Method and device for positioning low-frequency oscillation source of wind-thermal bundling system

Abstract

The invention discloses a method and a device for positioning a low-frequency oscillation source of a wind-thermal bundling system. The method comprises the following steps: acquiring multi-source signals of the wind-thermal bundling system; carrying out adaptive wavelet threshold anti-noise preprocessing; carrying out adaptive noise intensity set empirical mode decomposition; screening a wind-fire coupling key mode; calculating a wind-fire adaptive dissipated energy flow; and oscillation source determination based on phase coupling: constructing a multi-condition fusion determination logic in combination with a phase coupling relationship between a dissipated energy flow trend and a wind-fire key mode and a phase coupling relationship between the dissipated energy flow trend and an electric energy transmission signal so as to realize oscillation source type identification. According to the invention, accurate identification of oscillation sources (a wind power side, a thermal power side and a tie line) in a frequency band of 0.1-2Hz is realized, and both positioning accuracy and engineering real-time performance are considered.

Description

Technical Field

[0001] This invention belongs to the field of power system oscillation monitoring and control technology, specifically relating to a method and device for locating low-frequency oscillation sources in a wind-fire bundling system. Background Technology

[0002] Bundling wind and thermal power for grid connection has become a core mode for balancing renewable energy consumption and stable power system operation. This mode smooths out wind farm power fluctuations through the peak-shaving capacity of thermal power units, reduces carbon emissions by relying on the clean characteristics of wind power, and achieves stable power transmission to the main grid through interconnection lines. However, during power transmission, the dynamic characteristics of wind and thermal power units are prone to low-frequency oscillations: doubly-fed wind turbines have low equivalent inertia and fast rotor-side converter control response, while thermal power units have valve regulation delay and shaft inertia lag. The interaction between these two factors can easily generate low-frequency oscillations of 0.1 to 2 Hz. In addition, grid load fluctuations and equipment switching interference in the power transmission process introduce strong background noise into the system oscillation signal, further increasing the difficulty of locating the oscillation source.

[0003] Existing low-frequency oscillation source localization technologies are ill-suited for the power transmission scenarios of wind-thermal bundled systems: First, common methods such as the dissipative energy flow method based on empirical mode decomposition (EMD) and the impedance matching method based on variational mode decomposition (VMD) fail to consider the coupling characteristics of wind and thermal power and the laws of power transmission. They only analyze the active power signal of a single unit, ignoring key related signals such as wind turbine rotor speed, thermal power valve opening, and tie line power fluctuations, leading to positioning errors. Second, the noise immunity design has poor adaptability. Empirical mode decomposition (EEMD) relies on fixed-intensity white noise, which is prone to mode aliasing under strong interference noise in power transmission, making it impossible to accurately extract low-frequency oscillation components. Third, the oscillation source determination logic is simplistic, relying solely on the "rising / falling" trend of dissipative energy flow without considering the phase characteristics of wind-thermal coupling and power transmission. This makes it difficult to distinguish between "wind-generated oscillations," "thermal-generated oscillations," or "tie line-coupled oscillations," failing to meet the refined positioning requirements of wind-thermal bundled power transmission systems.

[0004] Therefore, it is necessary to develop a low-frequency oscillation source localization method for power transmission scenarios in wind-fire bundled systems, to solve the problems of insufficient adaptability of existing technologies, low extraction accuracy under strong interference noise, and one-sided judgment logic. Summary of the Invention

[0005] To address the technical challenges of "strong signal coupling, high noise interference, and one-sided judgment logic" in the location of low-frequency oscillation sources in wind-fired power bridging systems, this paper provides a method and device for locating low-frequency oscillation sources that features multi-source signal collaborative processing, strong noise resistance, and adaptability to wind-fired power coupling characteristics. This method enables accurate identification of oscillation sources (wind power side, thermal power side, and tie lines) in the 0.1 to 2 Hz frequency band, balancing positioning accuracy with real-time engineering performance.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: A method for locating the low-frequency oscillation source in a wind-fire bundling system includes: Step 1: Multi-source signal acquisition of the wind and fire bundling system Relying on a wideband measurement device with a sampling rate of no less than 1kHz, the original signals of the wind-fired power bundling system are collected simultaneously: wind power side signals, thermal power side signals, tie line signals and system parameters, to ensure coverage of the entire oscillation cycle; Step 2: Adaptive wavelet threshold noise reduction preprocessing Adaptive wavelet threshold denoising was performed on the original signals acquired in step 1 to obtain denoised signals. ; Step 3: Adaptive noise intensity set empirical mode decomposition Based on the denoised signal obtained in step 2 An improved adaptive noise intensity set empirical mode decomposition algorithm is used to perform mode decomposition, extract the intrinsic mode functions related to low-frequency oscillations in power transmission, and obtain candidate IMF components. Step 4: Screening of Key Modes for Wind-Fire Coupling Combining the inertial coupling characteristics of the wind-fire bundling system with the signal correlation features of power transmission, a three-dimensional screening index of "energy proportion - time-domain peak factor - inertial correlation degree" is designed to select the key modes that best reflect the low-frequency oscillation characteristics of power transmission from the candidate IMF components in step 3. ; Step 5: Calculation of wind-fire adapted dissipative energy flow Based on the key modes obtained in step 4 Combining the differences in dynamic characteristics between wind and thermal power units and the laws of power transmission, dissipation energy flow calculation models were constructed for the wind power side, thermal power side, and tie line, respectively. The phase coupling relationship between the dissipation energy flow trend and the key modes of wind and thermal power and power transmission signals was calculated. Step 6: Determining the Oscillation Source Based on Phase Coupling By combining the dissipated energy flow trend calculated in step 5 with the phase coupling relationship between the key wind and fire modes and the power transmission signal, a multi-condition fusion judgment logic is constructed to achieve oscillation source type identification.

[0007] A further improvement of this invention is that, in step 1: relying on a broadband measurement device with a sampling rate of not less than 1kHz, the wind power side signal, thermal power side signal, tie line signal, and system parameters of the wind-thermal bundling system are simultaneously acquired, including: Wind power side signal: Active power at the wind farm grid connection point Doubly fed fan rotor speed Rotor-side converter q shaft current Rotor-side converter dshaft voltage ; Thermal power side signal: Active power of thermal power units Thermal power unit speed Steam valve opening of thermal power unit Thermal power unit terminal voltage deviation ; Tie line signal: Active power of the tie line between the wind-fire bundling system and the main grid , tie line voltage amplitude System parameters: Rated inertia of thermal power unit J g Equivalent inertia of wind turbine units J w Connection line impedance .

[0008] A further improvement of this invention is that step 2: adaptive wavelet threshold denoising processing is performed on the multi-source signals acquired in step 1, including: 2.1 Wavelet decomposition: Choosing the db6 wavelet as the basis wavelet for the original signal Perform three-level wavelet decomposition to obtain the wavelet coefficients of each level. ,in j The number of decomposition layers, k This is the index of the wavelet coefficients for this layer; 2.2 Calculation of Noise and Signal Standard Deviation: Based on the first-level detail coefficients of wavelet decomposition, the noise standard deviation is calculated using a formula. : (1) In formula (1): median( () indicates the median operation. The detail coefficients of the first layer of wavelet decomposition are used; the original signal is calculated using the formula. Standard deviation : (2) In formula (2): N The number of signal sampling points. For the original signal i The values ​​of each sampling point; 2.3 Adaptive Threshold Design: Considering the difference in wavelet coefficient amplitude between the oscillating signal and the power transmission interference noise, a threshold that dynamically adjusts with the wavelet coefficient amplitude is designed, as shown in the following formula: (3) In formula (3): For the first j Layerk The absolute value of each wavelet coefficient is used to design a smaller threshold for oscillatory characteristic coefficients with larger amplitudes and a larger threshold for noise coefficients with smaller amplitudes. 2.4 Soft Thresholding and Signal Reconstruction: The wavelet coefficients are subjected to soft thresholding, as shown in the following formula: (4) In equation (4): sign( ) is a sign function, max( (,0) indicates taking the larger value between the input value and 0; the processed wavelet coefficients Wavelet reconstruction is performed to obtain the denoised signal. This serves as the input for subsequent mode decomposition.

[0009] A further improvement of the present invention is that step 3: based on the denoised signal obtained in step 2 An improved adaptive noise intensity set empirical mode decomposition algorithm is used to perform mode decomposition, extracting the intrinsic mode functions related to low-frequency oscillations in power transmission, including: 3.1 Adaptive Noise Intensity Design: Considering the fluctuation characteristics of the power transmission signal in the wind-fire bundled power transmission, the intensity of the added white noise is designed to decrease with the number of decompositions. The number of ensembles is set to 20. i White noise added during the decomposition strength The calculation formula is as follows: (5) In formula (5): For denoised signals The maximum absolute value, i The noise intensity varies with the number of decompositions. i The increment linearly decreases from 0.1 times the maximum absolute value of the signal to 0.01 times; 3.2 Multi-round EMD decomposition: For the first i Noisy signal after adding noise Perform EMD decomposition to obtain the first i The IMF components of the secondary decomposition and residual components , m This represents the number of IMFs in this decomposition. 3.3 Calculation of Average IMF Components and Selection of Candidate Modes: The arithmetic mean of the IMF components of the same order obtained from 20 decompositions is used to obtain the final IMF component, as shown in the following formula: (6) In equation (6), For the first k The final IMF component of the order, For the i-th decomposition, the first... k The final IMF components are analyzed using Fast Fourier Transform (FFT) to select IMF components with frequencies ranging from 0.1 to 2 Hz as candidate IMF components. , p The number of candidate components.

[0010] A further improvement of this invention lies in step 4: combining the inertial coupling characteristics of the wind-fire bundling system with the signal correlation characteristics of power transmission, a three-dimensional screening index of "energy proportion - time-domain peak factor - inertial correlation degree" is designed to screen the key modes that best reflect the low-frequency oscillation characteristics of power transmission from the candidate IMF components in step 3, including: 4.1 Energy percentage calculation: The energy percentage reflects the proportion of oscillation energy contained in the candidate IMF component within the total energy of all candidate components. A higher percentage indicates a greater contribution to the oscillation of power transmission. The formula is as follows: (7) In equation (7): For the first k One candidate IMF component t 0 represents the start time of oscillation. t 1 represents the termination time of the oscillation, and the numerator is the 1st oscillation. k The energy of each candidate component is divided by the denominator, which is the total energy of all candidate components. 4.2 Calculation of time-domain peak factor: The time-domain maxima factor reflects the amplitude stability of candidate IMF components. The smaller the maxima factor, the more stable the amplitude fluctuation of the component, and the more likely it is to be the true oscillation mode of the power transmission process. The formula is as follows: (8) In equation (8): For the first k The maximum absolute value of each candidate component is given by the denominator, which is the root mean square value of that component. 4.3 Inertia Correlation Calculation: The inertia correlation degree combines the inertia distribution of the wind and thermal power units with the correlation of power transmission signals, reflecting the degree of correlation between candidate components and the dynamic response of the wind and thermal power units and the power transmission status. The formula is as follows: (9) In equation (9): The rated inertia of the thermal power unit. This is the equivalent inertia of the wind turbine. For the correlation coefficient weight of thermal power, The weights of the wind power correlation coefficients are given by both, and they satisfy the following conditions: , The calculation formula is as follows: (10) In equation (10): corr( , ) is the function for calculating the Pearson correlation coefficient. The correlation coefficient between the candidate components and the rotational speed of the thermal power unit is given. The correlation coefficient between the candidate components and the power transmitted by the tie line. The correlation coefficient between the candidate components and the fan rotor speed is given. The correlation coefficient between the candidate components and the wind power transmission capacity; 4.4 Key Modality Screening: Set an energy percentage threshold of 0.1, a time-domain peak factor threshold of 5, and an inertia correlation threshold of 0.08 to filter candidates that simultaneously meet these criteria. , and The candidate components were identified as the key oscillation modes. If multiple components satisfy the conditions, choose the inertia correlation degree. The largest component; if no component simultaneously satisfies all three conditions, choose... and The components are used as key modes.

[0011] A further improvement of the present invention is that step 5: based on the key modes obtained in step 4 Based on the differences in dynamic characteristics between wind and thermal power units and the laws governing power transmission, dissipative energy flow calculation models are constructed for the wind power side, the thermal power side, and the tie line, including: 5.1 Calculation of dissipated energy flow on the thermal power plant side: Considering the transient impact of steam valve regulation on power transmission in thermal power units, steam valve opening is introduced. The term related to tie-line power is used as a correction term, and the formula is as follows: (11) In equation (11): For the dissipated energy flow on the thermal power side, The key modes are obtained by decomposing the active power signal of thermal power plants. For the rotational speed of the thermal power unit, This refers to the reactive power of thermal power units. This refers to the voltage deviation at the thermal power unit terminals. To transmit power to the tie line, This represents the maximum power of the tie line; 5.2 Calculation of energy dissipation flow on the wind power side: Considering the impact of rotor-side converter control on power transmission in doubly-fed wind turbines, the rotor-side converter control is introduced... q shaft current and d shaft voltage The product term, combined with the deviation between the rotor speed and the synchronous speed, adjusts the energy flow direction, as shown in the following formula: (12) In equation (12): This refers to the energy dissipation flow on the wind power side. The key modes are obtained by decomposing the active power signal of wind power. This refers to the rotor speed of the doubly fed fan. Synchronous speed; For rotor-side converter q shaft current, For rotor-side converter d shaft voltage; sign( ) is a symbolic function; To transmit power to wind power; This represents the maximum wind power output. 5.3 Calculation of energy dissipation flow in tie lines: Based on the impedance loss characteristics of the power transmission process, the real part of the tie line impedance is introduced to calculate the resistance loss, as shown in the following formula: (13) In equation (13), For the energy dissipation flow of the connection line, For the active power of the tie line, This refers to the amplitude of the tie line voltage. For the tie line impedance, This is the impedance magnitude. For tie line current, This represents the real part of the impedance.

[0012] A further improvement of this invention lies in step 6: combining the dissipated energy flow trend calculated in step 5 with the phase coupling relationship between the key wind and fire modes and the power transmission signal, a multi-condition fusion judgment logic is constructed to achieve oscillation source type identification, including: 6.1 Analysis of Dissipated Energy Flow Trends: Calculate the numerical value and rate of change of the dissipated energy flow on each side, if And the derivative of energy flow with respect to time This indicates that the thermal power plant is continuously injecting energy into the power transmission system; if and This indicates that the wind power side is continuously injecting energy into the power transmission system; if and This indicates that there is an energy coupling surplus in the transmission link of the tie line; among which, Calculated through numerical differentiation; 6.2 Calculation of phase coupling coefficient: The key modes of wind power were analyzed using Fast Fourier Transform. Key modes of thermal power and the transmission power of the connecting line Frequency analysis was performed to extract the phase information of the three components at the dominant oscillation frequency, and the phase difference between the two sets was calculated. Phase difference between wind power and thermal power : (14) Wind power-tether phase difference : (15) In equations (14) and (15), FFT( ) is the Fast Fourier Transform function, arg( () is the operation of taking the complex argument. , Both reflect the sequence of wind and fire response and power transmission oscillation; 6.3 Oscillation Source Determination: like , ,and , This indicates that the wind power side is continuously injecting energy, and the wind power oscillation response leads the thermal power and interconnection line transmission fluctuations, thus the oscillation source is determined to be the wind power side; like , ,and , This indicates that energy is continuously being injected into the thermal power plant, and the thermal power plant's oscillation response leads the fluctuations in wind power and transmission lines, thus identifying the oscillation source as the thermal power plant. like , ,and , This indicates that the energies on both sides of the wind and fire cancel each other out, and there is inter-system energy coupling oscillation in the transmission link of the connecting line. The source of the oscillation is determined to be the coupling of the connecting line.

[0013] A low-frequency oscillation source positioning device for a wind-fire bundling system includes: Multi-source signal acquisition unit of wind and fire baling system: Relying on a broadband measurement device with a sampling rate of not less than 1kHz, it synchronously acquires the original signals of the wind and fire baling system: wind power side signals, thermal power side signals, tie line signals and system parameters, to ensure coverage of the entire oscillation cycle; Adaptive wavelet threshold denoising preprocessing unit: Performs adaptive wavelet threshold denoising processing on the acquired raw signals to obtain denoised signals. ; Adaptive noise intensity set empirical mode decomposition unit: based on the obtained denoised signal An improved adaptive noise intensity set empirical mode decomposition algorithm is used to perform mode decomposition, extract the intrinsic mode functions related to low-frequency oscillations in power transmission, and obtain candidate IMF components. Wind-fire coupling key mode screening unit: Combining the inertial coupling characteristics of the wind-fire bundling system with the signal correlation characteristics of power transmission, a three-dimensional screening index of "energy proportion - time domain peak factor - inertial correlation degree" is designed to screen the key modes that best reflect the low-frequency oscillation characteristics of power transmission from the candidate IMF components. ; Wind- and fire-adaptive dissipative energy flow computation unit: based on the obtained key modes Combining the differences in dynamic characteristics between wind and thermal power units and the laws of power transmission, dissipation energy flow calculation models were constructed for the wind power side, thermal power side, and tie line, respectively. The phase coupling relationship between the dissipation energy flow trend and the key modes of wind and thermal power and power transmission signals was calculated. The phase-coupled oscillation source determination unit combines the calculated dissipated energy flow trend with the phase coupling relationship between the key wind and fire modes and the power transmission signal to construct a multi-condition fusion determination logic to achieve oscillation source type identification.

[0014] A further improvement of this invention lies in that, in the multi-source signal acquisition unit of the wind-fire baling system: relying on a broadband measurement device with a sampling rate of not less than 1kHz, the wind power side signal, thermal power side signal, tie line signal, and system parameters of the wind-fire baling system are simultaneously acquired, including: Wind power side signal: Active power at the wind farm grid connection point Doubly fed fan rotor speed Rotor-side converter q shaft current Rotor-side converter d shaft voltage ; Thermal power side signal: Active power of thermal power units Thermal power unit speed Steam valve opening of thermal power unit Thermal power unit terminal voltage deviation ; Tie line signal: Active power of the tie line between the wind-fire bundling system and the main grid , tie line voltage amplitude System parameters: Rated inertia of thermal power unit J g Equivalent inertia of wind turbine units J w Connection line impedance .

[0015] A further improvement of this invention lies in that, in the adaptive wavelet threshold denoising preprocessing unit: adaptive wavelet threshold denoising processing is performed on the acquired multi-source signals respectively, including: 2.1 Wavelet decomposition: Choosing the db6 wavelet as the basis wavelet for the original signal Perform three-level wavelet decomposition to obtain the wavelet coefficients of each level. ,in j The number of decomposition layers, k This is the index of the wavelet coefficients for this layer; 2.2 Calculation of Noise and Signal Standard Deviation: Based on the first-level detail coefficients of wavelet decomposition, the noise standard deviation is calculated using a formula. : (1) In formula (1): median( () indicates the median operation. The detail coefficients of the first layer of wavelet decomposition are used; the original signal is calculated using the formula. Standard deviation : (2) In formula (2): N The number of signal sampling points. For the original signal i The values ​​of each sampling point; 2.3 Adaptive Threshold Design: Considering the difference in wavelet coefficient amplitude between the oscillating signal and the power transmission interference noise, a threshold that dynamically adjusts with the wavelet coefficient amplitude is designed, as shown in the following formula: (3) In formula (3): For the first j Layer k The absolute value of each wavelet coefficient is used to design a smaller threshold for oscillatory characteristic coefficients with larger amplitudes and a larger threshold for noise coefficients with smaller amplitudes. 2.4 Soft Thresholding and Signal Reconstruction: The wavelet coefficients are subjected to soft thresholding, as shown in the following formula: (4) In equation (4): sign( ) is a sign function, max( (,0) indicates taking the larger value between the input value and 0; the processed wavelet coefficients Wavelet reconstruction is performed to obtain the denoised signal. This serves as the input for subsequent mode decomposition.

[0016] Compared with the prior art, the present invention has at least the following beneficial technical effects: 1. Strong scenario adaptability: For the power transmission characteristics of the wind and fire bundled system, a multi-source signal acquisition scheme (including transmission parameters such as tie line current) and a wind and fire adapted dissipative energy flow model (introducing transmission power normalization correction) are designed to solve the adaptability problem of the traditional one-size-fits-all method and accurately match the low-frequency oscillation positioning requirements of power transmission from 0.1 to 2Hz. 2. Excellent noise resistance: Through adaptive wavelet threshold denoising (dynamic threshold suppression of power transmission interference noise) and AN-EEMD decomposition (decreasing noise intensity design), the key mode extraction accuracy is ≥95% in a strong interference noise environment of 30 to 50 dB, which is significantly better than traditional EEMD (≤80%) and EMD (≤75%). 3. High positioning accuracy: The system introduces a three-dimensional screening index (including transmission power correlation) of "energy ratio - time domain peak factor - inertia correlation" and dual phase difference judgment logic to avoid misjudgment of single energy trend. The positioning accuracy is ≥98% in the 0.1 to 2Hz frequency band, and can clearly distinguish three types of oscillation sources: wind power side, thermal power side, and tie line. 4. Good real-time performance: The number of ensemble operations for AN-EEMD is optimized to 20 (compared to more than 100 for traditional EEMD). The calculation of each step is based on real-time transmission data from broadband measurements, and the single-machine calculation time is ≤50ms, which meets the real-time requirements of power system power transmission oscillation monitoring. Attached Figure Description

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

[0018] Figure 1 A flowchart of a method for locating a low-frequency oscillation source in a wind-fire bundling system.

[0019] Figure 2 This is a structural block diagram of a low-frequency oscillation source positioning device for a wind-fire bundling system. Detailed Implementation

[0020] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0021] In the description of this invention, it should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0022] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0023] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0024] The accompanying drawings illustrate various structural schematic diagrams according to embodiments disclosed in this invention. These drawings are not to scale, and some details have been enlarged for clarity, and some details may have been omitted. The shapes of the various regions and layers shown in the drawings, as well as their relative sizes and positional relationships, are merely exemplary and may deviate from reality due to manufacturing tolerances or technical limitations. Furthermore, those skilled in the art can design regions / layers with different shapes, sizes, and relative positions as needed.

[0025] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0026] Example 1 like Figure 1 As shown, the present invention provides a method for locating a low-frequency oscillation source in a wind-fire bundling system, comprising: This invention achieves low-frequency oscillation source localization in a wind-fire bundling system through six closely linked steps. Each step uses the output of the preceding step as input, forming a closed-loop collaborative localization process, as detailed below: Step 1: Multi-source signal acquisition of the wind and fire bundling system Relying on a broadband measurement device with a sampling rate of no less than 1kHz, the wind power side signal, thermal power side signal, tie line signal, and system parameters of the wind-thermal bundling system are collected simultaneously to ensure coverage of the entire oscillation cycle (oscillation start time). Until the oscillation ends This provides complete data support for subsequent analysis: Wind power side signal: Active power at the wind farm grid connection point (Reflecting fluctuations in wind power output), doubly-fed induction generator rotor speed (Demonstrating the dynamic response of the wind turbine), rotor-side converter q shaft current (Associated fan torque control), rotor-side converter d shaft voltage (Reflects the excitation status of the wind turbine); Thermal power side signal: Active power of thermal power units (Reflecting the power regulation capability of thermal power units), thermal power unit speed (Reflecting the unit's inertial response), steam valve opening of thermal power units (Related to thermal power generation power regulation delay), thermal power unit terminal voltage deviation (Reflects the voltage stability of the generating unit); Tie line signal: Active power of the tie line between the wind-fire bundling system and the main grid (Reflecting power interaction between systems), tie line voltage amplitude (Reflects the operating status of the tie line); System parameters: Rated inertia of the thermal power unit J g (Obtained from the unit nameplate parameters), equivalent inertia of the wind turbine unit J w (Calculated by multiplying the number of wind turbines in the wind farm by the inertia of each turbine), tie line impedance (Refer to the power grid parameter manual).

[0027] Step 2: Adaptive wavelet threshold noise reduction preprocessing To address the strong background noise generated by grid load fluctuations and equipment switching during power transmission, adaptive wavelet threshold denoising is performed on the multi-source signals acquired in step 1. This process eliminates noise interference while preserving low-frequency oscillation characteristics. The specific steps are as follows: 2.1 Wavelet decomposition: The db6 wavelet (which has excellent time-frequency focusing properties in non-stationary signal processing and is suitable for the oscillating signal characteristics of power transmission) was selected as the basis wavelet for processing the original signal. Perform three-level wavelet decomposition to obtain the wavelet coefficients of each level. ,in j The decomposition level (values ​​1, 2, or 3). k This is the index of the wavelet coefficients for this layer.

[0028] 2.2 Calculation of Noise and Signal Standard Deviation: Based on the first-level detail coefficients of wavelet decomposition (mainly dominated by power transmission interference noise), the noise standard deviation is calculated using a formula. : (1) In formula (1): median( () indicates the median operation. The detail coefficients of the first layer of wavelet decomposition are used; the original signal is calculated using the formula. Standard deviation (Reflecting the fluctuation intensity of the power transmission signal): (2) In formula (2): N This is the number of signal sampling points (obtained by multiplying the sampling rate of the broadband measurement device by the acquisition time). For the original signal i The values ​​of each sampling point.

[0029] 2.3 Adaptive Threshold Design: Considering the difference in wavelet coefficient amplitude between the oscillating signal and the power transmission interference noise, a threshold that dynamically adjusts with the wavelet coefficient amplitude is designed, as shown in the following formula: (3) In formula (3): For the first j Layer k The absolute value of each wavelet coefficient is used to design a smaller threshold for oscillation characteristic coefficients with larger amplitudes (preserving oscillation information) and a larger threshold for noise coefficients with smaller amplitudes (suppressing power transmission interference).

[0030] 2.4 Soft Thresholding and Signal Reconstruction: The wavelet coefficients are subjected to soft thresholding, as shown in the following formula: (4) In equation (4): sign( ) is the sign function (output 1 for positive input, -1 for negative input, and 0 for negative input), max( (,0) indicates taking the larger value between the input value and 0; the processed wavelet coefficients Wavelet reconstruction is performed to obtain the denoised signal. This serves as the input for subsequent mode decomposition.

[0031] Step 3: Adaptive noise intensity set empirical mode decomposition Based on the denoised signal obtained in step 2 An improved Adaptive Noise Intensity Set Empirical Mode Decomposition (AN-EEMD) algorithm is used for mode decomposition to extract intrinsic mode functions (IMFs) related to low-frequency oscillations in power transmission, avoiding mode aliasing caused by fixed noise intensity in traditional EEMD. The specific process is as follows: 3.1 Adaptive Noise Intensity Design: Considering the fluctuation characteristics of the power transmission signal in the wind-fire bundled power transmission, the intensity of the added white noise is designed to decrease with the number of decompositions (Ensemble count). The Ensemble count is set to 20 (balancing decomposition accuracy and computational efficiency). i White noise added during the decomposition strength The calculation formula is as follows: (5) In formula (5): For denoised signals The maximum absolute value, i The number of decompositions (ranging from 1 to 20) determines the noise intensity. i The increment linearly decreases from 0.1 times the maximum absolute value of the signal to 0.01 times, achieving the breaking of mode aliasing in the early stage of strong noise and the optimization of decomposition accuracy in the later stage of weak noise, adapting to the complex characteristics of oscillation signals in power transmission.

[0032] 3.2 Multi-round EMD decomposition: For the first i Noisy signal after adding noise The EMD decomposition process includes: identifying local maxima and minima of the signal (reflecting the peaks and valleys of power transmission oscillations); fitting the upper and lower envelopes using cubic spline interpolation; calculating the mean curve of the envelopes; subtracting the mean curve from the noisy signal to obtain the IMF component; repeating the above operations until the remaining components are monotonic or constant functions, thus obtaining the first EMD component. i The IMF components of the secondary decomposition ( m (The number of IMFs in this decomposition) and residual components .

[0033] 3.3 Calculation of Average IMF Components and Selection of Candidate Modes: The arithmetic mean of the IMF components of the same order obtained from 20 decompositions is used to obtain the final IMF component, as shown in the following formula: (6) In equation (6), For the first k The final IMF component of the order, For the i-th decomposition, the first... k The first-order IMF components were analyzed using Fast Fourier Transform (FFT) to identify IMF components with frequencies ranging from 0.1 to 2 Hz (corresponding to the low-frequency oscillation band of power transmission). These were designated as candidate IMF components. ( p (Number of candidate components).

[0034] Step 4: Screening of Key Modes for Wind-Fire Coupling Combining the inertial coupling characteristics of the wind-fire bundling system with the signal correlation features of power transmission, a three-dimensional screening index of "energy proportion - time domain peak factor - inertial correlation degree" is designed. From the candidate IMF components in step 3, the key modes that best reflect the low-frequency oscillation characteristics of power transmission are selected. The specific process is as follows: 4.1 Energy percentage calculation: The energy percentage reflects the proportion of oscillation energy contained in the candidate IMF component within the total energy of all candidate components. A higher percentage indicates a greater contribution to the oscillation of power transmission. The formula is as follows: (7) In equation (7): For the first k One candidate IMF component t 0 represents the start time of oscillation. t 1 represents the termination time of the oscillation, and the numerator is the 1st oscillation. k The energy of each candidate component (the square integral of the signal, reflecting the intensity of the oscillation energy) is divided into the denominator and the total energy of all candidate components.

[0035] 4.2 Calculation of time-domain peak factor: The time-domain maxima factor reflects the amplitude stability of candidate IMF components. The smaller the maxima factor, the more stable the component amplitude fluctuations, and the more likely it is to be a true oscillation mode of the power transmission process (rather than a sudden disturbance). The formula is as follows: (8) In equation (8): For the first k The maximum absolute value of each candidate component is given by the denominator, which is the root mean square value of that component (the square root of the energy average, reflecting the stability of the oscillation amplitude).

[0036] 4.3 Inertia Correlation Calculation: The inertia correlation degree combines the inertia distribution of the wind and thermal power units with the correlation of power transmission signals, reflecting the degree of correlation between candidate components and the dynamic response of the wind and thermal power units and the power transmission status. The formula is as follows: (9) In equation (9): The rated inertia of the thermal power unit. This is the equivalent inertia of the wind turbine. The weight of the thermal power correlation coefficient (corresponding to the impact of thermal power on power transmission oscillation). The weight of the wind power correlation coefficient (corresponding to the impact of wind power on power transmission oscillations) satisfies the following conditions: , The calculation formula is as follows: (10) In equation (10): corr( , ) is the function for calculating the Pearson correlation coefficient. The correlation coefficient between the candidate components and the rotational speed of the thermal power unit is given. The correlation coefficient between the candidate components and the power transmitted by the tie line. The correlation coefficient between the candidate components and the fan rotor speed is given. is the correlation coefficient between the candidate component and the wind power transmission capacity.

[0037] 4.4 Key Modality Screening: Set an energy percentage threshold of 0.1 (to ensure the component contains sufficient oscillating energy), a time-domain peak factor threshold of 5 (to ensure component amplitude stability), and an inertia correlation threshold of 0.08 (to ensure the component is strongly correlated with wind, fire, and the transport process). Screening should simultaneously meet these criteria. , and The candidate components were identified as the key oscillation modes. If multiple components satisfy the conditions, choose the inertia correlation degree. The largest component; if no component simultaneously satisfies all three conditions, choose... and The components are used as key modes.

[0038] Step 5: Calculation of wind-fire adapted dissipative energy flow Based on the key modes obtained in step 4 By combining the differences in dynamic characteristics between wind and thermal power units (thermal power valve regulation delay, wind power converter control response) and the laws of power transmission (tether line impedance loss, power interaction), dissipative energy flow calculation models are constructed for the wind power side, thermal power side, and tie line to improve the accuracy of energy calculation. The specific process is as follows: 5.1 Calculation of dissipated energy flow on the thermal power plant side: Considering the transient impact of steam valve regulation on power transmission in thermal power units, steam valve opening is introduced. The term related to tie-line power is used as a correction term, and the formula is as follows: (11) In equation (11): For the dissipated energy flow on the thermal power side, The key modes are obtained by decomposing the active power signal of thermal power plants. For the rotational speed of the thermal power unit, This refers to the reactive power of thermal power units. This refers to the voltage deviation at the thermal power unit terminals. To transmit power to the tie line, The maximum power of the tie line is normalized. The first term of the integral term reflects the energy exchange between the active power and the rotational speed of the thermal power plant. The second term reflects the energy loss caused by the combined effects of reactive power, voltage deviation, valve regulation, and transmission power. The integral of the difference is the total dissipated energy flow on the thermal power plant side.

[0039] 5.2 Calculation of energy dissipation flow on the wind power side: Considering the impact of rotor-side converter control on power transmission in doubly-fed wind turbines, the rotor-side converter control is introduced... q shaft current and d shaft voltage The product term, combined with the deviation between the rotor speed and the synchronous speed, adjusts the energy flow direction, as shown in the following formula: (12) In equation (12): This refers to the energy dissipation flow on the wind power side. The key modes are obtained by decomposing the active power signal of wind power. This refers to the rotor speed of the doubly fed fan. Synchronous rotation speed (3000 revolutions per minute for a 50Hz power grid, 1500 revolutions per minute for a 60Hz power grid); For rotor-side converter q shaft current, For rotor-side converter d shaft voltage; sign( ) is a symbolic function; To transmit power to wind power; The maximum wind power is represented by the normalized value. The first term of the integral term reflects the energy interaction between the active power of the wind power and the rotor speed. The second term reflects the energy loss due to the effect of the converter current and voltage and the proportion of wind power transmission. The sign function adjusts the sign of the loss term according to the speed deviation.

[0040] 5.3 Calculation of energy dissipation flow in tie lines: By directly incorporating the impedance loss characteristics of the power transmission process, the real part of the tie line impedance is introduced to calculate the resistance loss, as shown in the following formula: (13) In equation (13), For the energy dissipation flow of the connection line, For the active power of the tie line, This refers to the amplitude of the tie line voltage. For the tie line impedance, This is the impedance magnitude. For tie line current, The first term of the integral term reflects the energy exchange between the power transmitted by the tie line and the voltage impedance, while the second term reflects the resistance loss of the current passing through the real part of the impedance (the core loss term of power transmission). The integral of the difference is the energy dissipation flow of the tie line.

[0041] Step 6: Determining the Oscillation Source Based on Phase Coupling Combining the dissipated energy flow trend calculated in step 5 with the phase coupling relationship between the key wind and fire modes and the power transmission signal, a multi-condition fusion judgment logic is constructed to achieve oscillation source type identification. The specific process is as follows: 6.1 Analysis of Dissipated Energy Flow Trends: Calculate the numerical value and rate of change of the dissipated energy flow on each side, if And the derivative of energy flow with respect to time This indicates that the thermal power plant is continuously injecting energy into the power transmission system (possibly as an oscillation source); if and This indicates that the wind power side is continuously injecting energy into the power transmission system (potentially an oscillation source); if and This indicates that there is an energy coupling surplus in the transmission link of the tie line (which may be a source of oscillation); among them, Numerical differentiation is used to calculate (the energy difference between adjacent sampling points divided by the time interval, reflecting the rate of energy change). 6.2 Calculation of phase coupling coefficient: The key modes of wind power were analyzed using Fast Fourier Transform. Key modes of thermal power and the transmission power of the connecting line Frequency analysis was performed to extract the phase information of the three components at the dominant oscillation frequency (key mode frequency), and the phase difference between the two sets was calculated. Phase difference between wind power and thermal power : (14) Wind power-tether phase difference : (15) In equations (14) and (15), FFT( ) is the Fast Fourier Transform function, arg( This is the operation of taking the complex argument (phase). , Both reflect the sequence of wind and fire response and power transmission oscillation; 6.3 Oscillation Source Determination: like , ,and , This indicates that the wind power side is continuously injecting energy, and the wind power oscillation response leads the thermal power and interconnection line transmission fluctuations, thus the oscillation source is determined to be the wind power side; like , ,and , This indicates that energy is continuously being injected into the thermal power plant, and the thermal power plant's oscillation response leads the fluctuations in wind power and transmission lines, thus identifying the oscillation source as the thermal power plant. like , ,and (The values ​​of wind and fire energy are close, but their signs are opposite.) This indicates that the energies on both sides of the wind and fire cancel each other out, and there is inter-system energy coupling oscillation in the transmission link of the connecting line. The source of the oscillation is determined to be the coupling of the connecting line.

[0042] Example 2 like Figure 1 As shown, the present invention provides a method for locating a low-frequency oscillation source in a wind-fire bundling system, comprising: 1. Multi-source signal acquisition: A 10s signal was acquired using a broadband phasor measurement device with a sampling rate of 2kHz: Wind power side: (0.9 to 1.2 GW) (1500 to 1520 rad / min) (0 to 1.2kA) (0.9 to 1.1 kV); Thermal power plant side: (0.8 to 1.0 GW) (3000 to 3005 rad / min) (0.8 to 0.95) (-0.05 to 0.03 kV); Contact line: (1.7 to 2.1 GW) (215 to 225 kV) (7.5 to 9.3 kA); System parameters: , , .

[0043] 2. Adaptive wavelet threshold noise reduction preprocessing: With connecting lines For example, after the db6 wavelet 3-level decomposition, the median of the detail coefficients of the first level is 0.04, which is obtained from equation (1). σ =0.04 / 0.6745≈0.059; From equation (2), we get σ x≈0.15 (N=20000); Calculate the adaptive threshold using equation (3). (0.07 to 0.14), perform soft thresholding according to equation (4); reconstruct the result. The noise level drops below 15 decibels, and the oscillation characteristics of power transmission are fully preserved.

[0044] 3. Adaptive noise intensity set empirical mode decomposition: right , (Active power noise reduction signal from thermal power plant) is decomposed using AN-EEMD with 20 ensemble iterations; the noise intensity is obtained from equation (5). ( GW) After 20 decompositions, Nine IMF components were obtained, and candidate components IMF4 (1.9Hz), IMF5 (1.2Hz), and IMF6 (0.6Hz) were selected.

[0045] 4. Screening of key modes for wind-fire coupling: The energy percentage of IMF5 is obtained from equation (7). From equation (8), we obtain the IMF5 peak factor. From equation (10), we get From equation (9), we get IMF5 (1.2Hz) was selected as the key mode. .

[0046] 5. Calculation of dissipated energy flow for wind-fire compatible systems: Thermal power side (Equation 11): Based on IMF5 and thermal power signal calculation, the energy absorbed by thermal power is -120kJ; based on IMF5 and wind power signal calculation, the energy injected by wind power is 385kJ. Tie line (Equation 13): Based on tie line impedance and transmission parameters calculation, the tie line energy surplus is 255kJ.

[0047] 6. Oscillation Source Determination Based on Phase Coupling From equation (14) we get The source of the oscillation was determined to be the wind power side, which is consistent with the scenario of oscillation caused by wind power fluctuations in actual power transmission.

[0048] Example 3 like Figure 2 As shown, the present invention provides a low-frequency oscillation source positioning device for a wind-fire bundling system, comprising: Multi-source signal acquisition unit of wind and fire baling system: Relying on a broadband measurement device with a sampling rate of not less than 1kHz, it synchronously acquires the original signals of the wind and fire baling system: wind power side signals, thermal power side signals, tie line signals and system parameters, to ensure coverage of the entire oscillation cycle; Adaptive wavelet threshold denoising preprocessing unit: Performs adaptive wavelet threshold denoising processing on the acquired raw signals to obtain denoised signals. ; Adaptive noise intensity set empirical mode decomposition unit: based on the obtained denoised signal An improved adaptive noise intensity set empirical mode decomposition algorithm is used to perform mode decomposition, extract the intrinsic mode functions related to low-frequency oscillations in power transmission, and obtain candidate IMF components. Wind-fire coupling key mode screening unit: Combining the inertial coupling characteristics of the wind-fire bundling system with the signal correlation characteristics of power transmission, a three-dimensional screening index of "energy proportion - time domain peak factor - inertial correlation degree" is designed to screen the key modes that best reflect the low-frequency oscillation characteristics of power transmission from the candidate IMF components. ; Wind- and fire-adaptive dissipative energy flow computation unit: based on the obtained key modes Combining the differences in dynamic characteristics between wind and thermal power units and the laws of power transmission, dissipation energy flow calculation models were constructed for the wind power side, thermal power side, and tie line, respectively. The phase coupling relationship between the dissipation energy flow trend and the key modes of wind and thermal power and power transmission signals was calculated. The phase-coupled oscillation source determination unit combines the calculated dissipated energy flow trend with the phase coupling relationship between the key wind and fire modes and the power transmission signal to construct a multi-condition fusion determination logic to achieve oscillation source type identification.

[0049] In the multi-source signal acquisition unit of the wind-fired power baling system in this embodiment: relying on a broadband measurement device with a sampling rate of not less than 1kHz, the wind power side signal, thermal power side signal, tie line signal, and system parameters of the wind-fired power baling system are acquired simultaneously, including: Wind power side signal: Active power at the wind farm grid connection point Doubly fed fan rotor speed Rotor-side converter q shaft current Rotor-side converter d shaft voltage ; Thermal power side signal: Active power of thermal power units Thermal power unit speed Steam valve opening of thermal power unit Thermal power unit terminal voltage deviation ; Tie line signal: Active power of the tie line between the wind-fire bundling system and the main grid , tie line voltage amplitude System parameters: Rated inertia of thermal power unit J g Equivalent inertia of wind turbine units Jw Connection line impedance .

[0050] In the adaptive wavelet threshold denoising preprocessing unit of this embodiment: adaptive wavelet threshold denoising processing is performed on the acquired multi-source signals, including: 2.1 Wavelet decomposition: Choosing the db6 wavelet as the basis wavelet for the original signal Perform three-level wavelet decomposition to obtain the wavelet coefficients of each level. ,in j The number of decomposition layers, k This is the index of the wavelet coefficients for this layer; 2.2 Calculation of Noise and Signal Standard Deviation: Based on the first-level detail coefficients of wavelet decomposition, the noise standard deviation is calculated using a formula. : (1) In formula (1): median( () indicates the median operation. The detail coefficients of the first layer of wavelet decomposition are used; the original signal is calculated using the formula. Standard deviation : (2) In formula (2): N The number of signal sampling points. For the original signal i The values ​​of each sampling point; 2.3 Adaptive Threshold Design: Considering the difference in wavelet coefficient amplitude between the oscillating signal and the power transmission interference noise, a threshold that dynamically adjusts with the wavelet coefficient amplitude is designed, as shown in the following formula: (3) In formula (3): For the first j Layer k The absolute value of each wavelet coefficient is used to design a smaller threshold for oscillatory characteristic coefficients with larger amplitudes and a larger threshold for noise coefficients with smaller amplitudes. 2.4 Soft Thresholding and Signal Reconstruction: The wavelet coefficients are subjected to soft thresholding, as shown in the following formula: (4) In equation (4): sign( ) is a sign function, max( (,0) indicates taking the larger value between the input value and 0; the processed wavelet coefficients Wavelet reconstruction is performed to obtain the denoised signal. This serves as the input for subsequent mode decomposition.

[0051] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the scope of the invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0052] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can be appropriately combined to form other embodiments that can be understood by those skilled in the art. The above content is only for illustrating the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made based on the technical concept proposed in this invention shall fall within the scope of protection of the claims of this invention.

Claims

1. A method for locating a low-frequency oscillation source in a wind-fire bundling system, characterized in that, include: Step 1: Multi-source signal acquisition of the wind and fire bundling system Relying on a wideband measurement device with a sampling rate of no less than 1kHz, the original signals of the wind-fired power bundling system are collected simultaneously: wind power side signals, thermal power side signals, tie line signals and system parameters, to ensure coverage of the entire oscillation cycle; Step 2: Adaptive wavelet threshold noise reduction preprocessing Adaptive wavelet threshold denoising was performed on the original signals acquired in step 1 to obtain denoised signals. ; Step 3: Adaptive noise intensity set empirical mode decomposition Based on the denoised signal obtained in step 2 An improved adaptive noise intensity set empirical mode decomposition algorithm is used to perform mode decomposition, extract the intrinsic mode functions related to low-frequency oscillations in power transmission, and obtain candidate IMF components. Step 4: Screening of Key Modes for Wind-Fire Coupling Combining the inertial coupling characteristics of the wind-fire bundling system with the signal correlation features of power transmission, a three-dimensional screening index of "energy proportion - time-domain peak factor - inertial correlation degree" is designed to select the key mode that best reflects the low-frequency oscillation characteristics of power transmission from the candidate IMF components in step 3. ; Step 5: Calculation of wind-fire adapted dissipative energy flow Based on the key modes obtained in step 4 Combining the differences in dynamic characteristics between wind and thermal power units and the laws of power transmission, dissipation energy flow calculation models were constructed for the wind power side, thermal power side, and tie line, respectively. The phase coupling relationship between the dissipation energy flow trend and the key modes of wind and thermal power and power transmission signals was calculated. Step 6: Determining the Oscillation Source Based on Phase Coupling By combining the dissipated energy flow trend calculated in step 5 with the phase coupling relationship between the key wind and fire modes and the power transmission signal, a multi-condition fusion judgment logic is constructed to achieve oscillation source type identification.

2. The method for locating a low-frequency oscillation source in a wind-fire bundling system according to claim 1, characterized in that, Step 1: Using a broadband measurement device with a sampling rate of not less than 1kHz, simultaneously acquire the wind power side signals, thermal power side signals, tie line signals, and system parameters of the wind-thermal bundling system, including: Wind power side signal: Active power at the wind farm grid connection point Doubly fed fan rotor speed Rotor-side converter q shaft current Rotor-side converter d shaft voltage ; Thermal power side signal: Active power of thermal power units Thermal power unit speed Steam valve opening of thermal power unit Thermal power unit terminal voltage deviation ; Tie line signal: Active power of the tie line between the wind-fire bundling system and the main grid , tie line voltage amplitude System parameters: Rated inertia of thermal power unit J g Equivalent inertia of wind turbine units J w Connection line impedance .

3. The method for locating a low-frequency oscillation source in a wind-fire bundling system according to claim 2, characterized in that, Step 2: Perform adaptive wavelet threshold denoising processing on the multi-source signals acquired in Step 1, including: 2.1 Wavelet decomposition: Choosing the db6 wavelet as the basis wavelet for the original signal Perform a three-level wavelet decomposition to obtain the wavelet coefficients of each level. ,in j The number of decomposition layers, k This is the index of the wavelet coefficients for this layer; 2.2 Calculation of Noise and Signal Standard Deviation: Based on the first-level detail coefficients of wavelet decomposition, the noise standard deviation is calculated using a formula. : （1） In formula (1): median( () indicates the median operation. The detail coefficients of the first layer of wavelet decomposition are used; the original signal is calculated using the formula. Standard deviation : （2） In formula (2): N The number of signal sampling points. For the original signal i The values ​​of each sampling point; 2.3 Adaptive Threshold Design: Considering the difference in wavelet coefficient amplitude between the oscillating signal and the power transmission interference noise, a threshold that dynamically adjusts with the wavelet coefficient amplitude is designed, as shown in the following formula: （3） In formula (3): For the first j Layer k The absolute value of each wavelet coefficient is used to design a smaller threshold for oscillatory characteristic coefficients with larger amplitudes and a larger threshold for noise coefficients with smaller amplitudes. 2.4 Soft Thresholding and Signal Reconstruction: The wavelet coefficients are subjected to soft thresholding, as shown in the following formula: （4） In equation (4): sign( ) is a sign function, max( (,0) indicates taking the larger value between the input value and 0; the processed wavelet coefficients Wavelet reconstruction is performed to obtain the denoised signal. This serves as the input for subsequent mode decomposition.

4. The method for locating a low-frequency oscillation source in a wind-fire bundling system according to claim 3, characterized in that, Step 3: Based on the denoised signal obtained in Step 2 An improved adaptive noise intensity set empirical mode decomposition algorithm is used to perform mode decomposition, extracting the intrinsic mode functions related to low-frequency oscillations in power transmission, including: 3.1 Adaptive Noise Intensity Design: Considering the fluctuation characteristics of the power transmission signal in the wind-fire bundled power transmission, the intensity of the added white noise is designed to decrease with the number of decompositions. The number of ensembles is set to 20. i White noise added during the decomposition strength The calculation formula is as follows: （5） In formula (5): For denoised signals The maximum absolute value, i The noise intensity varies with the number of decompositions. i The increment linearly decreases from 0.1 times the maximum absolute value of the signal to 0.01 times; 3.2 Multi-round EMD decomposition: For the i Noisy signal after adding noise Perform EMD decomposition to obtain the first i The IMF components of the secondary decomposition and residual components , m This represents the number of IMFs in this decomposition. 3.3 Calculation of Average IMF Components and Selection of Candidate Modes: The arithmetic mean of the IMF components of the same order obtained from 20 decompositions is used to obtain the final IMF component, as shown in the following formula: （6） In equation (6), For the first k The final IMF component of the order, For the i-th decomposition, the first... k The final IMF components are analyzed using Fast Fourier Transform (FFT) to select IMF components with frequencies ranging from 0.1 to 2 Hz as candidate IMF components. , p The number of candidate components.

5. The method for locating a low-frequency oscillation source in a wind-fire bundling system according to claim 4, characterized in that, Step 4: Combining the inertial coupling characteristics of the wind-fire bundling system with the signal correlation characteristics of power transmission, design a three-dimensional screening index of "energy proportion - time domain peak factor - inertial correlation degree" to select the key modes that best reflect the low-frequency oscillation characteristics of power transmission from the candidate IMF components in Step 3, including: 4.1 Energy percentage calculation: The energy percentage reflects the proportion of oscillation energy contained in the candidate IMF component within the total energy of all candidate components. A higher percentage indicates a greater contribution to the oscillation of power transmission. The formula is as follows: （7） In equation (7): For the first k Candidate IMF components t 0 represents the start time of the oscillation. t 1 represents the termination time of the oscillation, and the numerator is the 1st oscillation. k The energy of each candidate component is divided by the denominator, which is the total energy of all candidate components. 4.2 Calculation of time-domain peak factor: The time-domain maxima factor reflects the amplitude stability of candidate IMF components. The smaller the maxima factor, the more stable the amplitude fluctuation of the component, and the more likely it is to be the true oscillation mode of the power transmission process. The formula is as follows: （8） In equation (8): For the first k The maximum absolute value of each candidate component is given by the denominator, which is the root mean square value of that component. 4.3 Inertia Correlation Calculation: The inertia correlation degree combines the inertia distribution of the wind and thermal power units with the correlation of power transmission signals, reflecting the degree of correlation between candidate components and the dynamic response of the wind and thermal power units and the power transmission status. The formula is as follows: （9） In equation (9): The rated inertia of the thermal power unit. This is the equivalent inertia of the wind turbine. For the correlation coefficient weight of thermal power, The weights of the wind power correlation coefficients are given by both, and they satisfy the following conditions: , The calculation formula is as follows: （10） In equation (10): corr( , ) is the function for calculating the Pearson correlation coefficient. The correlation coefficient between the candidate components and the rotational speed of the thermal power unit is given. The correlation coefficient between the candidate components and the power transmitted by the tie line. The correlation coefficient between the candidate components and the fan rotor speed is given. The correlation coefficient between the candidate components and the wind power transmission capacity; 4.4 Key Modality Screening: Set an energy percentage threshold of 0.1, a time-domain peak factor threshold of 5, and an inertia correlation threshold of 0.08 to filter candidates that simultaneously meet these criteria. , and The candidate components were identified as the key oscillation modes. If multiple components satisfy the conditions, choose the inertia correlation degree. The largest component; if no component simultaneously satisfies all three conditions, choose... and The components are used as key modes.

6. The method for locating a low-frequency oscillation source in a wind-fire bundling system according to claim 5, characterized in that, Step 5: Based on the key modes obtained in Step 4 Based on the differences in dynamic characteristics between wind and thermal power units and the laws governing power transmission, dissipative energy flow calculation models are constructed for the wind power side, the thermal power side, and the tie line, including: 5.1 Calculation of dissipated energy flow on the thermal power plant side: Considering the transient impact of steam valve regulation on power transmission in thermal power units, steam valve opening is introduced. The term related to tie-line power is used as a correction term, and the formula is as follows: （11） In equation (11): For the dissipated energy flow on the thermal power side, The key modes are obtained by decomposing the active power signal of thermal power plants. This refers to the rotational speed of the thermal power unit. This refers to the reactive power of thermal power units. This refers to the voltage deviation at the thermal power unit terminals. To transmit power to the tie line, This represents the maximum power of the tie line; 5.2 Calculation of energy dissipation flow on the wind power side: Considering the impact of rotor-side converter control on power transmission in doubly-fed wind turbines, the rotor-side converter control is introduced... q shaft current and d shaft voltage The product term, combined with the deviation between the rotor speed and the synchronous speed, adjusts the energy flow direction, as shown in the following formula: （12） In equation (12): This refers to the energy dissipation flow on the wind power side. The key modes are obtained by decomposing the active power signal of wind power. This refers to the rotor speed of the doubly fed fan. Synchronous speed; For rotor-side converter q shaft current, For rotor-side converter d shaft voltage; sign( ) is a symbolic function; To transmit power to wind power; This represents the maximum wind power output. 5.3 Calculation of energy dissipation flow in tie lines: Based on the impedance loss characteristics of the power transmission process, the real part of the tie line impedance is introduced to calculate the resistance loss, as shown in the following formula: （13） In equation (13), For the energy dissipation flow of the connecting line, For the active power of the tie line, This refers to the amplitude of the tie line voltage. For the tie line impedance, This is the impedance magnitude. For tie line current, This represents the real part of the impedance.

7. The method for locating a low-frequency oscillation source in a wind-fire bundling system according to claim 6, characterized in that, Step 6: Combining the dissipated energy flow trend calculated in Step 5 with the phase coupling relationship between the key wind and fire modes and the power transmission signal, construct a multi-condition fusion judgment logic to achieve oscillation source type identification, including: 6.1 Analysis of Dissipated Energy Flow Trends: Calculate the numerical value and rate of change of the dissipated energy flow on each side, if And the derivative of energy flow with respect to time This indicates that the thermal power plant is continuously injecting energy into the power transmission system; if and This indicates that the wind power side is continuously injecting energy into the power transmission system; if and This indicates that there is an energy coupling surplus in the transmission link of the tie line; among which, Calculated through numerical differentiation; 6.2 Calculation of phase coupling coefficient: The key modes of wind power were analyzed using Fast Fourier Transform. Key modes of thermal power and the transmission power of the connecting line Frequency analysis was performed to extract the phase information of the three components at the dominant oscillation frequency, and the phase difference between the two sets was calculated. Wind power-thermal power phase difference : （14） Wind power-tether phase difference : （15） In equations (14) and (15), FFT( ) is the Fast Fourier Transform function, arg( () is the operation of taking the complex argument. , Both reflect the sequence of wind and fire response and power transmission oscillation; 6.3 Oscillation Source Determination: like , ,and , This indicates that the wind power side is continuously injecting energy, and the wind power oscillation response leads the thermal power and interconnection line transmission fluctuations, thus the oscillation source is determined to be the wind power side; like , ,and , This indicates that energy is continuously being injected into the thermal power plant, and the oscillation response of the thermal power plant leads the fluctuations in wind power and transmission lines, thus identifying the oscillation source as the thermal power plant. like , ,and , This indicates that the energies on both sides of the wind and fire cancel each other out, and there is inter-system energy coupling oscillation in the transmission link of the connecting line. The source of the oscillation is determined to be the coupling of the connecting line.

8. A low-frequency oscillation source positioning device for a wind-fire bundling system, characterized in that, include: Multi-source signal acquisition unit of wind and fire baling system: Relying on a broadband measurement device with a sampling rate of not less than 1kHz, it synchronously acquires the original signals of the wind and fire baling system: wind power side signals, thermal power side signals, tie line signals and system parameters, to ensure coverage of the entire oscillation cycle; Adaptive wavelet threshold denoising preprocessing unit: Performs adaptive wavelet threshold denoising processing on the acquired raw signals to obtain denoised signals. ; Adaptive noise intensity set empirical mode decomposition unit: based on the obtained denoised signal An improved adaptive noise intensity set empirical mode decomposition algorithm is used to perform mode decomposition, extract the intrinsic mode functions related to low-frequency oscillations in power transmission, and obtain candidate IMF components. Wind-fire coupling key mode screening unit: Combining the inertial coupling characteristics of the wind-fire bundling system with the signal correlation characteristics of power transmission, a three-dimensional screening index of "energy proportion - time domain peak factor - inertial correlation degree" is designed to screen the key modes that best reflect the low-frequency oscillation characteristics of power transmission from the candidate IMF components. ; Wind- and fire-adaptive dissipative energy flow computation unit: based on the obtained key modes Combining the differences in dynamic characteristics between wind and thermal power units and the laws of power transmission, dissipation energy flow calculation models were constructed for the wind power side, thermal power side, and tie line, respectively. The phase coupling relationship between the dissipation energy flow trend and the key modes of wind and thermal power and power transmission signals was calculated. The phase-coupled oscillation source determination unit combines the calculated dissipated energy flow trend with the phase coupling relationship between the key wind and fire modes and the power transmission signal to construct a multi-condition fusion determination logic to achieve oscillation source type identification.

9. A low-frequency oscillation source positioning device for a wind-fire bundling system according to claim 1, characterized in that, In the multi-source signal acquisition unit of the wind-fired power baling system: relying on a broadband measurement device with a sampling rate of not less than 1kHz, the wind power side signal, thermal power side signal, tie line signal, and system parameters of the wind-fired power baling system are acquired simultaneously, including: Wind power side signal: Active power at the wind farm grid connection point Doubly fed fan rotor speed Rotor-side converter q shaft current Rotor-side converter d shaft voltage ; Thermal power side signal: Active power of thermal power units Thermal power unit speed Steam valve opening of thermal power unit Thermal power unit terminal voltage deviation ; Tie line signal: Active power of the tie line between the wind-fire bundling system and the main grid , tie line voltage amplitude System parameters: Rated inertia of thermal power unit J g Equivalent inertia of wind turbine units J w Connection line impedance .

10. A low-frequency oscillation source positioning device for a wind-fire bundling system according to claim 9, characterized in that, In the adaptive wavelet threshold denoising preprocessing unit: adaptive wavelet threshold denoising processing is performed on the acquired multi-source signals, including: 2.1 Wavelet decomposition: Choosing the db6 wavelet as the basis wavelet for the original signal Perform a three-level wavelet decomposition to obtain the wavelet coefficients of each level. ,in j The number of decomposition layers, k This is the index of the wavelet coefficients for this layer; 2.2 Calculation of Noise and Signal Standard Deviation: Based on the first-level detail coefficients of wavelet decomposition, the noise standard deviation is calculated using a formula. : （1） In formula (1): median( () indicates the median operation. The detail coefficients of the first layer of wavelet decomposition are used; the original signal is calculated using the formula. Standard deviation : （2） In formula (2): N The number of signal sampling points. For the original signal i The values ​​of each sampling point; 2.3 Adaptive Threshold Design: Considering the difference in wavelet coefficient amplitude between the oscillating signal and the power transmission interference noise, a threshold that dynamically adjusts with the wavelet coefficient amplitude is designed, as shown in the following formula: （3） In formula (3): For the first j Layer k The absolute value of each wavelet coefficient is used to design a smaller threshold for oscillatory characteristic coefficients with larger amplitudes and a larger threshold for noise coefficients with smaller amplitudes. 2.4 Soft Thresholding and Signal Reconstruction: The wavelet coefficients are subjected to soft thresholding, as shown in the following formula: （4） In equation (4): sign( ) is a sign function, max( (,0) indicates taking the larger value between the input value and 0; the processed wavelet coefficients Wavelet reconstruction is performed to obtain the denoised signal. This serves as the input for subsequent mode decomposition.

Images