A high-precision double-end fault positioning method for distribution network against line parameter error

Abstract

The application relates to the technical field of power distribution network fault positioning, in particular to a power distribution network double-end fault high-precision positioning method resisting line parameter errors, which comprises the following steps: collecting three-phase currents and three-phase voltages of nodes at two ends of a power distribution network transmission line, analyzing waveform characteristics and energy distribution characteristics of traveling wave signals of zero-mode and line-mode components in different frequency band IMF components, constructing wave head consistency indexes and multi-source propagation consistency indexes of the IMF components, adaptively adjusting wavelet denoising thresholds of the IMF components based on the two indexes, and improving the denoising effect of the IMF components; performing modal reconstruction according to the denoised modal components, and positioning faults according to the reconstructed zero-mode components and line-mode components; the method solves the problem that the traditional wavelet denoising cannot fully consider the propagation characteristic differences of wide-band zero-mode and line-mode components under different frequency bands, and the denoising effect is poor; and the accuracy of power distribution network transmission line fault positioning is improved.

Description

Technical Field

[0001] This application relates to the field of distribution network fault location technology, specifically to a high-precision fault location method for dual-end faults in distribution networks that is resistant to line parameter errors. Background Technology

[0002] With the accelerated advancement of energy transition, the large-scale integration of distributed power sources has profoundly altered the traditional structure and operational characteristics of distribution networks. The widespread integration of new energy nodes such as photovoltaics, wind power, and energy storage has transformed distribution networks from traditional "single-source radial networks" into "multi-source complex networks." This increased complexity in network topology not only leads to a significant increase in the frequency of faults but also makes them more sudden and random. Therefore, accurate fault location after a distribution network line fault occurs is crucial for rapid repair and ensuring power supply reliability.

[0003] In existing fault location methods, double-ended traveling wave ranging based on the traveling wave method locates the fault by analyzing the propagation time difference between the traveling wave signal generated at the fault point and the nodes at both ends of the line. This method boasts high ranging accuracy and has been widely used. However, distribution network transmission lines have numerous branches and complex operating environments, making traveling wave waveforms susceptible to noise interference and distortion. Traditional wavelet threshold denoising methods rely solely on the statistical characteristics of the signal to uniformly denoise and enhance components across all frequency bands. This fails to fully consider the differences in propagation characteristics of broadband zero-mode and line-mode components across different frequency bands. Consequently, some true traveling wave characteristics in certain frequency bands are easily misjudged as noise and weakened. This results in the denoised mode components failing to accurately reflect the true propagation time of the wavefront, ultimately leading to deviations in fault location results. Summary of the Invention

[0004] To address the aforementioned technical problems, this application provides a high-precision fault location method for dual-end faults in distribution networks that is resistant to line parameter errors, thereby resolving the existing issues.

[0005] The high-precision fault location method for double-ended faults in a distribution network that is resistant to line parameter errors in this application adopts the following technical solution: One embodiment of this application provides a high-precision fault location method for a distribution network with resistance to line parameter errors. The method includes the following steps: Step 1: Identify the faulty power transmission lines in the distribution network by using zero-sequence current. For the faulty power transmission lines, obtain the three-phase current and three-phase voltage at the nodes at both ends of the power transmission line. Decouple the three-phase current and three-phase voltage respectively to obtain the corresponding zero-mode component and line-mode component. Step 2: Perform mode decomposition on the zero-mode component and the line-mode component respectively to obtain the corresponding mode components; perform wavelet decomposition on each mode component, wherein the wavelet threshold is adaptively adjusted according to the energy distribution characteristics in the mode component to obtain the denoised mode components. The adaptive adjustment of the wavelet threshold specifically includes: determining each suspected traveling wave signal segment in the modal component based on the extreme value distribution in the Teager energy sequence of each modal component; and adaptively adjusting the original wavelet threshold based on the degree of disorder in the distribution of the differences between the peak values ​​of all adjacent suspected traveling wave signal segments in the modal component, as well as the differences in the energy of the suspected traveling wave signal segments of the corresponding modal components of current and voltage. Step 3: Modal reconstruction of the zero-mode and line-mode components of the current is performed using the denoised modal components to obtain the denoised zero-mode and line-mode components. Traveling wave front detection is then performed on each component to determine the initial travel wave front propagation time of the zero-mode and line-mode components of the current at both ends. Combined with the length of the transmission line, the traveling wave method is used for fault location.

[0006] Preferably, the suspected traveling wave signal segment specifically refers to: For any maximum point in the Teager energy sequence of the modal component, search outwards from the maximum point. When the energy operator value is found to be less than a preset first threshold on each side for the first time, the search stops. The time interval between the time points when the search stops on both sides of the maximum point is taken as the time interval corresponding to a suspected traveling wave signal segment. The signals in each of the time intervals in the modal component are taken as each suspected traveling wave signal segment of the modal component.

[0007] Preferably, the adaptive adjustment of the original wavelet threshold based on the degree of disorder in the distribution of differences between the peak values ​​of all adjacent suspected traveling wave signal segments in the modal components, and the differences in the energy of the suspected traveling wave signal segments of the corresponding modal components of current and voltage, includes: The wavefront consistency index of the modal component is determined based on the degree of disorder in the distribution of the differences between the peak values ​​of all adjacent suspected traveling wave signal segments in the modal component. The multi-source propagation consistency index of the modal components is constructed based on the difference in energy between the suspected traveling wave signal segments of the corresponding modal components of current and voltage. The wavelet threshold is adaptively adjusted based on the wavehead consistency index and the multi-source propagation consistency index.

[0008] Preferably, the calculation process of the wavehead consistency index is as follows: The peak values ​​of each suspected traveling wave signal segment in the modal component are obtained, and the amplitude ratio between the peak value of each suspected traveling wave signal segment and the peak value of the next suspected traveling wave signal segment in the modal component is calculated. The wavefront consistency index of the modal component is determined based on the degree of disorder in the distribution of the amplitude ratio.

[0009] Preferably, the wavefront consistency index for determining the modal components based on the degree of disorder in the distribution of the amplitude ratio is specifically as follows: The normalized value of the coefficient of variation of all the amplitude ratios in the modal components is used as the wavefront consistency index of the modal components.

[0010] Preferably, the calculation process of the multi-source propagation consistency index is as follows: The energy of each suspected traveling wave signal segment is obtained by amplitude integration, and an energy sequence is formed based on the normalized values ​​of the energy of all suspected traveling wave signal segments in the modal components. The multi-source propagation consistency index of the corresponding modal components is determined based on the difference between the energy sequences of the corresponding modal components of current and voltage.

[0011] Preferably, the determination of the multi-source propagation consistency index of the corresponding mode component based on the difference between the energy sequences of the corresponding mode components of current and voltage specifically involves: The difference distance between the energy sequences of the corresponding modal components of current and voltage is calculated by a distance metric algorithm, and the normalized distance is used as the multi-source propagation consistency index of the corresponding modal components.

[0012] Preferably, the adaptive adjustment of the wavelet threshold based on the wavefront consistency index and the multi-source propagation consistency index includes: Calculate the sum of the wavefront consistency index and the multi-source propagation consistency index corresponding to each modal component; adaptively adjust the original wavelet threshold for wavelet denoising based on the sum.

[0013] Preferably, the adaptive adjustment of the original wavelet threshold for wavelet denoising based on the sum value specifically involves: The product of the Sigmoid function value (with the sum as input) and the original wavelet threshold is used as the adaptively adjusted wavelet threshold.

[0014] Preferably, the fault location includes: The line mode components include a 1-mode component and a 2-mode component. The modulus maxima of the denoised current 1-mode component and the 2-mode component are extracted respectively. The wavefront propagation time of the initial traveling wave of the line mode component corresponding to the maximum value of the two modulus maxima is selected. Combined with the wavefront propagation time of the initial traveling wave of the zero-mode component and the transmission line length, fault location is performed.

[0015] This application has at least the following beneficial effects: This application collects three-phase current and three-phase voltage at both ends of a power distribution network transmission line; analyzes the waveform characteristics and energy distribution characteristics of the zero-mode and line-mode components of the current and voltage data in the traveling wave signals of the IMF components at different frequency bands to construct wavefront consistency index and multi-source propagation consistency index, which are used to reflect the severity of noise impact on the zero-mode and line-mode components at different frequency bands. Based on these two indices, the wavelet denoising threshold of the IMF components at each frequency band is adaptively adjusted to improve the denoising effect of the IMF components, facilitating the accuracy of subsequent traveling wave wavefront propagation time calibration; and adjusts the wavelet denoising threshold of the IMF components in accordance with the denoised modal components. Modal reconstruction is performed on the modal components. Based on the reconstructed zero-mode and line-mode components at both ends of the transmission line, the wavefront propagation time of the initial traveling wave is calculated. Combined with the length of the transmission line, the double-ended traveling wave method is used for fault location. This method solves the problem that traditional methods only rely on the statistical characteristics of the signal to uniformly denoise and enhance the components of each frequency band, without fully considering the differences in the propagation characteristics of broadband zero-mode and line-mode components in different frequency bands. This can easily lead to the true traveling wave characteristics in some frequency bands being misjudged as noise and weakened, thus making the denoised modal components unable to accurately reflect the true wavefront propagation time. This method improves the accuracy of fault location in distribution network transmission lines. Attached Figure Description

[0016] Figure 1 A flowchart illustrating a high-precision fault location method for a distribution network with resistance to line parameter errors, provided in this application; Figure 2 This is a schematic diagram illustrating the calculation process of the wave head consistency index in this application; Figure 3 This is a schematic diagram illustrating the calculation process of the multi-source propagation consistency index in this application. Detailed Implementation

[0017] The following description, in conjunction with the accompanying drawings, details the specific scheme of the high-precision fault location method for dual-end faults in a distribution network that is resistant to line parameter errors provided in this application.

[0018] One embodiment of this application provides a high-precision fault location method for dual-end faults in a distribution network that is resistant to line parameter errors.

[0019] Specifically, a high-precision fault location method for double-ended faults in distribution networks that is resistant to line parameter errors is provided below. Please refer to [link / reference]. Figure 1 The method includes the following steps: Step 1: Identify the faulty power distribution network transmission line by using the zero-sequence current. For the faulty transmission line, obtain the three-phase current and three-phase voltage at the nodes at both ends of the transmission line. Decouple the three-phase current and three-phase voltage respectively to obtain the corresponding zero-mode component and line-mode component.

[0020] This application takes a single-phase grounding fault in a distribution network as an example and uses the traveling wave method for fault location. First, power data is acquired, specifically by setting a power data statistical period; using zero-sequence current transformers, broadband current transformers, and capacitive voltage dividers, the neutral point zero-sequence current, three-phase current, and three-phase voltage data of the transmission line nodes in the distribution network are synchronously collected. The time-series data of the zero-sequence current, three-phase current, and three-phase voltage at both ends of the transmission line within the statistical period are obtained. The acquired power data is transmitted to the power grid central station via a 5G wireless communication module, enabling remote data interaction.

[0021] It should be noted that, regarding the setting of the statistical duration, since this application uses the traveling wave method for fault location, and the traveling wave travels extremely fast in the transmission line, the statistical duration is set based on the power frequency period in this application. The statistical duration is set to J times the power frequency period, where J∈[1,5]. The specific value can be determined according to the application scenario. This application does not impose any specific restrictions. In this embodiment, J is taken as 2, that is, the statistical duration is 40ms.

[0022] The transmission line node refers to a key node in the distribution network that has two or more branch connections, such as switchgear and ring main units.

[0023] In this embodiment, the sampling frequency of current and voltage data is set to 10MHz. Implementers can flexibly adjust it according to actual needs and site conditions. This application does not impose specific limitations.

[0024] For acquiring power data at both ends of any transmission line, time synchronization protocols such as GPS or PTP are used to ensure that the power data collected at both ends of the line are strictly aligned in time, meeting the synchronization accuracy requirements for subsequent fault location.

[0025] Furthermore, based on the zero-sequence current of the transmission lines within the aforementioned statistical period, the power frequency zero-sequence current amplitude comparison method is used to detect single-phase grounding faults and identify the transmission lines with single-phase grounding faults. The power frequency zero-sequence current amplitude comparison method for detecting single-phase grounding faults is a well-known technique in the power system field, and its specific process will not be elaborated further.

[0026] Furthermore, for transmission lines where a single-phase ground fault is detected, the traveling wave method is used to locate the fault based on the three-phase current and three-phase voltage data of the nodes at both ends of the transmission line in the above statistical time period.

[0027] During fault location, the three-phase current and three-phase voltage data for each node in the acquired statistical time period are first decoupled using Karrenbauer transform to obtain the corresponding zero-mode and line-mode components. The line-mode components include 1-mode and 2-mode components. Taking three-phase current as an example, the zero-mode, 1-mode, and 2-mode components obtained from the three-phase current decoupling are expressed as follows: 1 modulus component The loop component reflecting the current difference between phase A and phase B, a two-modulus component. The loop component reflects the current difference between phase A and phase C. The Karrenbauer transformation is adapted to existing technology; the specific process will not be elaborated further. Considering the coupling between the three phases in a transmission line, not only will the faulty phase generate traveling waves, but the non-faulty phases will also generate traveling waves due to mutual inductance. Therefore, the zero-mode component, the first-mode component, and the second-mode component are all used for fault location analysis. Similarly, the zero-mode component, the first-mode component, and the second-mode component of the three-phase voltage are expressed as follows: Then, the zero-mode component and the linear-mode component are processed.

[0028] Step 2: Perform mode decomposition on the zero-mode component and the linear mode component respectively to obtain the corresponding mode components; perform wavelet decomposition on each mode component, wherein the wavelet threshold is adaptively adjusted according to the energy distribution characteristics in the mode component to obtain the denoised mode components.

[0029] With the large-scale integration of distributed power sources into distribution networks, the topology of distribution network lines is gradually exhibiting multi-branch characteristics. When a transmission line fault occurs, the traveling wave signal generated at the fault point propagates through the multi-branch lines, generating reflected and refracted waves at points of impedance discontinuity. This results in multiple traveling wave fronts in a single fault sampling signal, increasing the difficulty of identifying the traveling wave fronts.

[0030] Meanwhile, in the field operation environment of transmission lines, there are various interference sources such as electromagnetic interference and load fluctuations. The noise signals generated by these interference sources are superimposed on the traveling wave signal, causing the waveform of the traveling wave signal to be distorted, so that the abrupt change characteristics reflecting the arrival time of the wavefront are submerged by the noise signal.

[0031] To address the aforementioned issues, existing technologies typically employ traditional wavelet thresholding denoising methods to denoise the zero-mode and line-mode components of power signals. These methods set wavelet thresholds based on the statistical characteristics of coefficients in each frequency band, enhancing the signal components across different frequency bands. However, because these methods do not consider the differences in propagation attenuation and dispersion characteristics of downwave signals in distribution network lines across different frequency bands—for example, high-frequency components attenuate quickly and propagate over short distances, while low-frequency components attenuate slowly and propagate over long distances—it is easy for genuine traveling wave abrupt changes within some frequency bands to be misjudged as noise filtering due to statistical values ​​falling below the threshold. This results in a time-domain shift in the reconstructed traveling wave signal, ultimately affecting the accuracy of wavefront propagation time identification.

[0032] (1): The fault traveling wave signal has an extremely wide frequency band, typically covering tens of kHz to thousands of kHz. To conduct a detailed analysis of the characteristics of the traveling wave signal under different frequency bands, variational mode decomposition (VMD) is first used to decompose the zero-mode component and the line-mode component to obtain K intrinsic mode functions (IMFs). The value of K can be set by the implementer according to the implementation scenario, and this application does not impose any special restrictions; in this embodiment, the value of K is 10.

[0033] (2): To reflect the propagation consistency and noise interference level of downlink waves in each frequency band, signal extraction is performed on each IMF component to obtain each suspected traveling wave signal segment in the IMF component. Specifically: The Teager energy sequence of this IMF component is calculated using the Teager energy operator to enhance the local amplitude abrupt change characteristics of the traveling wavefront. The calculation of the Teager energy operator is a well-known technique, and the specific process will not be elaborated further.

[0034] Subsequently, local maxima in the Teager energy sequence are calculated and used as potential traveling wave candidate points. A search is then performed outwards from each potential traveling wave candidate point. During the search, 10% of the Teager energy operator value of the potential traveling wave candidate point is used as the first threshold. The search stops when the Teager energy operator value on each side is found to be less than the first threshold for the first time. Specifically, when searching from a potential traveling wave candidate point to any side of the sequence, if the Teager energy operator value is still greater than or equal to the first threshold when reaching the boundary of the starting or ending sampling point, this sampling point is directly used as the stop search time point for that side. The time points at which the search stops on both sides are obtained, and the time interval between these two time points is used as the time interval corresponding to a potential traveling wave signal segment in the IMF component. The signal within this time interval in the IMF component is then used as a potential traveling wave signal segment in the IMF component.

[0035] (3): In a preferred embodiment of this application, the amplitude attenuation characteristics of the suspected traveling wave signal segments in each IMF component are analyzed to construct a wavefront consistency index, such as... Figure 2 As shown, the specific calculation process of the wavehead consistency index is as follows: Taking the k-th IMF component of the three-phase current 1-mode component as an example, the peak amplitude of the n-th suspected traveling wave signal segment in this IMF component is denoted as... , That is, the amplitude of the maximum point within the suspected traveling wave signal segment, where , This represents the total number of suspected traveling wave signal segments in the IMF component. In particular, if the number of suspected traveling wave signal segments extracted from a certain low-frequency band IMF component is less than 3, it indicates that the signal characteristics of this frequency band are relatively smooth and do not have practical statistical significance. In this case, the IMF component will no longer participate in the subsequent wavelet threshold adaptive adjustment, but will be directly denoised based on the wavelet threshold obtained from noise estimation.

[0036] Furthermore, according to the temporal relationship of the suspected traveling wave signal segments, the ratio between the peak amplitude of each suspected traveling wave signal segment in the IMF component and the peak amplitude of its next suspected traveling wave signal segment is calculated sequentially, forming a peak-to-peak ratio sequence. The nth element of this peak-to-peak ratio sequence can be represented as... ,in .

[0037] Furthermore, the coefficient of variation of all elements in the peak-to-peak ratio sequence of this IMF component is calculated, denoted as . .

[0038] Furthermore, adopting and The same acquisition method is used to calculate the peak-to-peak ratio (PSRR) sequence coefficient of variation for each IMF component of the three-phase current data's first-order modulus component. Using the maximum value among all coefficients of variation as a benchmark, each coefficient of variation is normalized and mapped to the [0,1] interval. Specifically, if the maximum value among all coefficients of variation is equal to 0, the normalized value of each coefficient of variation is set to 0. The wavefront consistency index of each IMF component is defined as the normalized value of the PSRR sequence coefficient of variation for that IMF component. The wavefront consistency index of the k-th IMF component of the three-phase current data's first-order modulus component is denoted as... .

[0039] The initial traveling wave generated by a line fault undergoes energy attenuation during its subsequent propagation, forming secondary traveling waves through multiple refractions and reflections. Therefore, the amplitude of the traveling wave gradually decreases until it disappears, following a chronological order. If there is an impedance discontinuity in the transmission line, the attenuation of the reflected traveling wave formed after reflection and refraction at that point depends primarily on the impedance values ​​on both sides of the discontinuity. Since the impedance values ​​on both sides remain constant for a short period, the attenuation ratio of the peak amplitude of each reflected traveling wave is basically the same. Therefore, the peak ratios of adjacent suspected traveling wave signal segments should be relatively close.

[0040] In statistical analysis, the coefficient of variation (CV) is used to measure the dispersion of a data sequence. This scheme assesses the noise interference level of a frequency band by calculating the CV of the peak-to-peak ratio (PTR) sequence within a single IMF component. A larger CV indicates a higher degree of dispersion in the PTR distribution, deviating from the normal attenuation pattern of traveling waves in transmission lines. This suggests strong noise interference in the frequency band, with some suspected traveling wave signal segments exhibiting abnormally high amplitudes due to noise superposition, thus disrupting the attenuation characteristics of the traveling wave. Conversely, a smaller CV indicates that the suspected traveling wave signal segments within the IMF component conform to normal traveling wave attenuation characteristics in terms of timing, indicating weaker noise interference in the frequency band.

[0041] (4): In a preferred embodiment of this application, the energy distribution characteristics of the corresponding IMF components in the zero-mode and line-mode components of current and voltage data are analyzed to construct a multi-source propagation consistency index, such as... Figure 3 As shown, the specific calculation process of the multi-source propagation consistency index is as follows: Current and voltage data under the same modal components originate from the same fault traveling wave propagation process; even the modal components of non-faulty phases have current and voltage generated by inter-phase coupling mutual inductance excitation, so current and voltage under the same mode have propagation consistency in the fault traveling wave signal characterization. In addition, the IMF components obtained after VMD decomposition of each zero-mode component or line-mode component correspond to a relatively narrow frequency band, and the corresponding IMF components of the zero-mode components or line-mode components of current and voltage originate from fault transient components in a similar frequency range; if a large number of real traveling wave components are retained in the IMF component, the energy changes of the two should have a high degree of consistency.

[0042] Taking the kth IMF component of the 1-mode component of three-phase current and the kth IMF component of the 1-mode component of three-phase voltage as examples, based on the above analysis, firstly, for the time intervals corresponding to each suspected traveling wave signal segment detected in the kth IMF component of the 1-mode component of the three-phase current data, each time interval is mapped to the kth IMF component of the 1-mode component of the three-phase voltage data to determine the suspected traveling wave signal segment corresponding to the kth IMF component of the 1-mode component of the three-phase voltage data.

[0043] Furthermore, using the time interval of the nth suspected traveling wave signal segment in the kth IMF component of the three-phase current data as the upper and lower limits, amplitude integration is performed on this suspected traveling wave signal segment to calculate its energy. The energy of all suspected traveling wave signal segments in this IMF component is obtained sequentially, forming an energy sequence. Similarly, calculate the energy sequence of the k-th IMF component of the three-phase voltage data's first-order modulus component. And use maximum and minimum value normalization to respectively... and The element values ​​in the sequence are mapped to the [0,1] interval to facilitate comparison of two energy sequences. It should be noted that if the energy sequence... or energy sequence If the maximum and minimum values ​​are equal, then the energy sequence will be... or energy sequence The normalized value of all elements in the middle is set to 1.

[0044] Furthermore, the multi-source propagation consistency index of the k-th IMF component of the 1-mode component of the three-phase current data is calculated. , The expression is: ,in, This indicates that the difference distance between two sequences is calculated; in this embodiment, Euclidean distance is used for calculation. and The difference distance between them. It should be noted that implementers can also use other existing distance metric algorithms to calculate this. and This application does not impose specific limitations on the difference in distance between them.

[0045] In this scheme, through measurement and The differences were assessed to evaluate the propagation consistency of suspected traveling wave signal segments in the current and voltage IMF components within similar frequency bands. Since the k-th IMF component of the current and voltage IMF components retains similar frequency components and propagation patterns, the energy changes of the corresponding suspected traveling wave signal segments should exhibit high synchronicity. As the fault traveling wave continues to propagate, reflect, and refract in the line, although the corresponding traveling waves in the current and voltage exhibit synchronous energy attenuation due to line parameters, their energy evolution trends still maintain a high correlation; when noise interference is weak, the calculated... The noise level is relatively low. Conversely, if there is strong noise interference or mode aliasing in the k-th IMF component, the random noise does not satisfy the common propagation law of current and voltage, causing the energy distribution of the corresponding suspected traveling wave signal segment to deviate randomly to varying degrees, and the correlation between the two energy sequences decreases. It increases accordingly.

[0046] Furthermore, the multi-source propagation consistency index of all IMF components in the 1-mode component of the three-phase current data is calculated, and the maximum-minimum normalization method is used to... The values ​​are mapped to the [0,1] interval. It should be noted that during the normalization process, if the maximum and minimum values ​​of all multi-source propagation consistency indices are equal, then the normalized values ​​of all multi-source propagation consistency indices are set to 0.

[0047] (5): Based on the above analysis, the k-th IMF component of the 1-mode component of the three-phase current data is decomposed using the sym8 wavelet basis. In this embodiment, the number of decomposition levels is 6. In other embodiments of this application, the implementer can set the number of decomposition levels according to the actual situation. Furthermore, based on the wavefront consistency index... Consistency Index of Multi-Source Propagation Calculate the adaptive wavelet threshold of the k-th IMF component of the three-phase current data modulus 1. The expression is: In the formula, The original wavelet threshold used in wavelet thresholding denoising is obtained by noise estimation of the IMF components. The solution is a well-known technique for wavelet threshold denoising, and its specific expression will not be elaborated here. for The normalized value; The weights of the independent variables are The Sigmoid function, where Sensitivity factor, used to regulate Sensitivity to threshold adjustment, ensuring When the value is large, the function value can saturate quickly. The typical value range is: Its size can be set by the implementer according to the implementation scenario. This application does not impose any special restrictions. In this embodiment, The value is 3.

[0048] use As The weight can avoid Excessive value reduction leads to a loss of noise reduction capability. , A larger value indicates that the traveling wavefront in the IMF component of that frequency band is severely distorted due to noise interference. In this case, maintaining the original noise estimate is necessary. Perform noise reduction; conversely, when , A lower value indicates weaker noise interference in that frequency band, so the value should be appropriately reduced. This is to avoid misclassifying real traveling wavefronts with steep rising edges and rapid fluctuations as noise, thus weakening their feature information. Finally, based on the obtained adaptive wavelet threshold... Wavelet denoising is performed on the k-th IMF component of the three-phase current's first-mode component. Wavelet denoising is an existing technique, and its specific process will not be described in detail.

[0049] Step 3: Modal reconstruction of the zero-mode and line-mode components of the current is performed using the denoised modal components to obtain the denoised zero-mode and line-mode components. Traveling wave front detection is then performed on each component to determine the initial travel wave front propagation time of the zero-mode and line-mode components of the current at both ends. Combined with the length of the transmission line, the traveling wave method is used for fault location.

[0050] Step (1): Using the same denoising method as the kth IMF component of the three-phase current data 1-modulus component, denoise all IMF components of the three-phase current data 1-modulus component, and then reconstruct the signal of the denoised IMF components to obtain the denoised three-phase current data 1-modulus component. Furthermore, the zero-mode and two-mode components of the three-phase current data are denoised using the same denoising method as the 1-mode component of the three-phase current data.

[0051] Step (2): Initial traveling wave propagation time calibration is performed on the zero-mode component and line-mode component of the denoised three-phase current data. Taking the 1-mode component as an example, it is decomposed using Discrete Wavelet Transform (DWT). In this embodiment, a first-order Gaussian wavelet is used as the basis function, and the decomposition scale is set to 4. Then, the arrival time of the traveling wave front is determined by the detection method based on the modulus maxima, and the arrival time of the first traveling wave front is taken as the initial traveling wave front propagation time. The traveling wave front detection method based on the modulus maxima is an existing technology in the power field, and the specific process will not be described in detail.

[0052] The initial traveling wave propagation time of the zero-mode component and the second-mode component after denoising the three-phase current data are calibrated using the same initial traveling wave propagation time calibration method as the first-mode component.

[0053] It should be noted that in asymmetric fault scenarios, not all line-mode components contain obvious traveling wave abrupt changes. To accurately obtain the wavefront arrival time, the modulus maxima of the denoised 1-mode and 2-mode components of the three-phase current are first extracted. Then, the modulus maxima values ​​of the two line-mode components are compared. Finally, the line-mode component corresponding to the maximum value of the two modulus maxima is selected, and its corresponding initial traveling wave front propagation time is taken as the final determined initial traveling wave front propagation time of the three-phase current line-mode components. Specifically, when the modulus maxima of the denoised 1-mode component is equal to the modulus maxima of the 2-mode component, the time obtained from the extraction of the 1-mode component is fixed as the initial traveling wave front propagation time of the three-phase current line-mode components.

[0054] Step (3): The wavefront propagation time of the initial traveling wave of the line mode and zero mode components of the current data at both ends (end A and end B) of the transmission line is denoted as follows: The time of the fault is recorded as The total length of the transmission line is denoted as The distance from the fault point to end A is denoted as . The distance from the fault point to end B is denoted as Fault location was performed using a two-terminal traveling wave method based on zero-mode line mode time difference. The process included: The initial traveling wave of the fault traveling wave propagates to end A in a linear mode, satisfying the following conditions: Zero-mode propagation satisfies Similarly, the initial traveling wave of the fault traveling wave propagates to end B in a linear mode, satisfying the following conditions: Zero-mode propagation satisfies .

[0055] Then, by combining the propagation formulas for the linear mode and zero mode at end A, we obtain Relationship 1: Similarly, by combining the propagation formulas for the linear mode and zero mode at end B, we obtain Relationship 2: This eliminates the moment when the fault occurred.

[0056] Furthermore, by combining the formulas of relation 1 and relation 2, we obtain... Further simplification yields At this point, the wave velocities at both ends are eliminated, and the distance from end A to the location of the fault can be calculated solely based on the transmission line length and the propagation time of the initial traveling wave front in the line mode and zero mode components. This calculation method eliminates the interference caused by parameter errors in transmission lines leading to deviations in traveling wave propagation speed from theoretical values, thus preventing errors in fault location. It only requires the transmission line length and traveling wave propagation time for calculation, avoiding location errors caused by unstable traveling wave speeds due to large differences in line parameters. This enables high-precision fault location in transmission lines within distribution networks. The double-ended traveling wave method based on zero-mode line mode time difference is a well-known technique.

[0057] In summary, this application embodiment collects three-phase currents and three-phase voltages at both ends of a power distribution network transmission line; analyzes the waveform characteristics and energy distribution characteristics of the traveling wave signals in the IMF components of the current and voltage data in different frequency bands to construct wavefront consistency index and multi-source propagation consistency index, which are used to reflect the severity of noise impact on the zero-mode and line-mode components in different frequency bands. Based on these two indices, the wavelet denoising threshold of the IMF components in each frequency band is adaptively adjusted to improve the denoising effect of the IMF components, facilitating the accuracy of subsequent traveling wave wavefront propagation time calibration; and adjusts the wavelet denoising threshold of the IMF components in each frequency band based on the denoised modal components. Modal reconstruction is performed on the zero-mode and line-mode components. Based on the reconstructed zero-mode and line-mode components at both ends of the transmission line, the wavefront propagation time of the initial traveling wave is calculated. Combined with the length of the transmission line, the double-ended traveling wave method is used for fault location. This method solves the problem that traditional methods only rely on the statistical characteristics of the signal to uniformly denoise and enhance the components of each frequency band, without fully considering the differences in the propagation characteristics of broadband zero-mode and line-mode components in different frequency bands. This can easily lead to the true traveling wave characteristics in some frequency bands being misjudged as noise and weakened, thus making the denoised modal components unable to accurately reflect the true wavefront propagation time. This method improves the accuracy of fault location in distribution network transmission lines.

Claims

1. A high-precision fault location method for double-ended faults in a distribution network that is resistant to line parameter errors, characterized in that, The method includes the following steps: Step 1: Identify the faulty power transmission lines in the distribution network by using zero-sequence current. For the faulty power transmission lines, obtain the three-phase current and three-phase voltage at the nodes at both ends of the power transmission line. Decouple the three-phase current and three-phase voltage respectively to obtain the corresponding zero-mode component and line-mode component. Step 2: Perform mode decomposition on the zero-mode component and the line-mode component respectively to obtain the corresponding mode components; perform wavelet decomposition on each mode component, wherein the wavelet threshold is adaptively adjusted according to the energy distribution characteristics in the mode component to obtain the denoised mode components. The adaptive adjustment of the wavelet threshold specifically includes: determining each suspected traveling wave signal segment in the modal component based on the extreme value distribution in the Teager energy sequence of each modal component; and adaptively adjusting the original wavelet threshold based on the degree of disorder in the distribution of the differences between the peak values ​​of all adjacent suspected traveling wave signal segments in the modal component, as well as the differences in the energy of the suspected traveling wave signal segments of the corresponding modal components of current and voltage. Step 3: Modal reconstruction of the zero-mode and line-mode components of the current is performed using the denoised modal components to obtain the denoised zero-mode and line-mode components. Traveling wave front detection is then performed on each component to determine the initial travel wave front propagation time of the zero-mode and line-mode components of the current at both ends. Combined with the length of the transmission line, the traveling wave method is used for fault location.

2. The high-precision fault location method for double-ended faults in a distribution network with resistance to line parameter errors as described in claim 1, characterized in that, The suspected traveling wave signal segment is specifically: For any maximum point in the Teager energy sequence of the modal component, search outwards from the maximum point. When the energy operator value is found to be less than a preset first threshold on each side for the first time, the search stops. The time interval between the time points when the search stops on both sides of the maximum point is taken as the time interval corresponding to a suspected traveling wave signal segment. The signals in each of the time intervals in the modal component are taken as each suspected traveling wave signal segment of the modal component.

3. The high-precision fault location method for double-ended faults in a distribution network with resistance to line parameter errors as described in claim 1, characterized in that, The adaptive adjustment of the original wavelet threshold based on the degree of disorder in the distribution of differences between the peak values ​​of all adjacent suspected traveling wave signal segments in the modal components, and the differences in the energy of the suspected traveling wave signal segments of the corresponding modal components of current and voltage, includes: The wavefront consistency index of the modal component is determined based on the degree of disorder in the distribution of the differences between the peak values ​​of all adjacent suspected traveling wave signal segments in the modal component. The multi-source propagation consistency index of the modal components is constructed based on the difference in energy between the suspected traveling wave signal segments of the corresponding modal components of current and voltage. The wavelet threshold is adaptively adjusted based on the wavehead consistency index and the multi-source propagation consistency index.

4. The high-precision fault location method for double-end faults in a distribution network with resistance to line parameter errors as described in claim 3, characterized in that, The calculation process for the wavehead consistency index is as follows: Obtain the peak value of each suspected traveling wave signal segment in the modal component, and calculate the amplitude ratio between the peak value of each suspected traveling wave signal segment and the peak value of the next suspected traveling wave signal segment in the modal component. The wavefront consistency index of the modal components is determined based on the degree of disorder in the distribution of the amplitude ratio.

5. The high-precision fault location method for a distribution network with resistance to line parameter errors as described in claim 4, characterized in that, The wavefront consistency index for determining the modal component based on the degree of disorder in the distribution of the amplitude ratio is specifically as follows: The normalized value of the coefficient of variation of all the amplitude ratios in the modal components is used as the wavefront consistency index of the modal components.

6. The high-precision fault location method for double-ended faults in a distribution network with resistance to line parameter errors as described in claim 3, characterized in that, The calculation process for the multi-source propagation consistency index is as follows: The energy of each suspected traveling wave signal segment is obtained by amplitude integration, and an energy sequence is formed based on the normalized values ​​of the energy of all suspected traveling wave signal segments in the modal components. The multi-source propagation consistency index of the corresponding modal components is determined based on the difference between the energy sequences of the corresponding modal components of current and voltage.

7. The high-precision fault location method for a distribution network with resistance to line parameter errors as described in claim 6, characterized in that, The determination of the multi-source propagation consistency index of the corresponding mode component based on the difference between the energy sequences of the corresponding mode components of current and voltage is specifically as follows: The difference distance between the energy sequences of the corresponding modal components of current and voltage is calculated by a distance metric algorithm, and the normalized distance is used as the multi-source propagation consistency index of the corresponding modal components.

8. The high-precision fault location method for a distribution network with resistance to line parameter errors as described in claim 3, characterized in that, The adaptive adjustment of the wavelet threshold based on the wavefront consistency index and the multi-source propagation consistency index includes: Calculate the sum of the wavefront consistency index and the multi-source propagation consistency index corresponding to each modal component; adaptively adjust the original wavelet threshold for wavelet denoising based on the sum.

9. The high-precision fault location method for a distribution network with resistance to line parameter errors as described in claim 8, characterized in that, The adaptive adjustment of the original wavelet threshold for wavelet denoising based on the sum value specifically involves: The product of the Sigmoid function value (with the sum as input) and the original wavelet threshold is used as the adaptively adjusted wavelet threshold.

10. A high-precision fault location method for a distribution network with resistance to line parameter errors as described in claim 1, characterized in that, The fault location includes: The line mode components include a 1-mode component and a 2-mode component. The modulus maxima of the denoised current 1-mode component and the 2-mode component are extracted respectively. The wavefront propagation time of the initial traveling wave of the line mode component corresponding to the maximum value of the two modulus maxima is selected. Combined with the wavefront propagation time of the initial traveling wave of the zero-mode component and the transmission line length, fault location is performed.

Images