A method and system for suppressing synchronous resonance in a wind farm

Abstract

The present application relates to the technical field of synchronous resonance suppression, and discloses a method and a system for suppressing synchronous resonance in a wind farm. The method includes: performing a group oscillation characteristic decomposition process on the system impedance characteristics and key frequency monitoring points to obtain the disturbance sensitivity index of the fan group and the oscillation source positioning result; performing an impedance reconstruction process on the disturbance sensitivity index and the hierarchical early warning threshold parameters to obtain the cooperative damping parameter of the fan group and the power dynamic distribution coefficient; performing a resonance link tracking process on the oscillation source positioning result, the disturbance sensitivity index of the fan group, and the power dynamic distribution coefficient to obtain the key branch impedance compensation value and the reactive power regulation coefficient, and performing a closed-loop correction process to obtain the suppression control instruction and the dynamic correction parameter. The present application realizes the accurate identification and effective suppression of resonance problems within the scope of the wind farm by establishing a complete technical solution for resonance characteristic identification, risk assessment, group oscillation analysis, and cooperative control.

Description

Technical Field

[0001] The present application relates to the field of synchronous resonance suppression, and in particular to a method for suppressing synchronous resonance in a wind farm and a system thereof. Background Art

[0002] Wind farms are an important part of the development of clean energy, but with the continuous expansion of wind farm scale and the continuous increase in grid-connected capacity, the problem of synchronous resonance has become increasingly prominent. Existing wind farm synchronous resonance suppression technologies mainly include passive damping method, active damping method and hybrid damping method. The passive damping method increases system damping by adding resistive elements to the system, and has the characteristics of simple structure and high reliability; the active damping method simulates virtual impedance through control strategies to suppress resonance, and has the advantages of strong flexibility and good adaptability; the hybrid damping method combines the characteristics of passive and active damping, and to a certain extent achieves the improvement of resonance suppression effect.

[0003] However, the existing technology has the following deficiencies: First, traditional resonance suppression methods are often designed for a single wind turbine or a local area, lacking systematic consideration of the group vibration characteristics within the entire wind farm, resulting in unsatisfactory suppression effects under complex working conditions; second, existing suppression strategies mostly use fixed parameter control, which is difficult to adapt to the dynamic changes in wind farm operating conditions, and is prone to problems of over-suppression or under-suppression; third, research on resonance source positioning and propagation path analysis is not in-depth enough, making it difficult to accurately locate and effectively control resonance problems; fourth, there is a lack of in-depth analysis of the mutual influence between wind turbine groups, resulting in unsatisfactory results in coordinated control. Summary of the invention

[0004] The present application provides a method and system for suppressing synchronous resonance of a wind farm, which is used to solve the technical problem that the prior art lacks systematic consideration of the group vibration characteristics of wind farms, resulting in unsatisfactory resonance suppression effects. By establishing a complete technical solution for resonance characteristic identification, risk assessment, group vibration analysis, and coordinated control, accurate identification and effective suppression of resonance problems within the wind farm are achieved.

[0005] In a first aspect, the present application provides a method for suppressing synchronous resonance of a wind farm, the method comprising:

[0006] Perform resonance characteristic identification processing on the electrical quantity data of each node in the wind farm to obtain the resonance frequency interval parameters and system impedance characteristics;

[0007] Performing risk classification processing on the resonance frequency interval parameters to obtain graded warning threshold parameters and key frequency monitoring points;

[0008] Perform a group oscillation characteristic decomposition process on the system impedance characteristic and the key frequency monitoring points to obtain the disturbance sensitivity index of the wind turbine group and the oscillation source positioning result;

[0009] Perform an impedance reconstruction process on the disturbance sensitivity index and the hierarchical early warning threshold parameters to obtain the cooperative damping parameter of the wind turbine group and the power dynamic distribution coefficient;

[0010] Perform a resonant link tracking process on the oscillation source positioning result, the disturbance sensitivity index of the wind turbine group, and the power dynamic distribution coefficient to obtain the key branch impedance compensation value and the reactive power regulation coefficient;

[0011] Perform a closed-loop correction process on the cooperative damping parameter of the wind turbine group, the key branch impedance compensation value, and the reactive power regulation coefficient to obtain the suppression control instruction and the dynamic correction parameter.

[0012] In a second aspect, the present application provides a wind farm synchronous resonance suppression system, and the wind farm synchronous resonance suppression system includes:

[0013] An identification module, configured to perform a resonance characteristic identification process on the electrical quantity data of each node in the wind farm to obtain the resonance frequency interval parameter and the system impedance characteristic;

[0014] A division module, configured to perform a risk level division process on the resonance frequency interval parameter to obtain the hierarchical early warning threshold parameter and the key frequency monitoring point;

[0015] A decomposition module, configured to perform a group oscillation characteristic decomposition process on the system impedance characteristic and the key frequency monitoring point to obtain the disturbance sensitivity index of the wind turbine group and the oscillation source positioning result;

[0016] A reconstruction module, configured to perform an impedance reconstruction process on the disturbance sensitivity index and the hierarchical early warning threshold parameter to obtain the cooperative damping parameter of the wind turbine group and the power dynamic distribution coefficient;

[0017] A tracking module, configured to perform a resonant link tracking process on the oscillation source positioning result, the disturbance sensitivity index of the wind turbine group, and the power dynamic distribution coefficient to obtain the key branch impedance compensation value and the reactive power regulation coefficient;

[0018] A correction module, configured to perform a closed-loop correction process on the cooperative damping parameter of the wind turbine group, the key branch impedance compensation value, and the reactive power regulation coefficient to obtain the suppression control instruction and the dynamic correction parameter.

[0019] In the technical solution provided by this application, by performing resonance characteristic identification processing on the electrical quantity data of each node in the wind farm, the resonance frequency interval parameters and the system impedance characteristics are obtained, realizing a comprehensive and accurate identification of the resonance characteristics of the wind farm, and laying a data foundation for subsequent suppression control; secondly, by performing risk level classification processing on the resonance frequency interval parameters, the hierarchical warning threshold parameters and the key frequency monitoring points are obtained, establishing a hierarchical risk warning mechanism, which can timely detect potential resonance risks; thirdly, by performing group oscillation characteristic decomposition processing on the system impedance characteristics and the key frequency monitoring points, the fan group disturbance sensitivity index and the oscillation source positioning result are obtained, realizing an in-depth analysis of the group oscillation characteristics of the wind farm and the precise positioning of the oscillation source; further, by performing impedance reconstruction processing on the disturbance sensitivity index and the hierarchical warning threshold parameters, the fan group cooperative damping parameters and the power dynamic distribution coefficient are obtained, constructing a cooperative control strategy based on the group characteristics, and improving the integrity of the suppression effect; then, by performing resonance link tracking processing on the oscillation source positioning result, the fan group disturbance sensitivity index and the power dynamic distribution coefficient, the key branch impedance compensation value and the reactive power regulation coefficient are obtained, realizing the precise tracking and directional suppression of the resonance propagation path; finally, by performing closed-loop correction processing on the fan group cooperative damping parameters, the key branch impedance compensation value and the reactive power regulation coefficient, the suppression control command and the dynamic correction parameters are obtained, establishing an adaptive closed-loop control mechanism, ensuring the real-time optimization and dynamic adjustment of the suppression strategy. Generally speaking, the present invention realizes the systematic suppression of the synchronous resonance in the wind farm by constructing a complete technical system of resonance characteristic identification, risk assessment, group oscillation analysis, and cooperative control, improves the accuracy and reliability of the suppression effect, enhances the adaptability and robustness of the control strategy, and provides an effective guarantee for the safe and stable operation of the wind farm. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0021] Figure 1 It is a schematic diagram of an embodiment of the method for suppressing synchronous resonance in a wind farm according to an embodiment of this application;

[0022] Figure 2 It is a schematic diagram of an embodiment of the system for suppressing synchronous resonance in a wind farm according to an embodiment of this application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] The embodiments of the present application provide a method and a system for suppressing synchronous resonance in a wind farm. Terms such as "first", "second", "third", "fourth", etc. (if any) in the specification, claims and the above-mentioned drawings of the present application are used to distinguish similar objects, and do not have to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments described here can be implemented in an order other than that illustrated or described here. In addition, the term "including" or "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0024] For ease of understanding, the specific process of the embodiments of the present application will be described below. Please refer to Figure 1 , an embodiment of the method for suppressing synchronous resonance in a wind farm in the embodiments of the present application includes:

[0025] Step S101: Perform resonance characteristic identification processing on the electrical quantity data of each node in the wind farm to obtain resonance frequency interval parameters and system impedance characteristics;

[0026] Step S102: Perform risk level classification processing on the resonance frequency interval parameters to obtain classification warning threshold parameters and key frequency monitoring points;

[0027] Step S103: Perform group oscillation characteristic decomposition processing on the system impedance characteristics and key frequency monitoring points to obtain the disturbance sensitivity index of the fan group and the oscillation source positioning result;

[0028] Step S104: Perform impedance reconstruction processing on the disturbance sensitivity index and classification warning threshold parameters to obtain the cooperative damping parameter of the fan group and the power dynamic distribution coefficient;

[0029] Step S105: Perform resonance link tracking processing on the oscillation source positioning result, the disturbance sensitivity index of the fan group and the power dynamic distribution coefficient to obtain the key branch impedance compensation value and the reactive power regulation coefficient;

[0030] Step S106: Perform closed-loop correction processing on the cooperative damping parameter of the fan group, the key branch impedance compensation value and the reactive power regulation coefficient to obtain the suppression control instruction and the dynamic correction parameter.

[0031] It can be understood that the execution subject of the present application can be a system for suppressing synchronous resonance in a wind farm, or a terminal or a server. Specifically, it is not limited here. The embodiments of the present application will be described by taking the server as the execution subject as an example.

[0032] Specifically, for the resonance characteristic identification process, by collecting the voltage and current data of each node in the wind farm, synchronously timestamp the sampled data to ensure the chronological correspondence of the data. The voltage and current data here contain amplitude and phase angle information. After performing data standardization processing on the voltage data, the voltage frequency components and amplitude characteristic quantities are obtained through spectrum analysis. For the collected voltage frequency components, correlation analysis is carried out to calculate the coupling degree index between nodes. For example, when the voltage frequency components at the outlet of a certain wind turbine are 49.8 Hz, 50.2 Hz, and 50.5 Hz, the coupling relationship between nodes is determined by calculating the amplitude ratio and phase difference of adjacent frequency components. Similarly, the same processing process is also carried out for the current data, and the branch impedance characteristic quantities are obtained by analyzing the correlation of the current frequency components. Based on the node coupling degree index and the branch impedance characteristic quantities, resonance mode analysis is carried out to identify the candidate resonance frequency interval. By segmentally screening the candidate resonance frequency interval, the target resonance frequency interval is determined. Within the target resonance frequency interval, combining the voltage and current frequency component information, the impedance characteristic curves of each node are calculated, and then the system impedance characteristics and the parameters of the resonance frequency interval are extracted.

[0033] For the obtained resonance frequency interval parameters, risk intensity statistics are carried out to obtain the risk distribution sequence of frequency bands. Through segmented level calculation and processing, the risk level is divided into different levels to obtain the initial value of risk classification. The boundary points of the initial value of risk classification are calibrated to determine the demarcation points of the risk level, and the classification threshold reference group is obtained through thresholding processing. For example, the risk level is divided into three levels: low risk (0 - 30%), medium risk (30% - 70%), and high risk (70% - 100%), and the corresponding threshold reference values are 0.3, 0.7, and 1.0 respectively. After parameter correction and dynamic modification, the classification warning threshold parameters are finally obtained. At the same time, sensitivity analysis is carried out on the resonance frequency interval parameters to obtain the resonance frequency sensitivity matrix, and the monitoring point evaluation index is determined through eigenvalue decomposition, and finally the key frequency monitoring points are selected. After obtaining the system impedance characteristics and the key frequency monitoring points, the group oscillation characteristics are decomposed. First, the system impedance characteristics are grouped, and the wind turbines with similar characteristics are grouped into the same group to calculate the inter-group coupling matrix. Combining the information of the key frequency monitoring points, the group oscillation eigenvectors are extracted, and the group oscillation mode parameters are obtained through modal decomposition. Sensitivity analysis and normalization processing are carried out on the group oscillation mode parameters to obtain the disturbance sensitivity index of the wind turbine group. At the same time, through time delay calculation and energy flow analysis, the oscillation propagation sequence and the oscillation energy distribution matrix are identified, and then the oscillation source localization result is determined.

[0034] Based on the disturbance sensitivity index and the hierarchical early warning threshold parameters, impedance reconstruction processing is carried out. The disturbance sensitivity indexes are sorted to obtain the sensitivity level sequence, which is then matched with the hierarchical early warning threshold parameters to form the sensitivity threshold matrix. By analyzing the impedance characteristics and parameter decomposition, the damping characteristic set is obtained, and then damping distribution and dynamic coordination are carried out to finally obtain the coordinated damping parameters of the wind turbine group. At the same time, the power characteristics are calculated based on the damping reference sequence, and through distributed analysis and dynamic calibration, the power dynamic distribution coefficient is obtained. For the oscillation source location result, the disturbance sensitivity index of the wind turbine group, and the power dynamic distribution coefficient, the resonant link tracking is carried out. The oscillation propagation path is determined through link association processing and hierarchical division is performed to obtain the link hierarchical sequence. Combining with the power dynamic distribution coefficient, feature matching and node identification are carried out to determine the key node set. Branch analysis and reference calculation are performed on the key node set to obtain the branch impedance reference value, and then the compensation amount is calculated and differential correction is carried out to obtain the key branch impedance compensation value. Through reactive power characteristic analysis and coefficient calibration, the reactive power regulation coefficient is finally obtained.

[0035] Finally, closed-loop correction processing is carried out on the coordinated damping parameters of the wind turbine group, the key branch impedance compensation value, and the reactive power regulation coefficient. First, parameter combination and matching are carried out to obtain the parameter combination sequence. Through stability evaluation and interval division, the control reference interval is determined. After parameter calibration and response characteristic analysis, the control reference value and response characteristic parameters are generated. Based on the response characteristic parameters, control instructions are generated and the effectiveness of the control instructions is ensured through closed-loop verification. Finally, through parameter correction and dynamic analysis, the dynamic correction parameters are obtained.

[0036] For example: A wind farm contains 50 wind turbines. The voltage frequency components collected at the outlet of the No. 1 wind turbine are mainly concentrated at 49.8 Hz, 50.2 Hz, and 50.5 Hz, and the voltage amplitudes are 35 kV, 0.8 kV, and 0.5 kV respectively. Through spectrum analysis and correlation calculation, the coupling degree index with adjacent wind turbines is obtained as 0.85. Combining with the analysis of current data, the impedance characteristic quantity of this branch is calculated as 0.4 Ω. After risk level division, the risk level of this frequency interval is medium risk, and the corresponding early warning threshold is 0.7. Through group oscillation characteristic decomposition, the disturbance sensitivity index of this wind turbine is determined as 0.65, belonging to a relatively sensitive wind turbine group. Based on impedance reconstruction, the coordinated damping parameter of this wind turbine is calculated as 0.3, and the power dynamic distribution coefficient is 0.8. In the resonant link tracking, the propagation level between this wind turbine and the oscillation source is identified as 2, the required impedance compensation value is 0.2 Ω, and the reactive power regulation coefficient is 0.4. Finally, through closed-loop correction, the generated suppression control instruction is to reduce the power output by 15%, and the dynamic correction parameter is 0.25.

[0037] In the embodiments of the present application, by performing resonance characteristic identification processing on the electrical quantity data of each node in the wind farm, resonance frequency interval parameters and system impedance characteristics are obtained, realizing a comprehensive and accurate identification of the resonance characteristics of the wind farm and laying a data foundation for subsequent suppression control; secondly, by performing risk level classification processing on the resonance frequency interval parameters, classification warning threshold parameters and key frequency monitoring points are obtained, and a hierarchical risk warning mechanism is established, which can timely detect potential resonance risks; thirdly, by performing group oscillation characteristic decomposition processing on the system impedance characteristics and key frequency monitoring points, the disturbance sensitivity index of the fan group and the oscillation source positioning result are obtained, realizing an in-depth analysis of the group oscillation characteristics of the wind farm and the accurate positioning of the oscillation source; further, by performing impedance reconstruction processing on the disturbance sensitivity index and classification warning threshold parameters, the collaborative damping parameter of the fan group and the power dynamic distribution coefficient are obtained, and a collaborative control strategy based on group characteristics is constructed, improving the integrity of the suppression effect; then, by performing resonance link tracking processing on the oscillation source positioning result, the disturbance sensitivity index of the fan group and the power dynamic distribution coefficient, the impedance compensation value of the key branch and the reactive power regulation coefficient are obtained, realizing the accurate tracking and directional suppression of the resonance propagation path; finally, by performing closed-loop correction processing on the collaborative damping parameter of the fan group, the impedance compensation value of the key branch and the reactive power regulation coefficient, suppression control instructions and dynamic correction parameters are obtained, and an adaptive closed-loop control mechanism is established, ensuring the real-time optimization and dynamic adjustment of the suppression strategy. Generally speaking, the present invention realizes the systematic suppression of synchronous resonance in the wind farm by constructing a complete technical system of resonance characteristic identification, risk assessment, group oscillation analysis, and collaborative control, improves the accuracy and reliability of the suppression effect, enhances the adaptability and robustness of the control strategy, and provides an effective guarantee for the safe and stable operation of the wind farm.

[0038] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0039] (1) Sampling the voltage data of each node in the wind farm to obtain a voltage amplitude sequence and a phase angle sequence, and performing synchronous timestamp marking processing on the voltage amplitude sequence and the phase angle sequence to obtain a standardized voltage data group;

[0040] (2) Performing spectrum analysis processing on the standardized voltage data group to obtain the voltage frequency components and amplitude characteristic quantities of each node, and performing correlation analysis processing on the voltage frequency components to obtain the node coupling degree index;

[0041] (3) Sampling the current data of each node in the wind farm to obtain a current amplitude sequence and a phase angle sequence, and performing synchronous timestamp marking processing on the current amplitude sequence and the phase angle sequence to obtain a standardized current data group;

[0042] (4)Perform spectral analysis on the standardized current data set to obtain the current frequency components and amplitude characteristic quantities of each node, and perform correlation analysis on the current frequency components to obtain the branch impedance characteristic quantities;

[0043] (5)Perform resonance mode analysis on the node coupling degree index and the branch impedance characteristic quantities to obtain the candidate resonance frequency interval, and perform segmented screening on the candidate resonance frequency interval to obtain the target resonance frequency interval;

[0044] (6)Perform impedance calculation on the voltage frequency components and current frequency components within the target resonance frequency interval to obtain the impedance characteristic curves of each node, and perform feature extraction on the impedance characteristic curves of each node to obtain the system impedance characteristics;

[0045] (7)Perform threshold analysis on the system impedance characteristics to obtain the resonance frequency interval parameters.

[0046] Specifically, collect the voltage data of each key node in the wind farm, including positions such as the fan outlet, the booster station bus, and the grid connection point. Through a high-precision synchronous sampling device, the sampling frequency is set to 10 kHz and the sampling duration is 10 seconds to obtain the amplitude sequence and phase angle sequence of the voltage. For the collected original data, perform synchronous timestamp marking, that is, add a time tag accurate to the microsecond level to each sampling point to ensure the timeliness of the data. Then perform normalization processing on the data, convert the voltage amplitude to the range of 0-1 to obtain the standardized voltage data set. Perform fast Fourier transform (FFT) analysis on the standardized voltage data to calculate the frequency spectrum distribution. In the spectral analysis, set the frequency resolution to 0.1 Hz and the analysis range to 0-1000 Hz to obtain the voltage frequency components and the corresponding amplitude characteristic quantities of each node. Perform correlation analysis on the obtained voltage frequency components, and calculate the coupling degree index between nodes by calculating the amplitude ratio and phase difference of the same-frequency components between different nodes. This index reflects the degree of correlation of voltage fluctuations between different nodes. Use the same sampling method to obtain the current data of each node. Similarly, collect 10 seconds of data at a sampling frequency of 10 kHz, and record the amplitude sequence and phase angle sequence of the current. Add synchronous timestamps to the current data and perform normalization processing to obtain the standardized current data set. The normalization processing ensures the consistency of the voltage and current data, facilitating subsequent analysis. Perform FFT analysis on the standardized current data set in the same way to obtain the frequency components and amplitude characteristic quantities of the current. By analyzing the correlation of the current frequency components between different nodes and combining the analysis results of the voltage data, calculate the branch impedance characteristic quantities. The branch impedance characteristic quantities reflect the impedance characteristics of each branch in the wind farm network and provide basic data for subsequent resonance analysis.

[0047] Based on the node coupling degree index and branch impedance characteristic quantities, resonance mode analysis is carried out. First, establish the state - space model of the system, and obtain the natural frequencies and oscillation modes of the system through eigenvalue decomposition. By analyzing the damping characteristics and participation factors of the modes, the candidate resonance frequency intervals are identified. Then, the candidate intervals are segmented and screened, and according to characteristics such as damping ratio and energy distribution, the target resonance frequency intervals with actual resonance risks are screened out. Within the determined target resonance frequency intervals, impedance calculations are performed on the voltage frequency components and current frequency components. By calculating the ratio of voltage to current at each frequency point, the impedance characteristic curves of each node are obtained. Feature extraction is performed on the impedance characteristic curves, including analyzing features such as peak points, valley points, and inflection points of the curves, and considering the changing trends of impedance amplitude and phase angle. Finally, the system impedance characteristics are obtained. Finally, threshold analysis is performed on the system impedance characteristics. By setting thresholds for parameters such as impedance amplitude and phase - angle change rate, the resonance risk levels of different frequency intervals are divided, and the resonance frequency interval parameters are obtained.

[0048] For example, at the outlet of the No. 1 wind turbine in a certain wind farm, after the sampled voltage data is standardized and subjected to FFT analysis, obvious frequency components are found at 49.8 Hz, 50.2 Hz, and 50.5 Hz, with amplitudes of 0.95, 0.03, and 0.02 (normalized values) respectively. Through correlation analysis, the coupling degree index between this node and the adjacent No. 2 wind turbine at 50.2 Hz is calculated to be 0.82, indicating a strong coupling relationship between the two nodes at this frequency point. At the same time, the current data analysis shows that the current component amplitudes at the same frequency points are 0.88, 0.04, and 0.03, and the impedance characteristic quantity of this branch at 50.2 Hz is calculated to be 0.75. Through resonance mode analysis, the 48 - 52 Hz is identified as the candidate resonance frequency interval, and 50.2 Hz ± 0.3 Hz is the target resonance frequency interval. Within this interval, the impedance characteristic curve shows that the impedance amplitude reaches a peak of 3.5 at 50.2 Hz, and the phase angle undergoes a - 180 - degree change. After feature extraction, the system impedance characteristic value of this node is 2.8. Finally, through threshold analysis, the 50.2 Hz ± 0.3 Hz frequency interval is divided into a high - risk resonance interval, and the resonance frequency interval parameters are set as [49.9 Hz, 50.5 Hz].

[0049] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0050] (1) Perform risk - intensity statistical processing on the resonance frequency interval parameters to obtain a frequency - band risk distribution sequence, and perform segmented - level calculation processing on the frequency - band risk distribution sequence to obtain the initial value of risk classification;

[0051] (2)Perform boundary point calibration on the initial risk grading value to obtain the risk level demarcation point, and perform thresholding on the risk level demarcation point to obtain the grading threshold reference group;

[0052] (3)Perform parameter correction on the grading threshold reference group to obtain the early warning threshold reference value, and perform dynamic correction on the early warning threshold reference value to obtain the grading early warning threshold parameter;

[0053] (4)Perform frequency sensitivity analysis on the resonance frequency interval parameters to obtain the resonance frequency sensitivity matrix, and perform eigen-decomposition on the resonance frequency sensitivity matrix to obtain the monitoring point evaluation index;

[0054] (5)Perform priority ranking on the monitoring point evaluation index to obtain the monitoring point candidate sequence, and perform screening on the monitoring point candidate sequence to obtain the key frequency monitoring points.

[0055] Specifically, perform risk intensity statistics on the resonance frequency interval parameters. By statistically analyzing features such as the harmonic amplitude value, duration, and occurrence frequency at each frequency point, calculate the risk intensity value of each frequency point. For continuous frequency intervals, divide them at a frequency interval of 0.1 Hz, and statistically analyze the risk characteristic parameters in each small interval to form a frequency band risk distribution sequence. Perform segmented level calculation on the obtained frequency band risk distribution sequence. Based on the distribution characteristics of the risk intensity value, use the clustering analysis method to initially divide the risk level into three levels: low risk, medium risk, and high risk to obtain the initial risk grading value. For the initial risk grading value, perform boundary point calibration. First, identify the transition region between adjacent risk levels and find the best demarcation point within the transition region. By calculating the gradient change of the risk intensity value, select the point with the largest gradient change as the risk level demarcation point. After determining the demarcation point, perform thresholding on it, map the risk intensity value to the 0-1 interval, and set the grading threshold reference group. Among them, the demarcation threshold from low risk to medium risk is set to 0.3, and the demarcation threshold from medium risk to high risk is set to 0.7.

[0056] Perform parameter calibration processing on the hierarchical threshold reference group. Considering the actual operating status and historical resonance data of the wind farm, adjust the reference threshold. By analyzing the risk level distribution in historical resonance events, calculate the statistical characteristics of each level, including mean, variance, etc., and obtain the reference value of the warning threshold accordingly. At the same time, combine the real-time operation data of the wind farm to dynamically correct the reference value of the warning threshold, and adjust the threshold parameters in real time according to factors such as load level and wind condition changes, and finally obtain the hierarchical warning threshold parameters. When determining the warning threshold, perform frequency sensitivity analysis on the resonance frequency interval parameters. By calculating the influence degree of each frequency point on the system resonance characteristics, establish a frequency sensitivity matrix. The specific analysis methods include: calculating the resonance magnification of each frequency point, the influence degree on the system damping, and the coupling strength with other frequency points. Organize these characteristics into a matrix form, where the matrix elements represent the sensitivity relationship between different frequency points. Perform eigenvalue decomposition on this sensitivity matrix, extract the main eigenvectors, and obtain the monitoring point evaluation index reflecting the importance of each frequency point.

[0057] Based on the monitoring point evaluation index, perform priority sorting processing. First, sort the evaluation indexes according to the numerical size to establish a preliminary candidate sequence of monitoring points. Considering the spatial distribution and coverage of the monitoring points, set the screening rules: the frequency interval between adjacent monitoring points is not less than 0.5 Hz, and the evaluation index value of the monitoring point is not lower than the threshold. Filter the candidate sequence through the screening rules to finally determine the key frequency monitoring points.

[0058] For example: The resonance frequency interval of a wind farm is [49.5 Hz, 50.5 Hz]. Through risk intensity statistics, it is found that the harmonic amplitude value at 50.2 Hz is on average 15% of the rated value, the probability that the duration exceeds 1 minute is 30%, and the occurrence frequency is 2 times per hour. Divide this interval at 0.1 Hz intervals to obtain the frequency band risk distribution sequence. After segmented level calculation, divide the risk intensity value below 0.4 into low risk, 0.4 - 0.7 into medium risk, and above 0.7 into high risk. In the boundary point calibration, by calculating the gradient of the risk intensity value, it is determined that there are obvious gradient changes at the risk intensity values of 0.45 and 0.75, and these two points are set as demarcation points. Considering the historical operation data, adjust the low-risk threshold to 0.35 and the high-risk threshold to 0.8 to form a hierarchical threshold reference group. In the frequency sensitivity analysis, it is calculated that the resonance magnification of the 50.2 Hz frequency point is 2.5 times, the influence degree on the system damping is 0.6, and the coupling strength with adjacent frequency points is 0.4. The comprehensive evaluation index value of this point is 0.85. After sorting the evaluation indexes of all frequency points, select the points with an evaluation index value greater than 0.7 and an interval not less than 0.5 Hz as monitoring points, and finally determine 49.8 Hz, 50.2 Hz, and 50.5 Hz as the key frequency monitoring points.

[0059] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0060] (1) Perform a grouping process on the system impedance characteristics to obtain a set of impedance characteristic groups, and perform a mutual coupling calculation process on the set of impedance characteristic groups to obtain an inter-group coupling matrix;

[0061] (2) Perform a feature extraction process on the inter-group coupling matrix and key frequency monitoring points to obtain a group oscillation eigenvector, and perform a modal decomposition process on the group oscillation eigenvector to obtain group oscillation modal parameters;

[0062] (3) Perform a sensitivity analysis process on the group oscillation modal parameters to obtain a node sensitivity sequence, and perform a normalization process on the node sensitivity sequence to obtain a fan group disturbance sensitivity index;

[0063] (4) Perform a time delay calculation process on the group oscillation modal parameters to obtain an oscillation propagation sequence, and perform an energy flow analysis process on the oscillation propagation sequence to obtain an oscillation energy distribution matrix;

[0064] (5) Perform a source point identification process on the oscillation energy distribution matrix to obtain a set of oscillation source candidates, and perform a localization analysis process on the set of oscillation source candidates to obtain an oscillation source localization result.

[0065] Specifically, group the system impedance characteristics, and perform clustering analysis based on the similarity of impedance characteristic curves. By calculating the Euclidean distance of impedance amplitude and phase angle, fans with similar characteristics are grouped into the same group to form a set of impedance characteristic groups. Perform a mutual coupling calculation on this grouped set, analyze the coupling strength between different groups, calculate the impedance mutual inductance coefficient between groups, and construct an inter-group coupling matrix. Each element of this matrix represents the coupling relationship strength between different groups. For the inter-group coupling matrix and key frequency monitoring point data, perform a feature extraction process. Calculate the eigenvalues and eigenvectors of the inter-group coupling matrix at each key frequency point to obtain a group oscillation eigenvector reflecting the characteristics of group oscillation. Perform a modal decomposition on the group oscillation eigenvector, use the singular value decomposition method, extract the main oscillation modes, and calculate the frequency, damping ratio, and amplitude of each mode to form group oscillation modal parameters.

[0066] Perform a sensitivity analysis on the group oscillation modal parameters. By calculating the participation factor of each node in the main oscillation mode, obtain a node sensitivity sequence. This sequence reflects the influence degree of each node in the group oscillation process. Perform a normalization process on the node sensitivity sequence, map the sensitivity value to the 0-1 interval, and obtain a standardized fan group disturbance sensitivity index.

[0067] The time delay calculation process and feature correlation process are key steps in this solution, and their mathematical descriptions are as follows:

[0068]

[0069] wherein, represents the oscillation propagation time delay between nodes i and j; is the angular frequency of the k-th oscillation mode; and are the phase angles of nodes i and j in the k-th mode respectively; is the coherence coefficient of nodes i and j in the k-th mode; is the weight coefficient of the k-th mode; is the attenuation coefficient of the k-th mode; t is the observation time; N is the total number of modes considered.

[0070] An oscillation propagation sequence is obtained through time delay calculation, and then energy flow direction analysis is performed on this sequence to calculate the energy transmission direction and intensity between each node, forming an oscillation energy distribution matrix. This matrix describes the propagation law of oscillation energy in the wind farm. Source point identification processing is performed on the oscillation energy distribution matrix, nodes with positive net energy output exceeding the threshold are screened out and included in the oscillation source candidate set. By calculating the energy contribution rate and influence range of each candidate point, location analysis is carried out, and finally the oscillation source location result is determined.

[0071] For example: A wind farm contains 30 wind turbines, which are divided into 3 groups based on impedance characteristic clustering analysis, with 10 wind turbines in each group. By calculating the mutual inductance coefficient of impedance between groups, a 3×3 inter-group coupling matrix is obtained, and the matrix element values range from 0.2 to 0.8. Feature extraction is performed at the key frequency point of 50.2 Hz, and a group oscillation eigenvector with eigenvalues of [2.5, 1.8, 0.7] is obtained. Through modal decomposition, 3 main oscillation modes are obtained, with frequencies of 50.2 Hz, 50.4 Hz, and 50.1 Hz respectively, and the corresponding damping ratios are 0.03, 0.05, and 0.08. Through sensitivity analysis, the participation factor of the No. 1 wind turbine in the dominant mode is calculated to be 0.85, and its perturbation sensitivity index is 0.92 after normalization, indicating that this wind turbine plays an important role in the group oscillation process. Time delay calculation shows that the oscillation propagation time delay between the No. 1 wind turbine and the No. 2 wind turbine is 15 ms, and energy flow direction analysis shows that the energy output ratio of the No. 1 wind turbine to the surrounding wind turbines is 1.5, exceeding the threshold of 1.2, so it is included in the oscillation source candidate set. Through location analysis, it is confirmed that the No. 1 wind turbine is the oscillation source, and its influence range covers 5 adjacent wind turbines.

[0072] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0073] (1) Perform sensitivity ranking processing on the perturbation sensitivity index to obtain a sensitivity level sequence, and perform matching processing on the sensitivity level sequence and the classification early warning threshold parameter to obtain a sensitivity threshold matrix;

[0074] (2) Perform impedance characteristic analysis and processing on the sensitivity threshold matrix to obtain impedance characteristic parameters, and perform parameter decomposition processing on the impedance characteristic parameters to obtain a damping characteristic set;

[0075] (3) Perform damping distribution processing on the damping characteristic set to obtain a damping reference sequence, and perform dynamic coordination processing on the damping reference sequence to obtain the coordinated damping parameters of the wind turbine group;

[0076] (4) Perform power characteristic calculation processing on the damping reference sequence to obtain a power regulation reference value, and perform distributed analysis processing on the power regulation reference value to obtain a power distribution vector;

[0077] (5) Perform coefficient calibration processing on the power distribution vector to obtain a distribution reference coefficient, and perform dynamic calibration processing on the distribution reference coefficient to obtain a power dynamic distribution coefficient.

[0078] Specifically, perform sensitivity ranking processing on the disturbance sensitivity index, perform descending order arrangement based on the magnitude of the disturbance sensitivity value to form a sensitivity level sequence. Match this sequence with the hierarchical warning threshold parameters to establish a two-dimensional sensitivity threshold matrix. The rows of the matrix represent different sensitivity levels, and the columns represent different warning levels. The matrix element values reflect the control reference values under the corresponding conditions. For the sensitivity threshold matrix, perform impedance characteristic analysis, extract the impedance amplitude and phase angle characteristics of each wind turbine at different frequency points to obtain impedance characteristic parameters. Perform decomposition processing on these parameters, decompose the impedance characteristics into resistive components and inductive components, comprehensively evaluate the damping contribution ability of each wind turbine, and form a damping characteristic set.

[0079] The damping distribution process is the core step of this solution, and its mathematical description is as follows:

[0080]

[0081] Among them, is the damping distribution coefficient of the i-th wind turbine; is the reference damping coefficient of the wind turbine; is the weight factor of the j-th resonance mode; is the participation degree of wind turbine i in mode j; is the attenuation coefficient of mode j; is the electrical distance from wind turbine i to the resonance center j; is the rated capacity of wind turbine i; is the energy density of mode j; M is the number of modes considered.

[0082] Perform damping allocation processing on the damping characteristic set, calculate the damping reference sequence according to the above formula, and this sequence reflects the damping responsibilities that each wind turbine should bear. Through dynamic coordination processing, considering the mutual influence among wind turbines, adjust the damping parameter configuration of each wind turbine, and finally obtain the collaborative damping parameters of the wind turbine group. Based on the damping reference sequence, perform power characteristic calculation, analyze the active power regulation ability and reactive power support ability of each wind turbine, and obtain the power regulation reference value. Conduct distributed analysis on the power regulation reference value, consider the spatial distribution and electrical distance of the wind turbines, calculate the power sharing ratio of each wind turbine, and form a power distribution vector.

[0083] Perform coefficient calibration on the power distribution vector, and determine the reference coefficient for power distribution according to factors such as the operating status and power generation efficiency of the wind turbines. Conduct dynamic calibration on the reference coefficient, and adjust the power distribution ratio of each wind turbine in real time to obtain the dynamic power distribution coefficient.

[0084] For example: A wind farm contains 20 wind turbines. Through sensitivity sorting, sort the disturbance sensitivity indicators from large to small. The sensitivity of the No. 1 wind turbine is 0.95, ranking first. Combine the early warning threshold parameters (low risk 0.3, medium risk 0.7) to construct a 4×3 sensitivity threshold matrix. Through impedance characteristic analysis, the impedance amplitude of the No. 1 wind turbine at 50.2 Hz is 2.5 Ω, and the phase angle is -45 degrees. The equivalent resistive component is decomposed to be 1.77 Ω. Through damping allocation calculation, considering that the rated capacity of the No. 1 wind turbine is 2 MW, the electrical distance to the resonance center is 2 bus segments, the weight factor of the dominant mode is 0.8, and the participation degree is 0.9, calculate its damping allocation coefficient to be 0.85. After dynamic coordination, determine the collaborative damping parameter of this wind turbine to be 0.75. The power characteristic calculation shows that the active power regulation ability of this wind turbine is 15% of the rated power, and the reactive power support ability is 30% of the rated capacity.

[0085] Through distributed analysis, determine the power sharing ratio of the No. 1 wind turbine to be 0.12. Considering its 90% power generation efficiency, calibrate and obtain the power distribution reference coefficient to be 0.108. Finally, through dynamic calibration, according to the real-time wind conditions, adjust its dynamic power distribution coefficient to 0.115 for guiding the real-time power regulation of the wind turbine.

[0086] In a specific embodiment, the process of executing step S105 may specifically include the following steps:

[0087] (1) Perform link association processing on the oscillation source positioning result and the wind turbine group disturbance sensitivity indicators to obtain the oscillation propagation path, and perform hierarchical division processing on the oscillation propagation path to obtain the link level sequence;

[0088] (2) Perform feature matching processing on the link-level sequence and the power dynamic allocation coefficient to obtain the link impedance feature quantity, and perform node identification processing on the link impedance feature quantity to obtain the set of key nodes;

[0089] (3) Perform branch analysis processing on the set of key nodes to obtain the branch impedance sequence, and perform reference calculation processing on the branch impedance sequence to obtain the branch impedance reference value;

[0090] (4) Perform compensation amount calculation processing on the branch impedance reference value to obtain the impedance compensation reference amount, and perform difference correction processing on the impedance compensation reference amount to obtain the key branch impedance compensation value;

[0091] (5) Perform reactive power characteristic analysis processing on the key branch impedance compensation value to obtain the reactive power compensation demand, and perform coefficient calibration processing on the reactive power compensation demand to obtain the reactive power regulation coefficient.

[0092] Specifically, perform link association processing on the oscillation source localization result and the fan group disturbance sensitivity index. Starting from the identified oscillation source node, based on the electrical connection relationship and disturbance propagation characteristics between nodes, trace the propagation path of the oscillation energy. According to the electrical distance and impedance characteristics between nodes, determine the priority path of oscillation propagation to form a complete oscillation propagation path diagram. Perform hierarchical division on the oscillation propagation path, with the oscillation source as the first layer, and divide the nodes along the way into different layers according to the propagation distance and attenuation characteristics to obtain the link-level sequence. For the link-level sequence and the power dynamic allocation coefficient, perform feature matching processing. Analyze the power distribution characteristics of each layer of nodes, calculate the impedance relationship between nodes, and obtain the link impedance feature quantity. This feature quantity reflects the impedance coupling relationship between nodes on the oscillation propagation path. Perform node identification on the link impedance feature quantity, and select the key nodes that have a greater impact on oscillation propagation according to the significance of the impedance feature and the importance of the node position to form the set of key nodes.

[0093] Perform branch analysis on the set of key nodes to calculate the branch impedance values between nodes. Considering the electrical characteristics of the branches, including resistance, inductance, and capacitance parameters, obtain the complete branch impedance sequence. Perform reference calculation on the branch impedance sequence, and determine the reference impedance value of each branch according to the impedance characteristics at the resonant frequency point. During the reference calculation process, consider the transmission capacity and stability margin requirements of the branches to obtain the branch impedance reference value. Perform compensation amount calculation on the branch impedance reference value, analyze the impedance deviation of each branch at the resonant frequency point, and calculate the required compensation amount. Normalize the compensation amount to obtain the impedance compensation reference amount. Perform difference correction on the impedance compensation reference amount, and adjust the distribution ratio of the compensation amount according to the actual operating state and load-bearing capacity of the branches to finally obtain the key branch impedance compensation value.

[0094] Perform reactive power characteristic analysis for the impedance compensation value of the key branch. Calculate the reactive power injection amount required to achieve impedance compensation, consider the reactive power regulation ability of the wind turbines and the grid operation constraints, and obtain the reactive power compensation demand. Calibrate the coefficient of the reactive power compensation demand, and determine the reactive power regulation coefficient according to the compensation priority of each branch and the principle of nearby compensation.

[0095] In a specific embodiment, the process of executing step S106 may specifically include the following steps:

[0096] (1) Perform parameter combination processing on the coordinated damping parameters of the wind turbine group and the impedance compensation value of the key branch to obtain an initial set of control parameters, and perform matching processing on the initial set of control parameters and the reactive power regulation coefficient to obtain a parameter combination sequence;

[0097] (2) Perform stability evaluation processing on the parameter combination sequence to obtain a stability margin index, and perform interval division processing on the stability margin index to obtain a control reference interval;

[0098] (3) Perform parameter calibration processing on the control reference interval to obtain a control reference value, and perform response characteristic analysis processing on the control reference value to obtain response characteristic parameters;

[0099] (4) Perform command generation processing on the response characteristic parameters to obtain a control command reference value, and perform closed-loop verification processing on the control command reference value to obtain a suppression control command;

[0100] (5) Perform parameter correction processing on the suppression control command to obtain a correction reference value, and perform dynamic analysis processing on the correction reference value to obtain dynamic correction parameters.

[0101] Specifically, perform parameter combination processing on the coordinated damping parameters of the wind turbine group and the impedance compensation value of the key branch, and integrate multiple parameters of the same control object together. For each wind turbine, weight and combine its coordinated damping parameters and the impedance compensation value of the connected branch according to the parameter weights to form an initial set of control parameters. Perform matching processing on this initial set and the reactive power regulation coefficient, and correct the control parameters according to the reactive power regulation ability to obtain a parameter combination sequence. Perform stability evaluation on the parameter combination sequence to analyze the influence of each parameter on the system stability. By calculating the small-signal stability margin, evaluate the influence of the change of control parameters on the system eigenvalues to obtain a stability margin index. This index includes characteristic quantities such as damping ratio and oscillation frequency deviation. Perform interval division on the stability margin index, and divide the control interval into a safe interval, a warning interval, and a dangerous interval according to the stability requirements of the system to obtain a control reference interval.

[0102] Parameter calibration is performed for the control reference interval, and the control parameters of each interval are corrected by combining the real-time operation data of the wind farm. By calculating the damping characteristics and impedance characteristics under the actual operating conditions, the control reference value is obtained. Response characteristic analysis is carried out on the control reference value, including calculating indicators such as response time, overshoot, and steady-state error, to form response characteristic parameters. Instruction generation processing is performed based on the response characteristic parameters, and a step-by-step control strategy is generated according to the dynamic response characteristics of the system. The control strategy is quantified into specific numerical instructions to obtain the control instruction reference value. Closed-loop verification is carried out on the control instruction reference value, and the control effect is verified through simulation analysis to ensure the effectiveness of the control instruction, and finally the suppression control instruction is obtained.

[0103] Parameter correction is performed on the suppression control instruction, and the control parameters are fine-tuned according to the real-time operation data. By calculating the control deviation and correction coefficient, the correction reference value is obtained. Dynamic analysis is carried out on the correction reference value, considering the time-varying characteristics of the control parameters, to determine the dynamic adjustment strategy of the parameters, and finally the dynamic correction parameters are obtained.

[0104] For example: For the No. 1 wind turbine in a certain wind farm, its cooperative damping parameter is 0.75, the impedance compensation value of the connecting branch is 0.25 Ω, and the reactive power regulation coefficient is 0.4. These parameters are combined according to the weight of 0.5:0.3:0.2 to obtain the initial control parameter of 0.53. After stability evaluation, it is calculated that the damping ratio of the system under this parameter is 0.05, the oscillation frequency deviation is 0.02, and the comprehensive stability margin index is 0.85. According to the stability margin index, the control interval is divided into three segments: the safe interval (0.8 - 1.0), the warning interval (0.6 - 0.8), and the dangerous interval (0 - 0.6). Since the current stability margin of 0.85 is in the safe interval, the control reference value is initially determined to be 0.50. Through response characteristic analysis, it is calculated that under this control value, the system response time is 100 ms, the overshoot is 5%, and the steady-state error is 2%. Based on the response characteristics, the generated control instructions include: adjusting the damping parameter to 0.70, the impedance compensation value to 0.22 Ω, and the reactive power regulation coefficient to 0.35. After closed-loop verification, this set of control instructions can reduce the harmonic amplitude value of the system by 80%. During the actual execution process, it is found that the system response is slightly lagging, and the calculated correction coefficient is 1.1. The control parameters are adjusted to: damping parameter 0.77, impedance compensation value 0.24 Ω, and reactive power regulation coefficient 0.38. The final dynamic correction parameters are set as: when the voltage fluctuation exceeds 5%, the damping parameter increases by 0.05, the impedance compensation value increases by 0.02 Ω, and the reactive power regulation coefficient increases by 0.03.

[0105] The above describes the synchronous resonance suppression method for the wind farm in the embodiment of the present application. Next, the synchronous resonance suppression system for the wind farm in the embodiment of the present application will be described. Please refer to Figure 2, an embodiment of the synchronous resonance suppression system for a wind farm in an embodiment of the present application includes:

[0106] An identification module, configured to perform resonance characteristic identification processing on the electrical quantity data of each node in the wind farm to obtain resonance frequency interval parameters and system impedance characteristics;

[0107] A division module, configured to perform risk level division processing on the resonance frequency interval parameters to obtain classification warning threshold parameters and key frequency monitoring points;

[0108] A decomposition module, configured to perform group oscillation characteristic decomposition processing on the system impedance characteristics and the key frequency monitoring points to obtain the disturbance sensitivity index of the fan group and the oscillation source positioning result;

[0109] A reconstruction module, configured to perform impedance reconstruction processing on the disturbance sensitivity index and the classification warning threshold parameters to obtain the cooperative damping parameter of the fan group and the power dynamic distribution coefficient;

[0110] A tracking module, configured to perform resonance link tracking processing on the oscillation source positioning result, the disturbance sensitivity index of the fan group, and the power dynamic distribution coefficient to obtain the key branch impedance compensation value and the reactive power regulation coefficient;

[0111] A correction module, configured to perform closed-loop correction processing on the cooperative damping parameter of the fan group, the key branch impedance compensation value, and the reactive power regulation coefficient to obtain a suppression control instruction and dynamic correction parameters.

[0112] Through the collaborative cooperation of the above-mentioned various components, by identifying and processing the harmonic resonance characteristics of the electrical quantity data at each node of the wind farm, the harmonic resonance frequency interval parameters and the system impedance characteristics are obtained, realizing the comprehensive and accurate identification of the harmonic resonance characteristics of the wind farm, and laying a data foundation for subsequent suppression control; secondly, by classifying the risk levels of the harmonic resonance frequency interval parameters, the hierarchical early warning threshold parameters and the key frequency monitoring points are obtained, and a hierarchical risk early warning mechanism is established, which can timely detect potential harmonic resonance risks; thirdly, by decomposing the group oscillation characteristics of the system impedance characteristics and the key frequency monitoring points, the disturbance sensitivity index of the wind turbine group and the oscillation source positioning result are obtained, realizing the in-depth analysis of the group oscillation characteristics of the wind farm and the accurate positioning of the oscillation source; further, by performing impedance reconstruction on the disturbance sensitivity index and the hierarchical early warning threshold parameters, the collaborative damping parameters of the wind turbine group and the power dynamic distribution coefficient are obtained, and a collaborative control strategy based on the group characteristics is constructed, improving the integrity of the suppression effect; then, by performing harmonic resonance link tracking on the oscillation source positioning result, the disturbance sensitivity index of the wind turbine group and the power dynamic distribution coefficient, the impedance compensation value of the key branch and the reactive power regulation coefficient are obtained, realizing the accurate tracking and directional suppression of the harmonic resonance propagation path; finally, by performing closed-loop correction on the collaborative damping parameters of the wind turbine group, the impedance compensation value of the key branch and the reactive power regulation coefficient, the suppression control command and the dynamic correction parameters are obtained, and an adaptive closed-loop control mechanism is established, ensuring the real-time optimization and dynamic adjustment of the suppression strategy. Generally speaking, the present invention realizes the systematic suppression of the synchronous harmonic resonance in the wind farm by constructing a complete technical system of harmonic resonance identification, risk assessment, group oscillation analysis and collaborative control, improves the accuracy and reliability of the suppression effect, enhances the adaptability and robustness of the control strategy, and provides an effective guarantee for the safe and stable operation of the wind farm.

[0113] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for suppressing synchronous resonance of a wind farm, characterized in that: The method for suppressing synchronous resonance of a wind farm comprises: Perform resonance characteristic identification processing on the electrical quantity data of each node in the wind farm to obtain the resonance frequency interval parameters and system impedance characteristics; Performing risk classification processing on the resonance frequency interval parameters to obtain graded warning threshold parameters and key frequency monitoring points; Performing group vibration characteristic decomposition processing on the system impedance characteristics and the key frequency monitoring points to obtain a wind turbine group disturbance sensitivity index and an oscillation source positioning result; Performing impedance reconstruction processing on the disturbance sensitivity index and the graded warning threshold parameter to obtain a wind turbine group coordinated damping parameter and a power dynamic allocation coefficient; Performing resonant link tracking processing on the oscillation source positioning result, the wind turbine group disturbance sensitivity index and the power dynamic allocation coefficient to obtain a key branch impedance compensation value and a reactive power adjustment coefficient; The wind turbine group coordinated damping parameters, the key branch impedance compensation value and the reactive power regulation coefficient are subjected to closed-loop correction processing to obtain suppression control instructions and dynamic correction parameters.

2. The method for suppressing synchronous resonance of a wind farm according to claim 1, characterized in that: The resonant characteristic identification processing is performed on the electrical quantity data of each node of the wind farm to obtain the resonant frequency interval parameters and the system impedance characteristics, including: Sampling the voltage data of each node of the wind farm to obtain a voltage amplitude sequence and a phase angle sequence, and performing synchronous time stamp marking on the voltage amplitude sequence and the phase angle sequence to obtain a standardized voltage data group; Performing spectrum analysis on the standardized voltage data group to obtain voltage frequency components and amplitude characteristic quantities of each node, and performing correlation analysis on the voltage frequency components to obtain an index of coupling degree between nodes; Sampling the current data of each node of the wind farm to obtain a current amplitude sequence and a phase angle sequence, and performing synchronous time stamp marking on the current amplitude sequence and the phase angle sequence to obtain a standardized current data group; Performing spectrum analysis on the standardized current data group to obtain the current frequency components and amplitude characteristic quantities of each node, and performing correlation analysis on the current frequency components to obtain the branch impedance characteristic quantities; Performing resonance modal analysis on the inter-node coupling index and the branch impedance characteristic quantity to obtain a candidate resonance frequency interval, and performing segmented screening on the candidate resonance frequency interval to obtain a target resonance frequency interval; Performing impedance calculation processing on the voltage frequency component and the current frequency component within the target resonant frequency interval to obtain an impedance characteristic curve of each node, and performing feature extraction processing on the impedance characteristic curve of each node to obtain a system impedance characteristic; Threshold analysis is performed on the impedance characteristics of the system to obtain resonance frequency interval parameters.

3. The method for suppressing synchronous resonance of a wind farm according to claim 1, characterized in that: The risk level classification process of the resonant frequency interval parameter to obtain the graded warning threshold parameter and the key frequency monitoring point includes: Performing risk intensity statistical processing on the resonance frequency interval parameter to obtain a frequency band risk distribution sequence, and performing segmented grade calculation processing on the frequency band risk distribution sequence to obtain a risk grading initial value; Performing boundary point calibration processing on the risk grading initial value to obtain risk grade demarcation points, and performing threshold processing on the risk grade demarcation points to obtain a grading threshold reference group; Performing parameter correction processing on the graded threshold reference group to obtain a warning threshold reference value, and dynamically correcting the warning threshold reference value to obtain a graded warning threshold parameter; Performing frequency sensitivity analysis on the resonance frequency interval parameters to obtain a resonance frequency sensitivity matrix, and performing characteristic decomposition on the resonance frequency sensitivity matrix to obtain a monitoring point evaluation index; Prioritizing the monitoring point evaluation indicators to obtain a monitoring point candidate sequence, and screening the monitoring point candidate sequence to obtain key frequency monitoring points.

4. The method for suppressing synchronous resonance of a wind farm according to claim 1, characterized in that: The group vibration characteristic decomposition processing is performed on the system impedance characteristic and the key frequency monitoring point to obtain the wind turbine group disturbance sensitivity index and the oscillation source positioning result, including: Performing group division processing on the impedance characteristics of the system to obtain an impedance characteristic grouping set, and performing mutual coupling calculation processing on the impedance characteristic grouping set to obtain an inter-group coupling matrix; Performing feature extraction processing on the inter-group coupling matrix and the key frequency monitoring point to obtain a group vibration feature vector, and performing modal decomposition processing on the group vibration feature vector to obtain a group vibration modal parameter; Performing sensitivity analysis on the group vibration modal parameters to obtain a node sensitivity sequence, and normalizing the node sensitivity sequence to obtain a wind turbine group disturbance sensitivity index; Performing time delay calculation processing on the group vibration modal parameters to obtain an oscillation propagation sequence, and performing energy flow analysis processing on the oscillation propagation sequence to obtain an oscillation energy distribution matrix; The oscillation energy distribution matrix is ​​subjected to source point identification processing to obtain an oscillation source candidate set, and the oscillation source candidate set is subjected to positioning analysis processing to obtain an oscillation source positioning result.

5. The method for suppressing synchronous resonance of a wind farm according to claim 1, characterized in that: The impedance reconstruction processing is performed on the disturbance sensitivity index and the graded warning threshold parameter to obtain the wind turbine group coordinated damping parameter and the power dynamic allocation coefficient, including: Performing sensitivity sorting processing on the disturbance sensitivity index to obtain a sensitivity level sequence, and performing matching processing on the sensitivity level sequence and the graded warning threshold parameter to obtain a sensitivity threshold matrix; Performing impedance characteristic analysis processing on the sensitivity threshold matrix to obtain impedance characteristic parameters, and performing parameter decomposition processing on the impedance characteristic parameters to obtain a damping characteristic set; Performing damping distribution processing on the damping characteristic set to obtain a damping reference sequence, and performing dynamic coordination processing on the damping reference sequence to obtain wind turbine group coordinated damping parameters; Performing power characteristic calculation processing on the damping reference sequence to obtain a power adjustment reference value, and performing distributed analysis processing on the power adjustment reference value to obtain a power allocation vector; The power allocation vector is subjected to coefficient calibration processing to obtain allocation reference coefficients, and the allocation reference coefficients are subjected to dynamic calibration processing to obtain power dynamic allocation coefficients.

6. The method for suppressing synchronous resonance of a wind farm according to claim 1, characterized in that: The resonant link tracking process is performed on the oscillation source positioning result, the wind turbine group disturbance sensitivity index and the power dynamic allocation coefficient to obtain the key branch impedance compensation value and the reactive power adjustment coefficient, including: Performing link association processing on the oscillation source positioning result and the disturbance sensitivity index of the wind turbine group to obtain an oscillation propagation path, and performing hierarchical division processing on the oscillation propagation path to obtain a link hierarchical sequence; Performing feature matching processing on the link level sequence and the power dynamic allocation coefficient to obtain a link impedance feature quantity, and performing node identification processing on the link impedance feature quantity to obtain a key node set; Performing branch analysis processing on the key node set to obtain a branch impedance sequence, and performing benchmark calculation processing on the branch impedance sequence to obtain a branch impedance benchmark value; Performing compensation amount calculation processing on the branch impedance reference value to obtain an impedance compensation reference amount, and performing difference correction processing on the impedance compensation reference amount to obtain a key branch impedance compensation value; The reactive characteristic analysis process is performed on the critical branch impedance compensation value to obtain the reactive compensation demand, and the reactive compensation demand is subjected to coefficient calibration process to obtain the reactive adjustment coefficient.

7. The method for suppressing synchronous resonance of a wind farm according to claim 1, characterized in that: The closed-loop correction process is performed on the wind turbine group coordinated damping parameter, the key branch impedance compensation value and the reactive power adjustment coefficient to obtain the suppression control instruction and the dynamic correction parameter, including: Performing parameter combination processing on the wind turbine group coordinated damping parameter and the key branch impedance compensation value to obtain an initial set of control parameters, and performing matching processing on the initial set of control parameters and the reactive power adjustment coefficient to obtain a parameter combination sequence; Performing stability evaluation processing on the parameter combination sequence to obtain a stability margin index, and performing interval division processing on the stability margin index to obtain a control reference interval; Performing parameter calibration processing on the control reference interval to obtain a control reference value, and performing response characteristic analysis processing on the control reference value to obtain a response characteristic parameter; Performing instruction generation processing on the response characteristic parameter to obtain a control instruction reference value, and performing closed-loop verification processing on the control instruction reference value to obtain a suppression control instruction; The suppression control instruction is subjected to parameter correction processing to obtain a correction reference value, and the correction reference value is subjected to dynamic analysis processing to obtain a dynamic correction parameter.

8. A wind farm synchronous resonance suppression system, used to implement the wind farm synchronous resonance suppression method according to any one of claims 1 to 7, characterized in that: The wind farm synchronous resonance suppression system comprises: An identification module is used to perform resonance characteristic identification processing on the electrical quantity data of each node of the wind farm to obtain the resonance frequency interval parameters and system impedance characteristics; A classification module is used to classify the resonant frequency interval parameters into risk levels to obtain graded warning threshold parameters and key frequency monitoring points; A decomposition module, used to perform group vibration characteristic decomposition processing on the system impedance characteristics and the key frequency monitoring points to obtain a wind turbine group disturbance sensitivity index and an oscillation source positioning result; A reconstruction module, used to perform impedance reconstruction processing on the disturbance sensitivity index and the graded warning threshold parameter to obtain a wind turbine group coordinated damping parameter and a power dynamic allocation coefficient; A tracking module, used to perform resonant link tracking processing on the oscillation source positioning result, the disturbance sensitivity index of the wind turbine group and the power dynamic allocation coefficient, so as to obtain a key branch impedance compensation value and a reactive power adjustment coefficient; The correction module is used to perform closed-loop correction processing on the wind turbine group coordinated damping parameters, the key branch impedance compensation value and the reactive power adjustment coefficient to obtain suppression control instructions and dynamic correction parameters.

Images