Multi-frequency oscillation identification methods, devices, electronic equipment and storage media

Abstract

This invention discloses a method, apparatus, electronic device, and storage medium for identifying multi-frequency oscillations, addressing the technical problems of poor high-frequency oscillation signal identification and mode aliasing in existing related technologies. The method includes: acquiring voltage signal data from a power system; performing preliminary variational mode decomposition (VMD) on the voltage signal data to obtain multiple frequency oscillation components, thereby avoiding mode aliasing in frequency oscillation analysis based on VMD; when system oscillation is determined based on multiple frequency oscillation components, performing secondary VMD on the voltage signal data by optimizing VMD parameters to obtain at least one dominant oscillation component; fitting the data based on the at least one dominant oscillation component, and calculating the oscillation characteristics of each dominant oscillation component based on the fitting results, thereby quickly and accurately identifying the oscillation characteristics of the oscillation signal by combining VMD parameter optimization and function fitting.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a method, apparatus, electronic device and storage medium for identifying multi-frequency oscillations. Background Technology

[0002] Flexible direct current (VSC-HVDC) transmission is widely used for large-capacity, long-distance power transmission and grid connection of new energy power plants due to its advantages such as low harmonic levels, absence of commutation failure, active and reactive power decoupling control, and ability to supply power to passive systems. With the commissioning of multiple flexible direct current systems in recent years, high-frequency resonance phenomena have frequently occurred in flexible direct current systems both domestically and internationally. Therefore, there is an urgent need to develop methods for detecting and identifying high-frequency oscillation signals.

[0003] Among existing methods for detecting high-frequency oscillation signals, the Fast Fourier Transform (FFT) with windowed interpolation is one of the most accurate algorithms for detecting steady-state signals. However, the FFT result is approximately an average result within a time window, which cannot effectively identify transient components and cannot obtain the time-frequency signal. The Hilbert-Huang Transform (HHT) is a time-frequency analysis method capable of detecting and identifying signals. However, HHT suffers from mode aliasing, which is insufficient for identifying high-frequency oscillation signals. Variational mode decomposition (VMD) can solve the mode aliasing problem to some extent. However, the VMD algorithm itself has several parameters that require further optimization, resulting in poor identification performance for high-frequency oscillation signals. Summary of the Invention

[0004] This invention provides a multi-frequency oscillation identification method, device, electronic device, and storage medium to solve or partially solve the technical problems of poor high-frequency oscillation signal identification and mode aliasing in existing related technologies.

[0005] This invention provides a multi-frequency oscillation identification method, the method comprising:

[0006] Collect voltage signal data from the power system;

[0007] The voltage signal data is subjected to preliminary variational mode decomposition to obtain multiple frequency oscillation components;

[0008] When the system is determined to be oscillating based on the multiple frequency oscillation components, the voltage signal data is subjected to secondary variational mode decomposition in combination with variational mode decomposition parameter optimization to obtain at least one dominant oscillation component.

[0009] The fitting is performed based on the at least one dominant oscillation component, and the oscillation characteristics of each dominant oscillation component are calculated based on the fitting results.

[0010] Optionally, the step of combining variational mode decomposition parameter optimization to perform secondary variational mode decomposition on the voltage signal data to obtain at least one dominant oscillation component includes:

[0011] Step S01: Determine the minimum envelope entropy value of each frequency oscillation component by calculation;

[0012] Step S02: Using the minimum envelope entropy value as the initial fitness value, the particle swarm optimization algorithm is used to optimize the parameters of the variational mode decomposition to determine the optimal parameter combination;

[0013] Step S03: Calculate the autocorrelation coefficient of each frequency oscillation component, the autocorrelation coefficient being used to determine whether the frequency oscillation component is a noise mode component or a non-noise mode component;

[0014] Step S04: Using the number of non-noise mode components as the number of the second variational mode decomposition, and based on the optimal parameter combination, perform variational mode decomposition again on the voltage signal data;

[0015] Step S05: Determine whether there are still noise mode components among the multiple frequency oscillation components generated after re-decomposition. If yes, repeat steps S01 to S04; otherwise, proceed to step S05.

[0016] Step S06: Based on the last variational mode decomposition, obtain multiple optimized frequency oscillation components, and select at least one dominant oscillation component from the multiple optimized frequency oscillation components.

[0017] Optionally, determining the minimum envelope entropy value of each frequency oscillation component by calculation includes:

[0018] The frequency oscillation components are normalized, and the envelope value of each normalized frequency oscillation component is calculated. The corresponding envelope entropy value is then calculated.

[0019] The minimum envelope entropy value is taken as the minimum envelope entropy value of each frequency oscillation component.

[0020] Optionally, the step of using a particle swarm optimization algorithm to optimize the parameters of the variational mode decomposition and determine the optimal parameter combination includes:

[0021] Bandwidth constraints and noise tolerance are used as the combination of parameters to be optimized in variational mode decomposition. Based on the initial fitness value, the combination of parameters to be optimized is optimized using the particle swarm optimization algorithm.

[0022] During the parameter optimization process, the combination of parameters to be optimized is limited to a pre-set range of parameter combination selections;

[0023] Repeat the iteration until the iteration reaches the convergence condition or the maximum number of iterations is reached, then stop the iteration and output the optimal parameter combination.

[0024] Optionally, selecting at least one dominant oscillation component from the plurality of optimized frequency oscillation components includes:

[0025] The frequency spectrum of the multiple optimized frequency oscillation components is obtained based on the Hilbert-Huang transform.

[0026] The amplitude of each optimized frequency oscillation component in the frequency spectrum is compared with a preset signal amplitude threshold, and the optimized frequency oscillation component that is greater than or equal to the preset signal amplitude threshold is taken as the dominant oscillation component.

[0027] Optionally, the preliminary variational mode decomposition of the voltage signal data to obtain multiple frequency oscillation components includes:

[0028] Based on a preset mode decomposition number, the voltage signal data is subjected to preliminary variational mode decomposition to obtain multiple frequency oscillation components, the number of which corresponds to the preset mode decomposition number.

[0029] Optionally, determining system oscillation based on the plurality of frequency oscillation components includes:

[0030] Each frequency oscillation component is compared with a preset signal amplitude threshold. When there is a frequency oscillation component that is greater than the preset signal amplitude threshold, it is determined that the power system is oscillating.

[0031] The present invention also provides a multi-frequency oscillation identification device, comprising:

[0032] The data acquisition module is used to acquire voltage signal data from the power system.

[0033] The preliminary variational mode decomposition module is used to perform preliminary variational mode decomposition on the voltage signal data to obtain multiple frequency oscillation components;

[0034] The second variational mode decomposition module is used to perform second variational mode decomposition on the voltage signal data when the system oscillation is determined based on the multiple frequency oscillation components, combined with variational mode decomposition parameter optimization, to obtain at least one dominant oscillation component.

[0035] An oscillation feature calculation module is used to fit the at least one dominant oscillation component and calculate the oscillation features of each dominant oscillation component according to the fitting results.

[0036] The present invention also provides an electronic device, the device comprising a processor and a memory:

[0037] The memory is used to store program code and transmit the program code to the processor;

[0038] The processor is configured to execute the multi-frequency oscillation identification method as described above, according to the instructions in the program code.

[0039] The present invention also provides a computer-readable storage medium for storing program code for executing the multi-frequency oscillation identification method as described in any of the preceding claims.

[0040] As can be seen from the above technical solutions, the present invention has the following advantages:

[0041] A method for identifying multi-frequency oscillations is provided. First, voltage signal data from a power system is acquired. Preliminary variational mode decomposition (VMD) is performed on the voltage signal data to obtain multiple frequency oscillation components, thus avoiding mode aliasing in frequency oscillation analysis. When system oscillation is determined based on multiple frequency oscillation components, a second VMD is performed on the voltage signal data, combined with VMD parameter optimization, to obtain at least one dominant oscillation component. Finally, a fitting process is performed based on at least one dominant oscillation component, and the oscillation characteristics of each dominant oscillation component are calculated according to the fitting results. By combining VMD parameter optimization and function fitting, the oscillation characteristics of the oscillation signal can be quickly and accurately identified. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a flowchart of the steps involved in a multi-frequency oscillation identification method.

[0044] Figure 2 A flowchart of an algorithm for optimizing variational mode decomposition parameters;

[0045] Figure 3 This is a schematic diagram of the overall process of a multi-frequency oscillation identification method;

[0046] Figure 4 This is a topology diagram of a flexible DC converter transmitting power to an AC system.

[0047] Figure 5 This is a schematic diagram showing the instantaneous value of phase a voltage on the AC output side of a flexible DC converter during 3 to 4 seconds of system operation.

[0048] Figure 6 This is a schematic diagram illustrating the fitness changes during particle swarm optimization of variational mode decomposition.

[0049] Figure 7 This is a waveform diagram of each frequency oscillation component after variational mode decomposition;

[0050] Figure 8 The frequency spectrum of a frequency oscillation component after undergoing a Hilbert-Huang transform based on variational mode decomposition;

[0051] Figure 9 This is a structural block diagram of a multi-frequency oscillation identification device. Detailed Implementation

[0052] This invention provides a multi-frequency oscillation identification method, apparatus, electronic device, and storage medium to solve or partially solve the technical problems of poor high-frequency oscillation signal identification and mode aliasing in existing related technologies.

[0053] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0054] As an example, among existing methods for detecting high-frequency oscillation signals, the Fast Fourier Transform (FFT) with windowed interpolation is one of the most accurate algorithms for steady-state signal detection. However, the FFT result is approximately an average result within a time window, which cannot effectively identify transient components and cannot obtain the time-frequency signal. The Hilbert-Huang Transform (HHT) is a time-frequency analysis method capable of detecting and identifying signals. However, HHT suffers from mode aliasing, which is insufficient for identifying high-frequency oscillation signals, especially when the high-frequency oscillation signal in a power system may have two adjacent frequency oscillation components, significantly reducing its identification effectiveness. Variational Mode Decomposition (VMD) can address mode aliasing to some extent. However, the VMD algorithm itself has several parameters that require further optimization, resulting in poor identification performance for high-frequency oscillation signals.

[0055] Therefore, one of the core inventive points of this invention is: addressing the deficiencies and shortcomings of existing technologies, it proposes a multi-frequency oscillation detection and identification method suitable for high-frequency bands. First, a mode decomposition number K is selected to perform variational mode decomposition on the acquired power system signal, splitting the signal into multiple oscillation frequency components (or signal components) at various frequencies. The method described in this invention (IMF) determines whether the system oscillates. For high-frequency oscillation components with two adjacent frequencies in the power system oscillation signal, the scheme provided by this invention avoids mode aliasing during mode decomposition compared to other methods. When the system oscillates, the bandwidth constraint and noise tolerance parameters in the variational mode decomposition are optimized using the particle swarm optimization algorithm. The value of the mode decomposition number K is redefined using the autocorrelation coefficient, thereby determining the three parameters of the variational mode decomposition method by combining the autocorrelation coefficient and the particle swarm optimization algorithm. Then, a second variational mode decomposition is performed based on the redefined parameters to obtain oscillation frequency components less affected by noise interference. Finally, Hilbert-Huang transform and fitting are performed on each oscillation frequency component obtained from the signal decomposition to further identify oscillation characteristic information such as frequency, amplitude, and damping ratio. Thus, based on the time-frequency domain information of each oscillation frequency component, transient information such as oscillation frequency, oscillation amplitude, and damping ratio in the power system oscillation signal can be identified, effectively solving the problem of high-frequency oscillation signal detection and identification in flexible DC transmission systems, and has strong adaptability.

[0056] Reference Figure 1 The diagram illustrates a flowchart of a multi-frequency oscillation identification method provided by an embodiment of the present invention, which may specifically include the following steps:

[0057] Step 101: Collect voltage signal data of the power system;

[0058] The first step is the preliminary variational mode decomposition and oscillation discrimination process of the signal.

[0059] In this embodiment of the invention, the sampling frequency is set to 25.6 kHz. The voltage value (voltage signal) at the AC output side of the flexible DC converter is sampled. The sampling method for the voltage signal data can be set by those skilled in the art based on actual conditions; for example, an interval sampling method can be used to collect the voltage signal, such as sampling the voltage value every 1 second or 200 ms.

[0060] When it is determined that the flexible DC transmission system is oscillating, the voltage signal data collected in this step will be used as the power system signal to analyze the power system oscillation.

[0061] Step 102: Perform preliminary variational mode decomposition on the voltage signal data to obtain multiple frequency oscillation components;

[0062] When performing oscillation analysis, a preliminary mode decomposition number K for variational mode decomposition can be given first. Then, the variational mode decomposition method is used to perform preliminary variational mode decomposition on the acquired voltage signal data.

[0063] Specifically, performing preliminary variational mode decomposition on the voltage signal data to obtain multiple frequency oscillation components can be achieved by: performing preliminary variational mode decomposition on the voltage signal data based on a preset mode decomposition number K to obtain multiple frequency oscillation components, wherein the number of frequency oscillation components corresponds to the preset mode decomposition number. That is, if the preset mode decomposition number is set to K, the number of frequency oscillation components obtained after preliminary variational mode decomposition is also K.

[0064] More specifically, a mode decomposition number K is first set. Variational mode decomposition (VMD) can decompose the acquired voltage signal into K frequency oscillation components (IMFs) based on the preset number K. Since this step mainly performs preliminary decomposition of the oscillating signal, the mode decomposition number K can be initially set to a relatively large value. In subsequent optimization steps, the mode decomposition number K is re-determined based on relevant criteria. In this embodiment, the preliminary mode decomposition number K is set to 10. Then, mode decomposition is performed on the acquired voltage signal to obtain 10 frequency oscillation components.

[0065] Since high-frequency oscillations in power systems frequently occur in the interconnection systems between flexible DC converters and AC power grids, and between flexible DC converters and new energy power plants, the high-frequency oscillation signals collected by the system may simultaneously contain oscillations at multiple frequencies. In this embodiment of the invention, variational mode decomposition is used to decompose the collected signals, which can avoid mode aliasing caused by two oscillation signal components having similar frequencies.

[0066] Step 103: When the system is determined to be oscillating based on the multiple frequency oscillation components, the voltage signal data is subjected to secondary variational mode decomposition in combination with variational mode decomposition parameter optimization to obtain at least one dominant oscillation component.

[0067] Based on the multiple frequency oscillation components obtained through preliminary variational mode decomposition, it is determined whether the system is oscillating. A threshold for signal amplitude can be preset. If any frequency oscillation component (except the fundamental frequency component) exceeds the preset signal amplitude threshold, the system is determined to be oscillating, and subsequent oscillation analysis steps are executed; otherwise, the oscillation detection procedure ends.

[0068] In a specific implementation, determining system oscillation based on multiple frequency oscillation components can be achieved by comparing each frequency oscillation component with a preset signal amplitude threshold. When there is a frequency oscillation component with an amplitude greater than the preset signal amplitude threshold, it is determined that the power system is oscillating.

[0069] The signal amplitude threshold can be set based on operational experience or in combination with actual detection conditions; this invention does not impose any restrictions on this.

[0070] When the system oscillates, the parameters of the variational mode decomposition method can be optimized to re-perform variational mode decomposition on the acquired voltage signal using the redefined relevant parameters, thereby obtaining frequency oscillation components that are less affected by noise and have higher accuracy.

[0071] See Figure 2 , Figure 2 The flowchart of an algorithm for optimizing variational mode decomposition parameters provided in an embodiment of the present invention is shown.

[0072] Combination Figure 2 In this step, the voltage signal data is subjected to secondary variational mode decomposition in combination with variational mode decomposition parameter optimization to obtain at least one dominant oscillation component. This process can be achieved by executing the following sub-steps S01 to S06:

[0073] Step S01: Determine the minimum envelope entropy value of each frequency oscillation component by calculation;

[0074] Furthermore, the minimum envelope entropy value of each frequency oscillation component is determined by calculation. Specifically, each frequency oscillation component is normalized, and the envelope value of each normalized frequency oscillation component is calculated. The corresponding envelope entropy value is then calculated. The minimum envelope entropy value (the smaller the envelope entropy value, the stronger the sparsity of the signal component and the more obvious the periodicity related to the oscillation) is taken as the minimum envelope entropy value of each frequency oscillation component.

[0075] Step S02: Using the minimum envelope entropy value as the initial fitness value, the particle swarm optimization algorithm is used to optimize the parameters of the variational mode decomposition and determine the optimal parameter combination;

[0076] The minimum envelope entropy value among all frequency oscillation components is used as the initial fitness value for the Particle Swarm Optimization (PSO) algorithm employed in subsequent parameter optimization. The PSO algorithm is then used to optimize the bandwidth constraint and noise tolerance parameters of the variational mode decomposition method.

[0077] In practice, increasing bandwidth constraints may cause mode aliasing, while decreasing bandwidth constraints may lead to an increase in noise and spurious modes. Increasing noise tolerance can reduce the occurrence of spurious modes, but excessively high noise tolerance can cause true modes to be ignored. Therefore, in this embodiment of the invention, based on operational experience with variational mode decomposition, selectable ranges for two parameters are pre-set. This allows the parameters to be optimized to be limited to a reasonable range during parameter optimization, avoiding parameters becoming too large or too small. Combining the parameter selection range, the particle swarm optimization algorithm can select the optimal combination of parameters in terms of bandwidth constraints and noise tolerance.

[0078] In the specific implementation, the particle swarm optimization algorithm is used to optimize the parameters of variational mode decomposition and determine the optimal parameter combination. This can be done by: taking the bandwidth constraint and noise tolerance as the parameter combination to be optimized in variational mode decomposition; optimizing the parameter combination based on the initial fitness value and combining it with the particle swarm optimization algorithm; restricting the parameter combination to be optimized to a pre-set parameter combination selection range during the parameter optimization process; repeating the iteration until the iteration reaches the convergence condition or the number of iterations reaches the maximum number, stopping the iteration, and outputting the optimal parameter combination.

[0079] Step S03: Calculate the autocorrelation coefficient of each frequency oscillation component. The autocorrelation coefficient is used to determine whether the frequency oscillation component is a noise mode component or a non-noise mode component.

[0080] Then, the autocorrelation coefficient can be used to determine the noise mode components in each frequency oscillation component.

[0081] First, the autocorrelation coefficient of each frequency oscillation component is calculated. Since noise occurs randomly at any given time, its autocorrelation coefficient is small, while the true frequency oscillation component (non-noise mode component) oscillates periodically according to its center frequency, resulting in a larger autocorrelation coefficient. Therefore, in this embodiment of the invention, the autocorrelation coefficient is used to distinguish between noise mode components and non-noise mode components among multiple frequency oscillation components.

[0082] Step S04: Using the number of non-noise mode components as the number of the second variational mode decomposition, and based on the optimal parameter combination, perform variational mode decomposition on the voltage signal data again;

[0083] The number of modes in variational mode decomposition is set to the number of non-noise mode components. At the same time, the bandwidth constraint and noise tolerance parameters of variational mode decomposition are adjusted to the optimal parameter combination obtained in the above parameter optimization steps. Then, variational mode decomposition is performed again on the acquired voltage signal data.

[0084] Step S05: Determine whether there are still noise mode components among the multiple frequency oscillation components generated after re-decomposition. If yes, repeat steps S01 to S04; otherwise, proceed to step S05.

[0085] If noisy mode components still exist among the multiple frequency oscillation components generated after re-decomposition, the aforementioned optimization steps are performed again to update the mode decomposition number K.

[0086] Step S06: Based on the last variational mode decomposition, obtain multiple optimized frequency oscillation components, and select at least one dominant oscillation component from the multiple optimized frequency oscillation components.

[0087] In practice, it is the dominant oscillation component that causes oscillations in the system, while the non-dominant oscillation components have a negligible impact on system oscillations and can be disregarded relative to the dominant oscillation component; therefore, they do not require calculation or analysis. Based on this, after obtaining multiple optimized frequency oscillation components, this embodiment of the invention further filters out the dominant oscillation component (mode) of the system. In subsequent oscillation identification, only the filtered dominant oscillation component needs to be analyzed for oscillation characteristics. This reduces the computational load and improves computational efficiency without affecting the identification accuracy.

[0088] In a specific implementation, selecting at least one dominant oscillation component from multiple optimized frequency oscillation components can be achieved by: obtaining the frequency spectrum of multiple optimized frequency oscillation components according to the Hilbert-Huang transform; comparing the amplitude of each optimized frequency oscillation component (excluding the fundamental frequency component) in the frequency spectrum with a preset signal amplitude threshold, and taking the optimized frequency oscillation component that is greater than or equal to the preset signal amplitude threshold as the dominant oscillation component of the signal.

[0089] Step 104: Fit the at least one dominant oscillation component, and calculate the oscillation characteristics of each dominant oscillation component based on the fitting results.

[0090] Finally, the dominant oscillation component can be further fitted with a function to obtain the information characteristics of the oscillation signal. Specifically, after fitting the dominant oscillation component, the oscillation characteristics such as the oscillation frequency, amplitude, and damping ratio of each dominant oscillation component are calculated based on the fitting results, so as to realize the detection and information identification of high-frequency oscillations.

[0091] In this embodiment of the invention, a method for detecting and identifying multi-frequency oscillations in the high-frequency band is proposed. First, a mode decomposition number K is selected to perform variational mode decomposition on the acquired power system signal, splitting the signal into oscillation frequency components at multiple frequencies. It is then determined whether the system is oscillating. For power system oscillation signals containing two adjacent high-frequency oscillation components, the scheme provided in this invention, compared to other methods, can avoid mode aliasing during mode decomposition. When system oscillation occurs, the bandwidth constraint and noise tolerance parameters in the variational mode decomposition are optimized using a particle swarm optimization algorithm. The value of the mode decomposition number K is then re-determined using the autocorrelation coefficient, thereby combining the autocorrelation coefficient and the particle swarm optimization algorithm. The optimization algorithm determines three parameters of the variational mode decomposition method. Then, based on the redefined parameters, a second variational mode decomposition is performed to obtain oscillation frequency components less affected by noise interference. Finally, Hilbert-Huang transform and fitting are performed on each oscillation frequency component obtained from the signal decomposition to further identify oscillation characteristic information such as frequency, amplitude, and damping ratio. Thus, based on the time-frequency domain information of each oscillation frequency component, transient information such as oscillation frequency, oscillation amplitude, and damping ratio in the power system oscillation signal can be identified, effectively solving the problem of high-frequency oscillation signal detection and identification in flexible DC transmission systems, and exhibiting strong adaptability.

[0092] For better illustration, refer to Figure 3 This diagram illustrates the overall flow of a multi-frequency oscillation identification method provided by an embodiment of the present invention. It should be noted that this embodiment only provides a brief description of the general flow of multi-frequency oscillation identification; the specific implementation process of each step can be understood by referring to the relevant content in the foregoing embodiments, and will not be elaborated upon here. It is understood that the present invention does not impose any limitations on this.

[0093] Step 1: Collect voltage signal data from the power system, and perform preliminary variational mode decomposition on the voltage signal data based on the preset mode decomposition number K to obtain K frequency oscillation components;

[0094] Step 2: Based on the K frequency oscillation components, determine whether the system is currently oscillating. If yes, proceed to step 3; otherwise, end the process directly.

[0095] Step 3: Combine variational mode decomposition parameter optimization to perform secondary variational mode decomposition on the voltage signal data;

[0096] Step 4: Perform function fitting on the dominant oscillation component obtained from the last variational mode decomposition to obtain oscillation characteristic information such as oscillation frequency, amplitude, and damping ratio, and end the process.

[0097] For ease of understanding, the following description uses a specific example to illustrate an embodiment of the present invention.

[0098] First, a simulation system is built in the power system electromagnetic transient simulation software PSCAD / EMTDC (Electro Magnetic Transient DC System). Figure 4 The topology diagram of the flexible DC converter transmitting power to the AC system shown is used for simulation verification.

[0099] The initial conditions and system parameters for the flexible DC system simulation are shown in Table 1. The AC system parameters are shown in Table 2. Detailed information on the IMFs of each frequency component of the high-frequency oscillation signal detected in this example is shown in Table 3.

[0100]

[0101] Table 1: Parameters of Flexible DC Transmission System

[0102]

[0103] Table 2: AC System Parameters

[0104]

[0105] Table 3: Detailed Information on IMF for Each Frequency Component

[0106] In this example, the following simulation scenario is set up:

[0107] The flexible DC converter operates under initial conditions and parameters. In the initial operating state of the AC system, switch S is open, with only the inductive branch (L1, R1) in operation. At 3.5 seconds, switch S of the AC system is closed, simultaneously energizing both the inductive and RC branches (C1, R1).

[0108] For example, Figure 5 This is a schematic diagram showing the instantaneous value of phase a voltage on the AC output side of the flexible DC converter during 3 to 4 seconds of system operation. Figure 6 This diagram illustrates the fitness changes during particle swarm optimization of variational mode decomposition. Figure 7 The waveforms of the oscillation components at each frequency after variational mode decomposition are shown. Figure 8 The frequency spectrum of the frequency oscillation component after undergoing the Hilbert-Huang transform based on variational mode decomposition is shown.

[0109] from Figure 5 The voltage waveform shows that the system is currently experiencing oscillations. Therefore, the identification method provided in this embodiment of the invention is first used to perform preliminary variational mode decomposition on the power system oscillation signal to obtain K frequency oscillation components (IMFs). Then, the envelope entropy value of each frequency oscillation component (IMF) is calculated using its envelope value, and the minimum envelope entropy value under the variational mode decomposition condition is determined.

[0110] Next, using the minimum envelope entropy value as the fitness function (initial fitness value), the optimal selection of the bandwidth constraint and noise tolerance parameters in variational mode decomposition is performed using the particle swarm optimization algorithm. In this example, after ten iterations, the bandwidth constraint and noise tolerance are finally determined to be 3000 and 0.01797, respectively. For details on the fitness changes during the particle swarm optimization process, please refer to [link to relevant documentation]. Figure 6 .

[0111] Then, the autocorrelation coefficient of each frequency oscillation component (IMF) is calculated. IMFs with an autocorrelation coefficient less than 0.2 are considered invalid mode components (i.e., noise mode components). Based on the number of non-noise mode components, the mode decomposition number K of the variational mode decomposition is reset. In this example, the final mode decomposition number is determined to be 4. Therefore, the acquired voltage signal can be re-decomposed into optimized frequency oscillation components (IMF') at 4 frequencies.

[0112] Finally, by combining Hilbert-Huang transform and function fitting, the amplitude of each optimized frequency oscillation component IMF' can be obtained. The amplitude is then compared with a pre-set signal amplitude threshold to screen out the dominant oscillation component. Furthermore, important characteristic information such as the oscillation frequency and damping ratio of the dominant oscillation component can be calculated.

[0113] Meanwhile, by using Fast Fourier Transform (FFT) to detect the voltage signal, it can be determined that there are two harmonics with large amplitudes at 1535Hz and 1635Hz, thus verifying the correctness of the method of the present invention. Furthermore, the method provided in this embodiment of the invention can detect transient components in the voltage signal more accurately than FFT and obtain important time-frequency domain information such as the damping ratio of the oscillation signal.

[0114] In summary, by adopting the method provided in the embodiments of the present invention, the mode aliasing phenomenon existing in the variational mode decomposition method can be overcome, the correct detection of multiple high-frequency oscillation components with similar frequencies can be achieved, and the amplitude and damping ratio of each frequency oscillation component can be quickly identified.

[0115] Reference Figure 9 The diagram illustrates a structural block diagram of a multi-frequency oscillation identification device provided in an embodiment of the present invention, which may specifically include:

[0116] Data acquisition module 901 is used to acquire voltage signal data of the power system;

[0117] The preliminary variational mode decomposition module 902 is used to perform preliminary variational mode decomposition on the voltage signal data to obtain multiple frequency oscillation components;

[0118] The second variational mode decomposition module 903 is used to perform second variational mode decomposition on the voltage signal data when the system oscillation is determined based on the multiple frequency oscillation components, and to obtain at least one dominant oscillation component by combining variational mode decomposition parameter optimization.

[0119] The oscillation feature calculation module 904 is used to fit based on the at least one dominant oscillation component, and calculate the oscillation features of each dominant oscillation component according to the fitting result.

[0120] In one optional embodiment, the second-order variational mode decomposition module 903 includes:

[0121] The minimum envelope entropy value determination module is used to perform step S01: determine the minimum envelope entropy value of each frequency oscillation component by calculation;

[0122] The parameter optimization module is used to execute step S02: using the minimum envelope entropy value as the initial fitness value, the particle swarm optimization algorithm is used to optimize the parameters of the variational mode decomposition to determine the optimal parameter combination;

[0123] The autocorrelation coefficient calculation module is used to perform step S03: calculate the autocorrelation coefficient of each frequency oscillation component, wherein the autocorrelation coefficient is used to determine whether the frequency oscillation component is a noise mode component or a non-noise mode component;

[0124] The second variational mode decomposition submodule is used to execute step S04: taking the number of non-noise mode components as the second variational mode decomposition number, and re-performing variational mode decomposition on the voltage signal data based on the optimal parameter combination;

[0125] The noise mode component determination module is used to execute step S05: determine whether there are still noise mode components among the multiple frequency oscillation components generated after re-decomposition. If yes, repeat steps S01 to S04; if no, jump to step S05.

[0126] The oscillation dominant component determination module is used to perform step S06: based on the last variational mode decomposition, obtain multiple optimized frequency oscillation components, and select at least one oscillation dominant component from the multiple optimized frequency oscillation components.

[0127] In one optional embodiment, the minimum envelope entropy value determination module is specifically used for:

[0128] The frequency oscillation components are normalized, and the envelope value of each normalized frequency oscillation component is calculated. The corresponding envelope entropy value is then calculated.

[0129] The minimum envelope entropy value is taken as the minimum envelope entropy value of each frequency oscillation component.

[0130] In one optional embodiment, the parameter optimization module includes:

[0131] The optimization module for the combination of parameters to be optimized is used to optimize the combination of parameters to be optimized by taking the bandwidth constraint and noise tolerance as the combination of parameters to be optimized by variational mode decomposition, and optimizing the parameters based on the initial fitness value and the particle swarm optimization algorithm.

[0132] The parameter combination constraint module is used to restrict the parameter combination to be optimized to a pre-set range of parameter combination selection during the parameter optimization process.

[0133] The optimal parameter combination output module is used for repeated iterations. When the iteration reaches the convergence condition or the number of iterations reaches the maximum number, the iteration stops and the optimal parameter combination is output.

[0134] In one alternative embodiment, the oscillation dominant component determination module includes:

[0135] A frequency spectrum generation module is used to obtain the frequency spectrum of the multiple optimized frequency oscillation components according to the Hilbert-Huang transform.

[0136] The oscillation dominant component determination submodule is used to compare the amplitude of each optimized frequency oscillation component in the frequency spectrum with a preset signal amplitude threshold, and to take the optimized frequency oscillation component that is greater than or equal to the preset signal amplitude threshold as the oscillation dominant component.

[0137] In one optional embodiment, the preliminary variational mode decomposition module 902 is specifically used for:

[0138] Based on a preset mode decomposition number, the voltage signal data is subjected to preliminary variational mode decomposition to obtain multiple frequency oscillation components, the number of which corresponds to the preset mode decomposition number.

[0139] In one optional embodiment, the second-order variational mode decomposition module 903 includes:

[0140] The system oscillation judgment module is used to compare each of the frequency oscillation components with a preset signal amplitude threshold one by one, and when there is a frequency oscillation component that is greater than the preset signal amplitude threshold, it is determined that the power system is oscillating.

[0141] As the device embodiment is basically similar to the method embodiment, it is described in a relatively simple way. For relevant details, please refer to the description of the method embodiment above.

[0142] This invention also provides an electronic device, which includes a processor and a memory:

[0143] The memory is used to store program code and transfer the program code to the processor;

[0144] The processor is used to execute the multi-frequency oscillation identification method of any embodiment of the present invention according to the instructions in the program code.

[0145] This invention also provides a computer-readable storage medium for storing program code for executing the multi-frequency oscillation identification method of any embodiment of this invention.

[0146] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0147] In the embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between devices or units through some interfaces, and may be electrical, mechanical, or other forms.

[0148] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0149] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0150] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0151] The above-described 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 the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying multi-frequency oscillations, characterized in that, include: Collect voltage signal data from the power system; The voltage signal data is subjected to preliminary variational mode decomposition to obtain multiple frequency oscillation components; When the system is determined to be oscillating based on the multiple frequency oscillation components, the voltage signal data is subjected to secondary variational mode decomposition in combination with variational mode decomposition parameter optimization to obtain at least one dominant oscillation component. Fitting is performed based on the at least one dominant oscillation component, and the oscillation characteristics of each dominant oscillation component are calculated based on the fitting results. The step of combining variational mode decomposition parameter optimization to perform secondary variational mode decomposition on the voltage signal data to obtain at least one dominant oscillation component includes: Step S01: Determine the minimum envelope entropy value of each frequency oscillation component by calculation; Step S02: Using the minimum envelope entropy value as the initial fitness value, the particle swarm optimization algorithm is used to optimize the parameters of the variational mode decomposition to determine the optimal parameter combination; Step S03: Calculate the autocorrelation coefficient of each frequency oscillation component, the autocorrelation coefficient being used to determine whether the frequency oscillation component is a noise mode component or a non-noise mode component; Step S04: Using the number of non-noise mode components as the number of the second variational mode decomposition, and based on the optimal parameter combination, perform variational mode decomposition again on the voltage signal data; Step S05: Determine whether there are still noise mode components among the multiple frequency oscillation components generated after re-decomposition. If yes, repeat steps S01 to S04; otherwise, proceed to step S05. Step S06: Based on the last variational mode decomposition, obtain multiple optimized frequency oscillation components, and select at least one dominant oscillation component from the multiple optimized frequency oscillation components.

2. The multi-frequency oscillation identification method according to claim 1, characterized in that, The step of calculating and determining the minimum envelope entropy value of each frequency oscillation component includes: The frequency oscillation components are normalized, and the envelope value of each normalized frequency oscillation component is calculated. The corresponding envelope entropy value is then calculated. The minimum envelope entropy value is taken as the minimum envelope entropy value of each frequency oscillation component.

3. The multi-frequency oscillation identification method according to claim 1, characterized in that, The step of using particle swarm optimization algorithm to optimize the parameters of variational mode decomposition and determine the optimal parameter combination includes: Bandwidth constraints and noise tolerance are used as the combination of parameters to be optimized in variational mode decomposition. Based on the initial fitness value, the combination of parameters to be optimized is optimized using the particle swarm optimization algorithm. During the parameter optimization process, the combination of parameters to be optimized is limited to a pre-set range of parameter combination selections; Repeat the iteration until the iteration reaches the convergence condition or the maximum number of iterations is reached, then stop the iteration and output the optimal parameter combination.

4. The multi-frequency oscillation identification method according to claim 1, characterized in that, The step of selecting at least one dominant oscillation component from the plurality of optimized frequency oscillation components includes: The frequency spectrum of the multiple optimized frequency oscillation components is obtained based on the Hilbert-Huang transform. The amplitude of each optimized frequency oscillation component in the frequency spectrum is compared with a preset signal amplitude threshold, and the optimized frequency oscillation component that is greater than or equal to the preset signal amplitude threshold is taken as the dominant oscillation component.

5. The multi-frequency oscillation identification method according to any one of claims 1 to 4, characterized in that, The preliminary variational mode decomposition of the voltage signal data to obtain multiple frequency oscillation components includes: Based on a preset mode decomposition number, the voltage signal data is subjected to preliminary variational mode decomposition to obtain multiple frequency oscillation components, the number of which corresponds to the preset mode decomposition number.

6. The multi-frequency oscillation identification method according to claim 5, characterized in that, The method of determining system oscillation based on the multiple frequency oscillation components includes: Each frequency oscillation component is compared with a preset signal amplitude threshold. When there is a frequency oscillation component that is greater than the preset signal amplitude threshold, it is determined that the power system is oscillating.

7. A multi-frequency oscillation identification device, characterized in that, include: The data acquisition module is used to acquire voltage signal data from the power system. The preliminary variational mode decomposition module is used to perform preliminary variational mode decomposition on the voltage signal data to obtain multiple frequency oscillation components; The second variational mode decomposition module is used to perform second variational mode decomposition on the voltage signal data when the system oscillation is determined based on the multiple frequency oscillation components, and to obtain at least one dominant oscillation component by combining variational mode decomposition parameter optimization. An oscillation feature calculation module is used to fit based on the at least one dominant oscillation component, and calculate the oscillation features of each dominant oscillation component according to the fitting result. The second-order variational mode decomposition module includes: The minimum envelope entropy value determination module is used to perform step S01: determine the minimum envelope entropy value of each frequency oscillation component by calculation; The parameter optimization module is used to execute step S02: taking the minimum envelope entropy value as the initial fitness value, using the particle swarm optimization algorithm to optimize the parameters of the variational mode decomposition, and determining the optimal parameter combination; The autocorrelation coefficient calculation module is used to perform step S03: calculate the autocorrelation coefficient of each frequency oscillation component, wherein the autocorrelation coefficient is used to determine whether the frequency oscillation component is a noise mode component or a non-noise mode component; The second variational mode decomposition submodule is used to execute step S04: taking the number of non-noise mode components as the second variational mode decomposition number, and re-performing variational mode decomposition on the voltage signal data based on the optimal parameter combination; The noise mode component determination module is used to execute step S05: determine whether there are still noise mode components among the multiple frequency oscillation components generated after re-decomposition. If yes, repeat steps S01 to S04; if no, jump to step S05. The oscillation dominant component determination module is used to perform step S06: based on the last variational mode decomposition, obtain multiple optimized frequency oscillation components, and select at least one oscillation dominant component from the multiple optimized frequency oscillation components.

8. An electronic device, characterized in that, The device includes a processor and a memory: The memory is used to store program code and transmit the program code to the processor; The processor is configured to execute the multi-frequency oscillation identification method according to any one of claims 1-6 according to the instructions in the program code.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store program code for executing the multi-frequency oscillation identification method according to any one of claims 1-6.

Images