Wind power plant subsynchronous oscillation suppression method and device based on energy flow and medium

Abstract

The invention discloses a wind power plant subsynchronous oscillation suppression method and device based on energy flow and a medium, and relates to the technical field of wind power integration stability control, and the method comprises the steps: drawing an oscillation energy vector graph on a wind power plant electrical wiring diagram based on an oscillation energy vector, and recognizing the direction, convergent point and divergent point of the oscillation energy vector; generating excitation point coordinates of subsynchronous oscillation; according to the excitation point coordinates of the subsynchronous oscillation and the electrical distance of each fan in the wind power plant, a targeted suppression strategy is dynamically formulated; and executing a targeted suppression strategy, integrating the electrical quantity change at the coordinates of the excitation point of the subsynchronous oscillation with wind power plant response data acquired in real time, and generating a subsynchronous oscillation suppression effect report. According to the invention, accurate mapping of the excitation point from the electrical node to the geographic coordinate is realized, a targeting target is provided for inhibition, and the beneficial effect of realizing efficient oscillation inhibition with the minimum power generation loss is achieved.

Description

Technical Field

[0001] This invention relates to the field of wind power grid connection stability control technology, and in particular to a method, equipment and medium for suppressing subsynchronous oscillations in wind farms based on energy flow. Background Technology

[0002] Early research focused on installing subsynchronous oscillation protection devices on thermal power units or using flexible AC transmission equipment to provide damping. With the increasing penetration of wind power, the technological focus has shifted to the wind farm side. The application of broadband measurement methods has made online monitoring of subsynchronous oscillations based on wide-area measurements possible. Existing technical solutions mostly focus on identifying oscillation modes based on acquired voltage and current signals, using signal processing methods such as the Prony algorithm and discrete Fourier transform, and designing additional damping controllers or formulating suppression strategies such as turbine tripping and adjusting reactive power reserves. In recent years, with the large-scale deployment of synchronous phasor measurement devices in wind farms, more accurate real-time capture of subsynchronous oscillation phenomena has been achieved, providing a data foundation for analyzing oscillation characteristics and formulating control strategies.

[0003] Existing technologies still have significant limitations in the precise location of oscillation sources and the accurate implementation of suppression strategies. Regarding oscillation source location, existing methods mostly rely on critical impedance ratio calculations based on impedance methods or participation factor calculations based on mode analysis, which typically only locate the oscillation source to a relatively broad region. In terms of generating and executing suppression strategies, existing technologies largely rely on preset fixed thresholds and static control logic, lacking adaptability and foresight. The suppression parameters are difficult to adapt to the dynamic changes in the actual operation of wind farms, leading to poor control effects or problems such as overly conservative or undercompensated approaches. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a wind farm subsynchronous oscillation suppression method based on energy flow, which solves the problems of ambiguous oscillation source location and lack of dynamic adaptability of suppression strategies in the prior art.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a method for suppressing subsynchronous oscillations in wind farms based on energy flow, comprising, Collect time series data of wind farms, preprocess the data by combining the geographical partitioning of wind turbine locations, extract parameters of subsynchronous oscillation mode, and generate trigger signals; Based on the trigger signal, multi-band voltage and current components are extracted, and the instantaneous value of oscillation energy flow is calculated based on the Poynting vector principle to generate the oscillation energy flow vector; Based on the oscillating energy flow vector, an oscillating energy flow vector map is plotted on the electrical wiring diagram of the wind farm, and the direction, convergence point and divergence point of the oscillating energy flow vector are identified to generate the coordinates of the excitation point of the subsynchronous oscillation. Based on the coordinates of the excitation point of the subsynchronous oscillation and the electrical distance between each wind turbine in the wind farm, a targeted suppression strategy is dynamically formulated. A targeted suppression strategy is implemented, which integrates the changes in electrical quantities at the excitation point coordinates of the subsynchronous oscillation with the real-time collected wind farm response data to generate a subsynchronous oscillation suppression effect report.

[0007] As a preferred embodiment of the wind farm subsynchronous oscillation suppression method based on energy flow described in this invention, the wind farm time series data includes three-phase voltage, three-phase current instantaneous values, wind turbine geographical zoning identifiers, and terrain feature parameters. The preprocessing includes eliminating high-frequency noise, filtering out the influence of fundamental power components, enhancing the signal-to-noise ratio, and terrain-region weighted correction.

[0008] As a preferred embodiment of the energy flow-based wind farm subsynchronous oscillation suppression method of the present invention, the specific steps for extracting parameters of the subsynchronous oscillation mode and generating a trigger signal are as follows: Based on the preprocessed wind farm time series data, the parameters of the subsynchronous oscillation mode are identified and extracted by orthogonal projection and least squares method, and a list of mode parameters with geographic partition identification is generated. Based on the list of modal parameters identified by the geographic zoning, the damping ratio of the subsynchronous oscillation mode of each zoning is compared with the safety threshold in real time, and a comprehensive risk assessment is conducted in conjunction with the rate of change of oscillation amplitude to generate a real-time risk level and trigger signal.

[0009] As a preferred embodiment of the wind farm subsynchronous oscillation suppression method based on energy flow described in this invention, the steps of extracting multi-frequency voltage and current components according to the trigger signal, calculating the instantaneous value of the oscillation energy flow based on the Poynting vector principle, and generating the oscillation energy flow vector are as follows: Call up the instantaneous values ​​of three-phase voltage and three-phase current aligned with the trigger signal timestamp, filter out the power frequency component and risk frequency component through a bandpass filter, and extract multi-frequency voltage and current components; Based on multi-band voltage and current components, and combined with the Poynting vector principle, the instantaneous value of oscillating energy flow is calculated and a broadband energy flow matrix is ​​constructed; Based on the broadband energy flow matrix, the instantaneous values ​​of the three-phase current are bound to the spatial coordinates of the measurement nodes to generate an oscillating energy flow vector containing spatial position and direction information.

[0010] As a preferred embodiment of the wind farm subsynchronous oscillation suppression method based on energy flow described in this invention, the following steps are taken: Based on the oscillation energy flow vector, an oscillation energy flow vector map is drawn on the wind farm electrical wiring diagram, and the direction, convergence point, and divergence point of the oscillation energy flow vector are identified to generate the coordinates of the excitation point of the subsynchronous oscillation. The oscillating energy flow vector is rendered on the wind farm electrical wiring diagram using computer graphics algorithms to generate an oscillating energy flow vector map. The direction of oscillating energy flow vectors in the oscillating energy flow vector spectrum is statistically analyzed, and the energy flow trend is determined. By applying graph theory algorithms to the topological structure, the convergence and divergence points of the oscillating energy flow vectors are identified. Based on the convergence and divergence points of the oscillating energy flow vector, detailed geographical information is obtained by associating it with the wind farm asset database. Through multi-source data fusion and coordinate mapping, the coordinates of the excitation point of the subsynchronous oscillation are generated.

[0011] As a preferred embodiment of the energy flow-based wind farm subsynchronous oscillation suppression method of the present invention, the specific steps of dynamically formulating a targeted suppression strategy based on the coordinates of the excitation point of the subsynchronous oscillation and the electrical distance between each wind turbine in the wind farm are as follows. Traverse the connected paths from the excitation point coordinates of the subsynchronous oscillation to the target wind turbine, filter the shortest electrical paths, and generate a list of electrical distances; Based on the electrical distance list, the suppression intensity coefficient is obtained through the attenuation function mapping method, and the overall suppression intensity is dynamically adjusted according to the real-time risk level to generate a preliminary suppression instruction set; Based on the initial set of suppression instructions, the power adjustment value is verified to identify changes in the total power across the entire field, and a targeted suppression strategy is dynamically formulated.

[0012] As a preferred embodiment of the energy flow-based wind farm subsynchronous oscillation suppression method of the present invention, the steps for obtaining the electrical quantity changes are as follows: Execute the targeted suppression strategy, record the precise timestamps of strategy execution and instruction confirmation feedback signals, and generate process data packets; Extract the instantaneous values ​​of three-phase voltage and three-phase current at the excitation point coordinates of the subsynchronous oscillation from the process data packet, and perform differential analysis with the baseline electrical quantity data before implementing the targeted suppression strategy to obtain the change in electrical quantity at the excitation point coordinates.

[0013] As a preferred embodiment of the energy flow-based wind farm subsynchronous oscillation suppression method of the present invention, the integration of the real-time acquired wind farm response data to generate a subsynchronous oscillation suppression effect report comprises the following specific steps. By combining the changes in electrical quantities at the excitation point coordinates with the real-time acquired wind farm response data, precise alignment and spatial correlation matching are performed to generate a multidimensional suppression response dataset. Based on a multidimensional suppression response dataset, the evaluation results are obtained through a multidimensional comprehensive evaluation method, and a report on the suppression effect of subsynchronous oscillations is generated.

[0014] In a second aspect, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the energy flow-based wind farm subsynchronous oscillation suppression method described in the first aspect of the present invention.

[0015] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the energy flow-based wind farm subsynchronous oscillation suppression method described in the first aspect of the present invention.

[0016] The beneficial effects of this invention are as follows: By converting electrical signals into energy flow vectors using the Poynting vector principle, dynamic visualization of oscillation energy is achieved, enabling precise capture of energy propagation paths and achieving the beneficial effect of non-disruptive, high-precision location of oscillation sources. Through graph theory algorithm analysis of the energy flow map and association with an asset database, a precise mapping of the excitation point from electrical nodes to geographical coordinates is realized. This provides a targeted objective for suppression, achieving the beneficial effect of efficient oscillation suppression with minimal power generation loss. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a wind farm subsynchronous oscillation suppression method based on energy flow.

[0019] Figure 2 This is a flowchart of data preprocessing and modal parameter extraction.

[0020] Figure 3 A flowchart for generating energy flow vectors and plotting spectra.

[0021] Figure 4 A flowchart for the formulation and execution of suppression strategies. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] Reference Figures 1-4 As one embodiment of the present invention, this embodiment provides a method for suppressing subsynchronous oscillations in wind farms based on energy flow, comprising the following steps: S1: Collect time series data of wind farms, preprocess the data by combining the geographical partitions of wind turbine locations, extract parameters of subsynchronous oscillation modes, and generate trigger signals; S1.1: Wind farm time series data includes instantaneous values ​​of three-phase voltage and three-phase current, wind turbine geographical zoning identifiers, and terrain feature parameters; Specifically, the acquisition of wind farm time series data begins with broadband measurements deployed at the wind farm grid connection point and wind turbine terminals. The internal sensors of the broadband measurements are coupled to the power lines through voltage transformers and current transformers to sense the instantaneous values ​​of three-phase voltage and three-phase current in analog form. After the analog signals are filtered out by an anti-aliasing filter to remove high-frequency noise, they are synchronously sampled and quantized by a high-precision analog-to-digital converter at a fixed sampling rate to generate discrete digital sequences. A high-precision synchronous clock source is added to the discrete digital sequence to provide timestamps, forming time-stamped samples of instantaneous three-phase voltage and current values. The geographic information table and wind turbine attribute table in the wind farm asset database are accessed, and the wind turbine's corresponding latitude and longitude coordinates, altitude, slope, aspect, and terrain zone number are matched according to the wind turbine number field. These terrain feature parameters and the wind turbine's geographic zone identifier are written into the timestamp frame corresponding to the wind turbine's sampled data. Through field binding, the instantaneous three-phase voltage and current values ​​are synchronized with the wind turbine's geographic zone identifier and terrain feature parameters, forming wind farm time series data containing instantaneous three-phase voltage and current values, wind turbine geographic zone identifier, and terrain feature parameters.

[0026] S1.2: Preprocessing includes eliminating high-frequency noise, filtering out the influence of fundamental power components, enhancing the signal-to-noise ratio, and terrain-zone weighted correction; Specifically, the preprocessing of wind farm time series data involves: passing the instantaneous values ​​of three-phase voltage and three-phase current through a low-pass filter with a cutoff frequency set at 400 Hz to eliminate high-frequency noise; using a band-stop filter with a center frequency of 50 Hz or constructing a power frequency notch filter in a synchronous rotating coordinate system to filter out the influence of the fundamental power component; applying a wavelet transform denoising algorithm to the processed signal, selecting appropriate wavelet basis functions and decomposition levels for multi-resolution identification, and performing soft thresholding on detail coefficients to enhance the signal-to-noise ratio; establishing a terrain zoning weighting matrix based on the wind turbine geographical zoning identifier and terrain feature parameters, with each row of the terrain zoning weighting matrix corresponding to a wind turbine geographical zoning number and each column corresponding to the sampling point of the instantaneous values ​​of three-phase voltage and three-phase current; obtaining the element values ​​of the terrain zoning weighting matrix based on the wind turbine terrain feature parameters; performing weighted correction on the terrain zoning weighting matrix; and rearranging the terrain-zoning weighted corrected signal in chronological order to form the terrain-weighted preprocessing result, resulting in a clean signal suitable for subsynchronous component extraction after terrain zoning weighting correction.

[0027] S1.3: Based on the preprocessed wind farm time series data, the parameters of the subsynchronous oscillation mode are identified and extracted by orthogonal projection and least squares method, and a list of mode parameters with geographic partition identification is generated; Specifically, based on the preprocessed wind farm time series data, it is divided into past and future parts in chronological order and arranged in matrix form. The past output row contains historical wind farm time series data, and the future output row contains future wind farm time series data. A block Hankel matrix containing the past and future output rows is constructed. The columns of the block Hankel matrix correspond one-to-one with the wind turbine geographical region identifiers, and the wind turbine geographical region identifier field is retained. An orthogonal projection operation is performed on the row space of the block Hankel matrix to obtain the projection matrix. The column space of the projection matrix is ​​decomposed (column space decomposition is a matrix factorization operation performed on the projection matrix, specifically decomposing the projection matrix into the product of three matrices using the singular value decomposition algorithm: the left singular vector matrix, the ... The state sequence is estimated from the diagonal matrix and the right singular vector matrix. The state space equation is constructed with the state sequence as input and the future output part of the block Hankel matrix as output. The state matrix and output matrix in the state space equation are solved by the least squares method. The state matrix is ​​decomposed into complex eigenvalues. The imaginary part of the eigenvalues ​​is converted into the oscillation frequency. The damping ratio is obtained according to the relationship between the real part of the eigenvalue and the modulus. The estimated mode shape information is bound to the wind turbine geographical zone identifier at the field level to form a record entry containing frequency value, damping ratio, mode shape information and wind turbine geographical zone identifier. The parameters of the identified synchronous oscillation modes of each order are arranged from low to high frequency to generate a modal parameter list containing frequency value, damping ratio and mode shape information.

[0028] S1.4: Based on the list of modal parameters identified by the geographic partition, the damping ratio of the subsynchronous oscillation mode of each partition is compared with the safety threshold in real time, and a comprehensive risk assessment is performed in combination with the rate of change of oscillation amplitude to generate a real-time risk level and trigger signal.

[0029] Furthermore, the damping ratio of each synchronous oscillation mode in the modal parameter list of the geographic zoning identifier is continuously monitored. Each damping ratio is compared with a safety threshold in real time. Simultaneously, the oscillation amplitude of the corresponding oscillation frequency is extracted from the preprocessed wind farm time series data according to the one-to-one correspondence between the geographic zoning identifier and the modal frequency, and the rate of change per unit time is obtained. Based on fuzzy logic rules, the degree to which the damping ratio is lower than the safety threshold and the positive and continuously increasing trend of the oscillation amplitude change rate are mapped using membership degree mapping and Mamdani minimum-maximum inference. The centroid method is used to defuzzify and obtain the zoning-level risk value. When the damping ratio corresponding to the same geographic zoning identifier is lower than the safety threshold, the risk value is determined. When the damping ratio is below the safety threshold for several consecutive periods and the rate of change of oscillation amplitude exceeds the critical value (e.g., ≥8% / second), the comprehensive risk assessment result is determined to be of high risk level, and a trigger signal is immediately generated, which includes the risk mode frequency value, geographical partition identifier, timestamp, and high risk level identifier. When the damping ratio is close to the safety threshold and the rate of change of oscillation amplitude is in the medium-speed growth range (e.g., 5% / second), the comprehensive risk assessment result is determined to be of medium risk level and a trigger signal is generated. When the damping ratio is higher than the safety threshold and the rate of change of oscillation amplitude is not positive (e.g., ≤0% / second), the comprehensive risk assessment result is determined to be of low risk level and a trigger signal is generated.

[0030] It should be noted that the safety threshold is usually in the range of 0.03 to 0.05. The safety threshold is a critical value of damping ratio determined by statistically analyzing the high percentile of damping ratio under stable operating conditions in historical wind farm operating data and combining it with safety margin requirements. It is used to assess the risk of subsynchronous oscillation in real time.

[0031] Fuzzy logic rules are a set of conditional statements based on fuzzy set theory, used to handle the continuous changes in two input variables: the degree to which the damping ratio is below the safety threshold and the rate of change of oscillation amplitude. Fuzzy logic rules transform precise input values ​​into fuzzy variables (for example, quantifying the difference between the damping ratio and the safety threshold using triangular or trapezoidal membership functions into linguistic values ​​such as "slightly low," "moderately low," and "severely low," and quantifying the rate of change of oscillation amplitude into linguistic values ​​such as "slowly increasing," "medium-speed increasing," and "rapidly developing"). Fuzzy logic rules activate the corresponding set of fuzzy rules, describing the input-output relationship in "IF-THEN" form (for example, if the damping ratio is severely low and the rate of change of oscillation amplitude is rapidly increasing, then the risk level is high risk). Fuzzy logic rules perform fuzzy inference using the Mamdani min-maximum inference method and use a centroid-weighted average algorithm to transform the fuzzy conclusions into precise comprehensive risk assessment results.

[0032] S2: Based on the trigger signal, extract the multi-band voltage and current components, and calculate the instantaneous value of the oscillation energy flow based on the Poynting vector principle to generate the oscillation energy flow vector; S2.1: Call up the instantaneous values ​​of three-phase voltage and three-phase current aligned with the trigger signal timestamp, filter out the power frequency component and risk frequency component through a bandpass filter, and extract multi-frequency voltage and current components; Specifically, the risk frequency parameter contained in the trigger signal is read, and the passband frequency range of the finite-length unit impulse response bandpass filter is configured. The lower limit of the passband is set as the difference between the risk frequency value and the set passband frequency offset (e.g., 5 Hz), and the upper limit of the passband is set as the weighted sum of the risk frequency value and the set passband frequency offset. The coefficients of the finite-length unit impulse response bandpass filter that meet the passband requirements are generated using the window function method (e.g., Hanning window). The instantaneous values ​​of the three-phase voltage and three-phase current are convolved with the filter coefficients respectively. The zero-phase filtering method is used to eliminate phase distortion, retain the frequency components within the passband, and filter out the power frequency component and high and low frequencies outside the passband. The composition lays the foundation for multi-band component extraction. Based on passband filtering, three frequency bands are set: low-frequency passband (example range 0.1Hz to 10Hz), subsynchronous passband (example range 10Hz to 50Hz), and high-frequency passband (example range 50Hz to 300Hz). Finite-length unit impulse response bandpass filters of the corresponding frequency bands are applied to the instantaneous values ​​of three-phase voltage and three-phase current, and convolution operations are performed. Zero-phase filtering is used to eliminate the phase difference of the filtering results of different frequency bands. The output includes multi-band voltage and current components, including low-frequency voltage and current components, subsynchronous voltage and current components, and high-frequency voltage and current components.

[0033] S2.2: Based on multi-band voltage and current components, combined with the Poynting vector principle, calculate the instantaneous value of oscillation energy flow and construct a broadband energy flow matrix; Specifically, based on the physical nature of the relationship between instantaneous power flow density and electromagnetic field strength in the Poynting vector principle, scalar multiplication is performed on the instantaneous values ​​of the subsynchronous voltage component and the subsynchronous current component at the same measurement point and time. Point-to-point multiplication is then performed on the instantaneous values ​​of phase A voltage and phase A current, phase B voltage and phase B current, and phase C voltage and phase C current in the three-phase circuit to obtain the instantaneous power of each phase. The three-phase instantaneous power is then summed to obtain the total instantaneous power value. This total instantaneous power value is the instantaneous value of the oscillation energy flow, which characterizes the magnitude and direction of the subsynchronous oscillation energy transmission. The instantaneous values ​​of the low-frequency oscillation energy flow, the subsynchronous frequency oscillation energy flow, and the high-frequency oscillation energy flow are then combined column-wise in a time-aligned manner to form a broadband energy flow matrix.

[0034] The formula for calculating the instantaneous value of oscillating energy flow is: ; in, This represents the instantaneous value of the oscillating energy flow; Indicates in time, Instantaneous value of phase voltage; Indicates in time, Instantaneous value of phase current; Indicates in time, Instantaneous value of phase voltage; Indicates in time, Instantaneous value of phase current; Indicates in time, Instantaneous value of phase voltage; Indicates in time, Instantaneous value of phase current.

[0035] S2.3: Based on the broadband energy flow matrix, the instantaneous values ​​of the three-phase currents are bound to the spatial coordinates of the measurement nodes to generate an oscillating energy flow vector containing spatial position and direction information.

[0036] Specifically, based on the broadband energy flow matrix, the instantaneous three-phase current data recorded at each measurement node are bound to the spatial coordinates of the measurement node. The absolute value of the instantaneous energy flow of each frequency band in the broadband energy flow matrix obtained at each measurement node is used as the vector magnitude. The spatial coordinate data of the measurement node (e.g., latitude and longitude coordinates of 120.35°E, 36.87°N, or a distance percentage relative to the starting point of the collector line of 0.65) is read. At the same measurement node, the phase ratios of the low-frequency voltage component and current component, the sub-synchronous frequency band voltage component and current component, and the high-frequency voltage component and current component are compared. The orientation relationship is such that when the phase of the current component lags behind the phase of the voltage component, the orientation angle is recorded as zero degrees (indicating that energy flows from the grid to the wind farm), and when the phase of the current component leads the phase of the voltage component, the orientation angle is recorded as 180 degrees (indicating that energy flows from the wind farm to the grid). The absolute value of the instantaneous value of the oscillation energy flow in each frequency band, the spatial coordinates of the measurement node, and the orientation angle are combined into a set of spatial vector data records. The low-frequency band oscillation energy flow vector, the subsynchronous frequency band oscillation energy flow vector, and the high-frequency band oscillation energy flow vector are combined according to the timestamp alignment method to form an oscillation energy flow vector containing spatial position, oscillation energy flow magnitude, and orientation angle.

[0037] S3: Based on the oscillating energy flow vector, draw the oscillating energy flow vector map on the wind farm electrical wiring diagram, identify the direction, convergence point and divergence point of the oscillating energy flow vector, and generate the coordinates of the excitation point of the subsynchronous oscillation; S3.1: The oscillating energy flow vector is rendered on the wind farm electrical wiring diagram using computer graphics algorithms to generate an oscillating energy flow vector map; Furthermore, an oscillating energy flow vector map is generated using computer graphics algorithms. The specific steps are as follows: In the coordinate space transformation stage, the geographic coordinates of the oscillating energy flow vectors are transformed into Cartesian coordinates using Mercator projection or Albers conic projection. Then, an affine transformation matrix is ​​used to convert the Cartesian coordinates into pixel coordinates on the computer screen, including translation to move the origin to the canvas center, rotation to align with the actual geographic north and the top of the screen, and scaling to adjust the display ratio according to the canvas size. In the vector graphics stage, a vector graphic element with an arrow is created for each oscillating energy flow vector. The arrow length is determined by linear interpolation or a piecewise function mapping based on the vector magnitude. The arrow rotation angle is obtained based on the vector direction angle, and the arrow shape is generated using a polygon filling algorithm. Then… Color mapping and rendering are performed. The oscillating energy flow vector magnitude is encoded using a continuous red-yellow-green gradient color scheme, mapping the minimum magnitude to green, the median to yellow, and the maximum to red. Bilinear interpolation is used to obtain the intermediate hue. A vertex shader program from the graphics application programming interface is used to color each arrow vertex, and a depth test is initiated to eliminate aliasing. Layer compositing and annotation are then performed. First, the wind farm electrical wiring diagram is drawn as the underlying background. Then, arrows are drawn sequentially using a painter's algorithm, following the order of oscillating energy flow vector magnitude from largest to smallest, to avoid occlusion. A legend is added to illustrate the correspondence between color and magnitude, a scale bar is added to indicate the actual distance, and a compass is added to indicate direction. This generates an oscillating energy flow vector map containing spatial distribution, vector direction, and energy intensity information, completing the entire rendering process.

[0038] It should be noted that depth testing for anti-aliasing is a comprehensive rendering method in computer graphics that combines depth buffering and anti-aliasing. Its core principle is to accurately manage the occlusion relationship of graphic elements in 3D space through depth testing (Z-test), ensuring that objects in the background are correctly occluded. At the same time, it uses multi-sampling anti-aliasing (MSAA) or post-processing anti-aliasing (such as FXAA) algorithms to sample and blend the pixels of the vector graphics edges multiple times. By smoothing the color transition, it eliminates the jagged edges caused by pixelation, so that the arrow graphics in the generated oscillating energy flow vector map can present smooth edges and clear layers even in complex superposition states.

[0039] S3.2: Statistically analyze the direction of oscillating energy flow vectors in the oscillating energy flow vector map, determine the energy flow trend, and use graph theory algorithms to identify the convergence and divergence points of oscillating energy flow vectors through topological structures; Specifically, the direction angles of all oscillating energy flow vectors in the oscillating energy flow vector spectrum are statistically analyzed to obtain a histogram of direction angle distribution. Using a histogram distribution analysis method based on circular statistics, the most frequently occurring dominant direction interval is identified, thus determining that the overall energy flow trend is propagation from southwest to northeast. Next, combining the topology of the wind farm electrical wiring diagram, the diagram is converted into a weighted directed graph network, where nodes represent measurement points, directed edges represent the direction of the oscillating energy flow vector, and edge weights represent the vector magnitude. A depth-first search algorithm from graph theory is applied to traverse the directed graph network, marking all points forming sinks. Nodes entering the structure are initially identified as convergence points of the oscillating energy flow vector. Simultaneously, a reverse breadth-first search algorithm is applied to trace back along the reverse direction of the vector from each node, marking the common origin node of all paths, and these nodes are initially identified as divergence points of the oscillating energy flow vector. The flow of the initially identified convergence and divergence points is verified by obtaining the net flow value of the adjacent edges of each candidate node. Nodes with net flow values ​​greater than the positive direction are confirmed as convergence points of the oscillating energy flow vector, and nodes with net flow values ​​less than the negative direction are confirmed as divergence points of the oscillating energy flow vector, generating a list of convergence and divergence points of the oscillating energy flow vector.

[0040] It should be noted that the graph theory algorithm mainly consists of two core parts: depth-first search (DFS) and reverse breadth-first search (BFS). The DFS recursively traverses all connected paths in the weighted directed graph network, prioritizing the exploration of the terminal nodes of each branch path. This identifies the merging structural nodes (i.e., the endpoints of multiple energy flow input paths) with in-degrees much greater than out-degrees, and preliminarily determines them as the convergence points of oscillating energy flow vectors. The BFS starts from each node and backtracks layer by layer along the reverse direction of the directed edges. Through a queue structure, it achieves hierarchical expansion traversal and quickly locates the common origin node of multiple backtracking paths (i.e., the node with an out-degree much greater than its in-degree and located at the starting point of energy flow divergence), and preliminarily determines it as the divergence point of oscillating energy flow vectors. These two algorithms work together to reveal the propagation path and distribution characteristics of oscillating energy flow in the wind farm topology network from the two dimensions of "convergence" and "source tracing," respectively.

[0041] S3.3: Based on the convergence and divergence points of the oscillation energy flow vector, detailed geographical information is obtained by associating with the wind farm asset database. Through multi-source data fusion and coordinate mapping, the coordinates of the excitation point of the subsynchronous oscillation are generated.

[0042] Specifically, the system accesses the equipment ledger in the wind farm asset database, precisely matches the node numbers in the list with the equipment codes in the ledger, and obtains the complete attribute records of the corresponding equipment, including fields such as equipment type, equipment code, and installation latitude and longitude coordinates. For successfully matched equipment nodes, their latitude and longitude coordinates are directly read as preliminary coordinates. For nodes located on transmission line segments, the system calls the line spatial database to query the precise latitude and longitude coordinates of the towers at both ends of the line segment. Based on the percentage position of the node in the line electrical parameters, the geographical coordinates of the point are determined using a linear interpolation algorithm. Next, multi-source data fusion is performed, integrating the node electrical attributes from the oscillation energy flow vector map, the equipment static attributes from the wind farm asset database, and the geospatial attributes from the coordinate mapping, and filling the data fields according to a standardized format. This generates a subsynchronous oscillation excitation point coordinate data package containing complete information such as the excitation point equipment type, equipment code, latitude and longitude coordinates, the electrical zone to which it belongs, and the oscillation energy intensity.

[0043] It should be noted that the wind farm asset database is a structured data collection that centrally stores information on all physical equipment and attributes of a wind farm. Its construction begins with collecting data on the entire lifecycle of equipment such as wind turbines, box-type transformers, collector lines, and towers. This includes nameplate parameters entered during the equipment procurement phase, latitude and longitude coordinates recorded during the installation phase, equipment codes generated during the commissioning phase, and updated ledger information during the operation and maintenance phase. The wind farm asset database adopts a relational structure, using equipment codes as primary keys to establish relationships between equipment ledger tables, geographic information tables, and electrical parameter tables, ensuring data consistency and traceability. The wind farm asset database maintains dynamic updates to the real-time status and static attributes of equipment by periodically synchronizing data with the data acquisition and monitoring control platform and the asset management platform. This forms a unified data source containing multi-dimensional information such as equipment spatial location, electrical characteristics, and operating status, providing accurate equipment coordinates and attribute support for oscillation source location.

[0044] The line spatial database is a structured database specifically designed to store the spatial geometric information and topological relationships of wind farm transmission lines. Its generation process begins with the precise latitude and longitude coordinates of transmission line towers at various voltage levels obtained through mapping using the Global Navigation Satellite System. This is combined with digitally acquired parameters such as span and sag from the line design drawings, and the electrical connections of the lines are integrated to construct a topological model. Its core function is to provide spatial coordinate calculation services for any point on the line segment (such as a fault point or oscillation measurement point). Through linear interpolation algorithms, combined with tower coordinates and electrical parameters (such as distance percentage), the precise geographical location of points on the line is determined, providing spatial data support for oscillation source location.

[0045] S4: Based on the coordinates of the excitation point of the subsynchronous oscillation and the electrical distance between each wind turbine in the wind farm, a targeted suppression strategy is dynamically formulated. S4.1: Traverse the connecting paths from the excitation point coordinates of the subsynchronous oscillation to the target wind turbine, filter the shortest electrical paths, and generate a list of electrical distances; Furthermore, taking the node corresponding to the excitation point coordinates as the starting point, Dijkstra's shortest path algorithm is used for traversal based on a weighted directed graph network. In the initialization phase, the distance of the starting node is set to zero, and the distances of other nodes are set to infinity. The starting node is then added to a priority queue. In the iteration phase, the node with the smallest current distance is taken from the priority queue, and all adjacent nodes of this node are traversed. The path distance from the current node to the adjacent node is obtained. If the new distance is less than the original distance of the adjacent node, the distance value is updated and the adjacent node is added to the priority queue. This process is repeated until all reachable nodes are processed. The shortest electrical path is obtained by backtracking the path record from the target wind turbine node. An electrical distance list containing the number of each target wind turbine node, the total cable length of the corresponding shortest electrical path, and the total equivalent reactance is generated.

[0046] S4.2: Based on the electrical distance list, the suppression intensity coefficient is obtained through the attenuation function mapping method, and the overall suppression intensity is dynamically adjusted according to the real-time risk level to generate a preliminary suppression instruction set; Specifically, the electrical distance value corresponding to each wind turbine node in the list is read, and the suppression intensity coefficient is obtained by applying the attenuation function mapping method. The attenuation function adopts the form of a negative exponential curve (for example, the suppression intensity coefficient of the wind turbine at the excitation point is set to 1.0, and the suppression intensity coefficient of the wind turbine directly upstream of the excitation point is set to 0.8). The overall suppression intensity is dynamically adjusted according to the real-time risk level. The real-time risk level is read. If the risk level is high, the final adjustment value is determined by combining the maximum power reduction limit with the suppression intensity coefficient. If the risk level is medium, 70% of the maximum power reduction limit is combined with the suppression intensity coefficient. A preliminary suppression instruction set is generated.

[0047] S4.3: Based on the initial set of suppression instructions, verify the power adjustment value to identify changes in the total power across the entire field, and dynamically formulate targeted suppression strategies.

[0048] Specifically, the process reads the power adjustment value of each wind turbine in the set, verifies whether these values ​​exceed the upper and lower limits of safe operation of the wind turbine (e.g., minimum power 0 kW and maximum power 1500 kW), and simulates whether the power adjustment leads to local voltage exceeding the limit. Then, it summarizes the power adjustment values ​​of all wind turbines to calculate the total power change of the entire field, and assesses whether the total power change exceeds the power fluctuation range allowed by the dispatch center (e.g., the total change should not exceed 100 kW, determined based on statistical analysis of historical wind farm operation data). If an anomaly is found during the verification, a dynamic adjustment process is initiated to smoothly correct the power adjustment values ​​exceeding the limit and to coordinate and optimize the total power change of the entire field, generating a standardized targeted suppression strategy. The targeted suppression strategy includes power reduction instructions for the excitation point wind turbines, gradient adjustment instructions for upstream wind turbines, and strategy execution timing parameters.

[0049] S5: Implement a targeted suppression strategy, integrate the changes in electrical quantities at the excitation point coordinates of the subsynchronous oscillation with the real-time collected wind farm response data, and generate a subsynchronous oscillation suppression effect report.

[0050] S5.1: Execute the targeted suppression strategy, record the precise timestamp of strategy execution and instruction confirmation feedback signal, and generate a process data packet; Specifically, a power adjustment command is issued to the target wind turbine controller. The moment the command is issued, a high-precision clock source is triggered to record the Coordinated Universal Time (UTC) timestamp. At the same time, the command status monitoring process is started to continuously receive the confirmation feedback signal returned by the wind turbine controller. The timestamp information, the content of the power adjustment command (including the wind turbine number, target power value, and rate of change limit), and the command confirmation feedback signal are associated and encapsulated. A process data package containing the strategy execution sequence number, command issuance timestamp, target power setpoint for each wind turbine, actual power change curve sampling points, and command execution status code is generated according to a standardized format.

[0051] S5.2: Extract the instantaneous values ​​of three-phase voltage and three-phase current at the excitation point coordinates of the subsynchronous oscillation from the process data packet, and perform differential analysis with the baseline electrical quantity data before the targeted suppression strategy is implemented to obtain the change in electrical quantity at the excitation point coordinates; Furthermore, the device code corresponding to the excitation point coordinates of the subsynchronous oscillation is extracted from the process data packet. Based on the device code, the specific measurement point is located in the wind farm measurement node list. The instantaneous values ​​of three-phase voltage and three-phase current recorded at the measurement point after the execution of the targeted suppression strategy are read. At the same time, the instantaneous values ​​of three-phase voltage and three-phase current under the same operating conditions before the execution of the targeted suppression strategy are retrieved from the wind farm asset database and the timestamps are precisely aligned. The voltage change is obtained by subtracting the voltage instantaneous value of each corresponding sampling point using the point-by-point difference method, and the current change is obtained by subtracting the current instantaneous value of each corresponding sampling point. The electrical quantity changes including the three-phase voltage change and the three-phase current change at the excitation point coordinates are obtained.

[0052] S5.3: Combine the changes in electrical quantities at the excitation point coordinates with the real-time acquired wind farm response data, perform precise alignment and spatial correlation matching, and generate a multidimensional suppression response dataset; Furthermore, electrical quantity change data at the excitation point coordinates of the subsynchronous oscillation are extracted from the process data packet, including the three-phase voltage change sequence and the three-phase current change sequence. Simultaneously, the response data of the entire field is collected in real time through the wind farm wide-area measurement device, including the instantaneous values ​​of the three-phase voltage, three-phase current, and active power of each node, and is precisely aligned. Using the timestamp provided by the high-precision synchronous clock source, the time series of electrical quantity change data and wind farm response data are aligned with millisecond precision to ensure that the time deviation of each data point is controlled within milliseconds. Then, spatial correlation matching is performed. According to the wind farm electrical wiring diagram, the latitude and longitude position of the excitation point coordinates are mapped with the spatial coordinates of the measurement nodes in the entire field to establish the topological connection relationship between the electrical quantity change of the excitation point and the response data of adjacent nodes. The time-aligned and spatially correlated data are integrated to generate a multi-dimensional suppression response dataset containing timestamps, spatial coordinates, electrical parameter changes, and response values.

[0053] S5.4: Based on the multidimensional suppression response dataset, the evaluation results are obtained through a multidimensional comprehensive evaluation method, and a report on the suppression effect of subsynchronous oscillations is generated.

[0054] Furthermore, based on the multidimensional suppression response dataset, the timestamp-aligned oscillation frequency sequence, damping ratio sequence, and oscillation energy flow amplitude sequence stored in the dataset are read. A multidimensional comprehensive evaluation method is applied to process the frequency dimension, assessing whether the absolute value of the shift in the dominant oscillation frequency before and after suppression is lower than the shift threshold (an exemplary range is 0.05Hz to 0.5Hz, determined based on power stability methods and statistical analysis of historical wind farm operation data). The damping dimension is evaluated, comparing the numerical change in the damping ratio before and after suppression and obtaining the percentage increase. The energy dimension is checked, measuring whether the attenuation rate of the oscillation energy flow amplitude reaches the target value. Then, a weighted comprehensive scoring mechanism is used, assigning weight coefficients to each dimension, and obtaining the overall suppression effect score through weighted summation. The results, including frequency shift statistics, damping ratio improvement curves, energy flow attenuation spectra, and comprehensive scores, are added with strategy execution timestamps, excitation point coordinates, and a wind turbine adjustment list to generate a subsynchronous oscillation suppression effect report containing an effect summary, data comparison, and conclusion recommendations.

[0055] This embodiment also provides a computer device applicable to the energy flow-based wind farm subsynchronous oscillation suppression method, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the energy flow-based wind farm subsynchronous oscillation suppression method proposed in the above embodiment.

[0056] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0057] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the energy flow-based wind farm subsynchronous oscillation suppression method proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0058] In summary, this invention utilizes the Poynting vector principle to transform electrical signals into energy flow vectors, achieving dynamic visualization of oscillating energy. This allows for precise capture of energy propagation paths, resulting in the beneficial effect of non-disruptive, high-precision oscillation source localization. By analyzing the energy flow map using graph theory algorithms and linking it to an asset database, a precise mapping of the excitation point from electrical nodes to geographic coordinates is achieved. This provides a targeted objective for suppression, achieving efficient oscillation suppression with minimal power generation loss.

[0059] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for suppressing subsynchronous oscillations in wind farms based on energy flow, characterized in that: include, Collect time series data of wind farms, preprocess the data by combining the geographical partitioning of wind turbine locations, extract parameters of subsynchronous oscillation mode, and generate trigger signals; Based on the trigger signal, multi-band voltage and current components are extracted, and the instantaneous value of oscillation energy flow is calculated based on the Poynting vector principle to generate the oscillation energy flow vector; Based on the oscillating energy flow vector, an oscillating energy flow vector map is plotted on the electrical wiring diagram of the wind farm, and the direction, convergence point and divergence point of the oscillating energy flow vector are identified to generate the coordinates of the excitation point of the subsynchronous oscillation. Based on the coordinates of the excitation point of the subsynchronous oscillation and the electrical distance between each wind turbine in the wind farm, a targeted suppression strategy is dynamically formulated. A targeted suppression strategy is implemented, which integrates the changes in electrical quantities at the excitation point coordinates of the subsynchronous oscillation with the real-time collected wind farm response data to generate a subsynchronous oscillation suppression effect report.

2. The method for suppressing subsynchronous oscillations in wind farms based on energy flow as described in claim 1, characterized in that: The wind farm time series data includes instantaneous values ​​of three-phase voltage and three-phase current, wind turbine geographical zoning identifiers, and terrain feature parameters; The preprocessing includes eliminating high-frequency noise, filtering out the influence of fundamental power components, enhancing the signal-to-noise ratio, and terrain-region weighted correction.

3. The method for suppressing subsynchronous oscillations in wind farms based on energy flow as described in claim 2, characterized in that: The specific steps for extracting parameters of the subsynchronous oscillation mode and generating a trigger signal are as follows. Based on the preprocessed wind farm time series data, the parameters of the subsynchronous oscillation mode are identified and extracted by orthogonal projection and least squares method, and a list of mode parameters with geographic partition identification is generated. Based on the list of modal parameters identified by the geographic zoning, the damping ratio of the subsynchronous oscillation mode of each zoning is compared with the safety threshold in real time, and a comprehensive risk assessment is conducted in conjunction with the rate of change of oscillation amplitude to generate a real-time risk level and trigger signal.

4. The method for suppressing subsynchronous oscillations in wind farms based on energy flow as described in claim 3, characterized in that: The steps are as follows: Based on the trigger signal, multi-frequency voltage and current components are extracted, and the instantaneous value of the oscillation energy flow is calculated based on the Poynting vector principle to generate the oscillation energy flow vector. Call up the instantaneous values ​​of three-phase voltage and three-phase current aligned with the trigger signal timestamp, filter out the power frequency component and risk frequency component through a bandpass filter, and extract multi-frequency voltage and current components; Based on multi-band voltage and current components, and combined with the Poynting vector principle, the instantaneous value of oscillating energy flow is calculated and a broadband energy flow matrix is ​​constructed; Based on the broadband energy flow matrix, the instantaneous values ​​of the three-phase current are bound to the spatial coordinates of the measurement nodes to generate an oscillating energy flow vector containing spatial position and direction information.

5. The method for suppressing subsynchronous oscillations in wind farms based on energy flow as described in claim 4, characterized in that: The specific steps for plotting the oscillation energy flow vector map on the wind farm electrical wiring diagram, identifying the direction, convergence point, and divergence point of the oscillation energy flow vector, and generating the excitation point coordinates of the subsynchronous oscillation are as follows: The oscillating energy flow vector is rendered on the wind farm electrical wiring diagram using computer graphics algorithms to generate an oscillating energy flow vector map. The direction of oscillating energy flow vectors in the oscillating energy flow vector spectrum is statistically analyzed, and the energy flow trend is determined. By applying graph theory algorithms to the topological structure, the convergence and divergence points of the oscillating energy flow vectors are identified. Based on the convergence and divergence points of the oscillating energy flow vector, detailed geographical information is obtained by associating it with the wind farm asset database. Through multi-source data fusion and coordinate mapping, the coordinates of the excitation point of the subsynchronous oscillation are generated.

6. The method for suppressing subsynchronous oscillations in wind farms based on energy flow as described in claim 5, characterized in that: The specific steps for dynamically formulating a targeted suppression strategy based on the coordinates of the excitation point of the subsynchronous oscillation and the electrical distance between each wind turbine in the wind farm are as follows. Traverse the connected paths from the excitation point coordinates of the subsynchronous oscillation to the target wind turbine, filter the shortest electrical paths, and generate a list of electrical distances; Based on the electrical distance list, the suppression intensity coefficient is obtained through the attenuation function mapping method, and the overall suppression intensity is dynamically adjusted according to the real-time risk level to generate a preliminary suppression instruction set; Based on the initial set of suppression instructions, the power adjustment value is verified to identify changes in the total power across the entire field, and a targeted suppression strategy is dynamically formulated.

7. The method for suppressing subsynchronous oscillations in wind farms based on energy flow as described in claim 6, characterized in that: The steps for obtaining the changes in electrical quantities are as follows: Execute the targeted suppression strategy, record the precise timestamps of strategy execution and instruction confirmation feedback signals, and generate process data packets; Extract the instantaneous values ​​of three-phase voltage and three-phase current at the excitation point coordinates of the subsynchronous oscillation from the process data packet, and perform differential analysis with the baseline electrical quantity data before implementing the targeted suppression strategy to obtain the change in electrical quantity at the excitation point coordinates.

8. The method for suppressing subsynchronous oscillations in wind farms based on energy flow as described in claim 7, characterized in that: The data is integrated with the real-time acquired wind farm response data to generate a subsynchronous oscillation suppression effect report. The specific steps are as follows: By combining the changes in electrical quantities at the excitation point coordinates with the real-time acquired wind farm response data, precise alignment and spatial correlation matching are performed to generate a multidimensional suppression response dataset. Based on a multidimensional suppression response dataset, the evaluation results are obtained through a multidimensional comprehensive evaluation method, and a report on the suppression effect of subsynchronous oscillations is generated.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the wind farm subsynchronous oscillation suppression method based on energy flow as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the energy flow-based wind farm subsynchronous oscillation suppression method as described in any one of claims 1 to 8.

Images