the requirements for the prediction accuracy of the inertia center frequency response characteristics and is superior to the improved SFR model method.

[0066] (3) Analyze the various active power disturbance faults that may occur in the steady-state operating system in the future and their probability of occurrence. Based on the disturbance probability curve, combined with the simplified relationship of the extreme value of the inertia center frequency deviation and the frequency safety constraint, calculate the multi-level inertia safety boundary based on the frequency deviation extreme value safety constraint.

[0067] (31) By analyzing the operating status of the wind power grid-connected power system, predict the probability of the system under the current stable operating state being subjected to various active power disturbances, and summarize the disturbance probability curve.

[0068] (32) Select the active disturbance magnitude dividing points with occurrence probabilities of 20%, 70%, and 100%, respectively. P 1. P 2. P 3. That is, by analyzing the real-time operating conditions of the current steady-state system, the magnitude of the active power disturbance that the stable operating system will suffer in the future is less than or equal to... P The probability of 1 is 20%, and the magnitude of the active disturbance is less than or equal to P The probability of 2 is 70%, and the magnitude of the active disturbance is less than or equal to P The probability of 3 is 100%.

[0069] (33) Based on P 1. P 2. P 3. Based on the probability boundary points of three active disturbances and the established frequency deviation extreme value safety constraints, the extreme value constraints of the fixed inertia center frequency deviation are calculated using the simplified relationship of the inertia center frequency deviation extreme value derived above. The critical active power disturbance magnitude that the system can withstand H fmax,limit1 , H fmax,limit2 , H fmax,limit3 This serves as a multi-level inertia safety boundary for a stable operating system, based on the extreme value safety constraint of the inertia center frequency deviation.

[0070]

[0071] (4) Using the same sampling method as above, calculate the multi-level inertia safety boundary based on the active disturbance probability curve for the maximum rate of frequency change. H rocof,limit1 , H rocof,limit2 , H rocof,limit3 These indicate that the system ensures the maximum rate of change of the center frequency of inertia meets safety constraints under 20%, 70%, and 100% of the anticipated active power disturbances, respectively. The critical equivalent inertia level.

[0072] (41) According to the rotor motion equation, the initial change rate of the center frequency of the moment of inertia at the instant of the active disturbance is Under the premise that the total active power disturbance of the system changes in an approximate step response form, the initial rate of change of the inertia center frequency is the maximum rate of change of frequency during the process from the moment the disturbance occurs until the frequency reaches its extreme value.

[0073] (42) Calculate the multi-level inertia safety boundary based on the active disturbance probability curve for the maximum rate of change of frequency using the same method described above. H rocof,limit1 , H rocof,limit2 , H rocof,limit3 These indicate the critical equivalent inertia levels at which the maximum rate of change of the center frequency of inertia satisfies safety constraints under 20%, 70%, and 100% of the anticipated active disturbances, respectively.

[0074]

[0075] (5) Based on the piecewise linear idea, considering the safety constraints of the extreme value of frequency deviation and the maximum rate of change of frequency, calculate the frequency stability assessment index based on inertia level (FSAI-I) corresponding to the real-time equivalent inertia level of the system under the multi-level inertia safety boundary.

[0076] (51) First, based on the piecewise linear idea, considering the frequency deviation extreme value safety constraint, calculate the frequency instability risk quantification index corresponding to the real-time equivalent inertia level of the system under the multi-level inertia safety boundary. .

[0077] (52) Then, based on the piecewise linear idea and considering the constraint of the maximum rate of change of frequency, the quantitative assessment index of frequency instability risk corresponding to the real-time equivalent inertia level of the system under the multi-level inertia safety boundary is calculated. .

[0078] (53) Finally, the minimum value of frequency instability risk quantitative assessment index 1 and index 2 is taken as the multi-dimensional fusion frequency instability risk quantitative assessment index. .

[0079] (6) During the online frequency stability assessment, the inertia level of the wind power grid-connected power system is assessed and updated in real time. Finally, the FSAI-I index is calculated online and the quantitative assessment results of the system frequency instability risk under steady state are analyzed.

[0080] The maximum rate of change of the inertia center frequency and the extreme value constraint of the deviation are set to 1 Hz / s and 0.75 Hz, respectively. Using the aforementioned active-frequency open-loop decoupling analysis model, a simplified relationship is established between the characteristic quantity of the inertia center frequency response and the equivalent inertia level and the magnitude of the active disturbance. The aforementioned method is then used to determine the multi-level inertia safety boundary of the current system under the constraint of the maximum rate of change of the inertia center frequency. H rocof,limit1 , H rocof,limit2 , H rocof,limit3 The values ​​are 2.1s, 13.2s, and 56s, respectively. Multi-level inertia safety boundaries under the constraint of extreme values ​​of inertia center frequency deviation. H fmax,limit1 , H fmax,limit2 , H fmax,limit3 The times were 3.5s, 15.2s, and 60.7s, respectively.

[0081] After determining the multi-level safety boundaries of the stable operating system, six different steady-state operating conditions were set by adjusting the inertial time constant and active power output of each generator. The FSAI-I1, FSAI-I2, and FSAI-I indices of the system were calculated under each of the six steady-state operating scenarios to quantitatively evaluate the system's frequency stability margin. Simultaneously, using the PSST toolbox, simulation tests were conducted on the frequency response process after all active power disturbances in the anticipated fault sets under the six steady-state operating scenarios, and the frequency response characteristics of the inertia center after each active power disturbance were recorded. The evaluation results of FSAI-I1, FSAI-I2, and FSAI-I were compared with the actual simulation results and recorded. Figure 7 , Figure 8 middle.

[0082] Depend on Figure 7 , Figure 8It can be seen that as the equivalent inertia level of the system increases and the FSAI-I1 and FSAI-I2 indices increase, the maximum rate of change of the inertia center frequency and the extreme value of the deviation of the stable operating system under the anticipated fault set both decrease overall. The medians of the two types of frequency response characteristics also decrease under all disturbance scenarios, indicating that the frequency stability margin of the stable operating system improves. Therefore, the changing trends of the FSAI-I1 and FSAI-I2 indices are the same as the changing trends of the system's frequency stability margin. The FSAI-I indices for six stable operating conditions are calculated by combining the FSAI-I1 and FSAI-I2 indices. The evaluation results of the FSAI-I indices under the six stable operating conditions are compared and analyzed with the simulation results of the proportion of anticipated fault set cases that meet the frequency safety constraints. Specifically, for stable operating conditions where the proportion of anticipated failure cases meeting frequency safety constraints is less than 20%, the FSAI-I index evaluation result is 0. For stable operating conditions where the proportion of anticipated failure cases meeting frequency safety constraints is greater than 20% but less than 70%, the FSAI-I index evaluation result is between 0 and 100%, and the FSAI-I index evaluation result increases with the increase of the proportion of cases meeting the requirements. For stable operating conditions where the proportion of anticipated failure cases meeting frequency safety constraints is greater than 70% but less than 100%, the FSAI-I index evaluation result is between 100 and 200%, and the FSAI-I index evaluation result increases with the increase of the proportion of cases meeting the requirements. Therefore, from the perspective of the instability ratio of the anticipated failure set, the FSAI-I evaluation index can quantify the frequency stability margin of a stable operating system to a certain extent.

[0083] This invention provides an electronic device, including: a computer-readable storage medium and a processor; The computer-readable storage medium is used to store executable instructions; The processor is configured to read executable instructions stored in the computer-readable storage medium and execute the method as described in any of the above embodiments.

[0084] This invention provides a computer-readable storage medium storing computer instructions that cause a processor to perform the method described in any of the above embodiments.

[0085] This invention provides a computer program product, including a computer program or instructions, which, when executed by a processor, implement the method described in any of the above embodiments.

[0086] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for quantitatively assessing the frequency instability risk of a wind-powered grid-connected power system under steady-state conditions, taking into account the probability of active power disturbances, characterized in that, include: S1, after the disturbance occurs, the steady-state power flow information of each power generation cluster in the system is input into the pre-trained speed regulation system frequency regulation parameter identification model based on XGBoost to obtain the speed regulation system frequency regulation parameters of each power generation cluster, so as to construct the equivalent analysis model transfer function of each power generation cluster. S2, combining the transfer function of the equivalent analysis model of each power generation cluster with the active-frequency open-loop decoupling model based on the inertia center, the relationship between the extreme value of the frequency deviation of the inertia center and the magnitude of the active disturbance and the equivalent inertia level is simplified in order to calculate the extreme value of the frequency deviation of the inertia center. S3, obtain the probability curve of active power disturbance that the system may be subjected to under the current stable operating state, calculate the first multi-level inertia safety boundary in combination with the extreme value safety constraint of the inertia center frequency deviation, and calculate the second multi-level inertia safety boundary in combination with the safety constraint of the maximum rate of change of the inertia center frequency. S4 calculates the frequency instability risk quantification index corresponding to the real-time equivalent inertia level of the system under the first and second multi-level inertia safety boundaries, respectively. S5, take the minimum value of the frequency instability risk quantification assessment index calculated in S4 as the final frequency instability risk quantification assessment result of the system.

2. The method as described in claim 1, characterized in that, The training process of the frequency modulation parameter identification model for the speed control system based on XGBoost is as follows: The steady-state power flow information of each power generation cluster in the system is obtained, and the speed regulation system frequency regulation parameters in the transfer function of the equivalent analysis model of each power generation cluster are obtained by fitting with the least squares method; wherein, the steady-state power flow information includes the output active power PG, generator output reactive power QG, load active power PL, load reactive power QL, voltage amplitude V and voltage phase angle δ before and after the disturbance of the power generation cluster. Using the steady-state power flow information as samples, and the corresponding speed control system frequency regulation parameters as labels, the speed control system frequency regulation parameter identification model based on XGBoost is trained.

3. The method as described in claim 1 or 2, characterized in that, For each synchronous power generation cluster, the equivalent analysis model of its speed regulation system is as follows: ;in, G i,gen (s) is i The transfer function of the equivalent analysis model of the frequency response process of the speed regulation system of the No. 1 synchronous power generation cluster. k i,0 , k i,1 and k i,2 All i Frequency regulation parameters of the speed regulation system of the No. 1 synchronous generator cluster; For each grid-connected wind turbine power generation cluster, the equivalent analysis model of its speed regulation system is as follows: ;in, G i,wind (s) is i The transfer function of the equivalent analysis model of the active frequency support process of the No. 1 wind power cluster. J i,NE , D i,NE and K i,NE They are respectively i Frequency regulation parameters of the speed regulation system of the No. 1 wind power generation cluster.

4. The method as described in claim 3, characterized in that, The simplified formula for calculating the extreme value of the inertia center frequency deviation is: ;in, , , , This refers to the total power deficit suffered by the system. H sys For the system's equivalent inertia, They are 1~ m The conversion factor between the base capacity of an individual synchronous generator cluster and the total base capacity of the system. for m +1~ m + n The conversion factor between the baseline capacity of the No. 1 wind power cluster and the total baseline capacity of the system. m This represents the total number of grid-connected synchronous generator sets. n This refers to the total number of grid-connected wind power generation clusters.

5. The method as described in claim 4, characterized in that, The first multi-level inertia safety boundary is: , The maximum value of the extreme value of the frequency deviation of the center of inertia. , , These are the dividing points where the system will experience active power disturbances with probabilities of 20%, 70%, and 100%, respectively. , , They are respectively with , , The corresponding first-level multi-layer inertia safety boundary; The quantitative assessment index for frequency instability risk corresponding to the real-time equivalent inertia level of the computing system under the first multi-level inertia safety boundary is: The second multi-level inertia safety boundary is: , , , They are respectively with , , The corresponding second-level inertia safety boundary; The quantitative assessment index for frequency instability risk corresponding to the real-time equivalent inertia level of the computing system under the second-level multi-level inertia safety boundary is as follows: 。 6. An electronic device, characterized in that, include: Computer-readable storage media and processors; The computer-readable storage medium is used to store executable instructions; The processor is configured to read executable instructions stored in the computer-readable storage medium and execute the method as described in any one of claims 1-5.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing a processor to perform the method as described in any one of claims 1-5.

8. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by a processor, they implement the method as described in any one of claims 1-5.

Images