Plateau mountain power transmission line fault type identification method and computer system

Abstract

The application relates to the technical field of relay protection, in particular to a fault type identification method and a computer system for a highland mountain power transmission line. A two-stage aggregation strategy is used to extract a fault event section from a denoised fault current signal, fault component currents are extracted for each fault event section, and average power ratios of fault duration stages and decay stages are calculated. The average power ratios are compared with a discrimination threshold value to determine the fault type, and reliable fault type identification under the highland mountain power transmission line is realized. The application aims to solve the problem of how to realize reliable lightning stroke identification in the special environment of highlands and mountains.

Description

Technical Field

[0001] This application relates to the field of relay protection technology, and in particular to a fault type identification method and computer system for transmission lines in high-altitude mountainous areas. Background Technology

[0002] Accurate fault identification of transmission lines is a key link in ensuring the safe and stable operation of the power grid. In particular, distinguishing between lightning strike faults and ordinary short-circuit faults is of great significance for guiding line inspection and maintenance, evaluating line lightning protection performance, and implementing differentiated protection strategies.

[0003] In the unique application scenario of high-altitude mountainous areas, the complex terrain, harsh climate, and frequent and intense lightning activity cause the fault current waveform to exhibit complex characteristics such as multiple impacts and variable attenuation, which can easily lead to missegmentation or mismerization of fault events using traditional methods. At the same time, the unique electrical environment caused by high altitude alters the characteristics of the fault current, reducing the accuracy of fixed threshold discrimination methods.

[0004] In view of this, this application proposes a fault type identification method for transmission lines in plateau and mountainous areas, aiming to achieve reliable lightning strike identification in the special environment of plateau and mountainous areas. Summary of the Invention

[0005] The main purpose of this application is to provide a fault type identification method for transmission lines in high-altitude mountainous areas, aiming to solve the problem of how to reliably identify lightning strikes in the special environment of high-altitude mountainous areas.

[0006] To achieve the above objectives, this application provides a method for fault type identification of transmission lines in high-altitude mountainous areas, the method comprising:

[0007] S10, acquire the denoised time-domain current signal after a fault occurs in a power transmission line in a high-altitude mountainous area;

[0008] S20, multiple time windows are slidably extracted on the denoised time-domain current signal with a preset analysis time window and sliding step size. Candidate fault event segments are obtained by first-level merging based on the time interval of adjacent time windows and waveform similarity, and target fault event segments are obtained by second-level aggregation based on the time interval of adjacent candidate fault event segments.

[0009] S30, extract the fault component current in the target fault event segment, and calculate the average power ratio of the fault duration stage to the decay stage in the fault component current.

[0010] S40, determine the fault type based on the relationship between the average power ratio and the preset power ratio threshold, wherein the fault type includes lightning strike faults and non-lightning strike faults.

[0011] Optionally, in S20, candidate fault event segments are obtained by first-level merging based on the time interval of adjacent time windows and waveform similarity, and target fault event segments are obtained by second-level aggregation based on the time interval of adjacent candidate fault event segments, including:

[0012] If the time interval between adjacent time windows is less than the first threshold and the waveform similarity is greater than the preset similarity threshold, then the two adjacent time windows are merged into one candidate fault event segment.

[0013] If the time interval between any two adjacent candidate fault event segments is less than the second threshold, then the adjacent candidate fault event segments are aggregated into one target fault event segment.

[0014] The first threshold and the second threshold are related to the altitude of the transmission line in the plateau and mountainous area, and the second threshold is greater than the first threshold.

[0015] Optionally, the duration of the attenuation phase is directly proportional to the duration of the fault duration phase, and this proportionality is adjusted by a proportional coefficient that can adapt to the attenuation rate of the fault current.

[0016] Optionally, the average power P during the fault duration phase d and the average power P during the decay phase e The specific calculation formula is as follows:

[0017]

[0018]

[0019] Among them, T e T represents the duration of the decay phase. d R is the duration of the fault duration, R is the equivalent positive sequence resistance calculated based on the line parameters, t0 is the fault start time within the fault event segment, and i f (t) represents the fault component current.

[0020] Optionally, the preset power ratio threshold is determined based on the altitude of the location of the plateau mountain transmission line and the average number of thunderstorm days per year. The calculation expression for the preset power ratio threshold is as follows:

[0021]

[0022] Where h is the altitude of the location of the route; ρ storm ρ represents the average number of thunderstorm days per year in the region; γ is a correction factor greater than 1, whose value varies with h and ρ. storm It increases with the increase of .

[0023] Optionally, in S10, the denoising step of the denoised time-domain current signal includes:

[0024] The acquired raw current waveform is filtered using a low-pass filter;

[0025] A threshold denoising method based on discrete wavelet transform is used to perform secondary denoising on the filtered signal, wherein the threshold function in the threshold denoising method is:

[0026]

[0027] In the formula, w represents the wavelet decomposition coefficients, sign() is the sign function, and max() is the maximum value function; λ The threshold for wavelet denoising.

[0028] Optionally, after step S40, the following steps are also included:

[0029] S50, determine the coefficient of variation of the power sequence during the fault duration phase, and calculate the identification confidence level based on the coefficient of variation;

[0030] S60, when the identification confidence level is lower than the preset confidence threshold, adjust the analysis time window or the sliding step size, and return to step S20.

[0031] Optionally, the S50 includes:

[0032] During the fault duration phase, multiple time periods are divided with a fixed sub-window length. The average power of the fault component current in each time period is calculated to obtain the power sequence.

[0033] Calculate the coefficient of variation (CV) of the power sequence:

[0034]

[0035] In the formula, σ P and μ P These are the standard deviation and mean of the power sequence, respectively.

[0036] The identification confidence level C is calculated based on the coefficient of variation:

[0037] .

[0038] Optionally, in S60, adjusting the analysis time window or the sliding step size includes:

[0039] Determine the fluctuation characteristics of the power sequence within the current analysis window;

[0040] If the fluctuation characteristics indicate that the power sequence is incomplete and the high-frequency components are truncated, then the length of the analysis time window is increased by a preset time window step value, while keeping the sliding step size unchanged;

[0041] If the fluctuation characteristics indicate that the fault component current includes non-fault section noise, then the sliding step size is reduced by a preset sliding step value, while keeping the analysis time window unchanged.

[0042] In addition, to achieve the above objectives, this application also provides a computer system, the computer system comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, it implements the steps of the fault type identification method for high-altitude mountain transmission lines as described in any of the preceding claims.

[0043] In addition, to achieve the above objectives, this application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the fault type identification method for high-altitude mountain transmission lines as described in any of the preceding claims.

[0044] This application has at least the following beneficial effects:

[0045] 1. Fault event segments are extracted from the denoised fault current signal using a two-level aggregation strategy. For each fault event segment, the fault component current is extracted and the average power ratio between the fault duration stage and the attenuation stage is calculated. The fault type is determined by comparing the average power ratio with the discrimination threshold, thus realizing reliable fault type identification for transmission lines in plateau and mountainous areas.

[0046] 2. By simultaneously limiting the time interval and waveform similarity, it is ensured that the merged time windows are continuous in time and consistent in waveform shape, thereby eliminating the interference of random noise.

[0047] 3. By introducing an adjustable proportional coefficient, a dynamic proportional relationship is established between the duration of the decay phase and the duration of the fault duration.

[0048] 4. Set a correction factor that is related to altitude and average number of thunderstorm days per year. When the line is located in a high-altitude or thunderstorm-prone area, the correction factor becomes more adaptable, thereby raising the discrimination threshold. Conversely, the correction factor is reduced to the normal baseline value.

[0049] 5. By improving the Stein unbiased risk estimation criterion, the optimal threshold is searched by minimizing the mean square error risk, thus avoiding signal distortion or noise interference caused by excessive noise reduction.

[0050] 6. The coefficient of variation reflects the dispersion of power fluctuations during the fault duration. The smaller the fluctuation, the more stable the signal and the more reliable the feature extraction, and the higher the confidence level. Conversely, the larger the fluctuation, the lower the confidence level. When the confidence level is lower than the preset threshold, the signal features are re-adapted by changing the analysis time window or sliding step size, and the aggregation and feature extraction process is executed again to ensure that the output fault type determination result is reliable. Attached Figure Description

[0051] Figure 1 This is a flowchart illustrating the fault type identification method for high-altitude mountain transmission lines according to an embodiment of this application.

[0052] Figure 2 This is a waveform diagram of the current of a high-altitude mountain transmission line under a lightning strike fault, as described in the embodiments of this application.

[0053] Figure 3 This is a waveform diagram of the current of a high-altitude mountain transmission line under a non-lightning fault, as described in the embodiments of this application.

[0054] Figure 4 This is a diagram showing the original waveforms of the three-phase current during a lightning strike fault in a high-altitude mountain transmission line, as described in the embodiments of this application.

[0055] Figure 5 This is the original waveform diagram of the three-phase current during a common short-circuit fault in a high-altitude mountain transmission line according to an embodiment of this application;

[0056] Figure 6 This is a schematic diagram of the hardware operating environment of the computer system involved in the embodiments of this application.

[0057] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0058] To better understand the above technical solutions, exemplary embodiments of this disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of this disclosure to those skilled in the art.

[0059] First Embodiment

[0060] Reference Figure 1 This embodiment provides a method for identifying fault types in high-altitude mountain transmission lines, the method comprising:

[0061] S10, acquire the denoised time-domain current signal after a fault occurs in a power transmission line in a high-altitude mountainous area;

[0062] In this embodiment, the original current waveform is easily mixed with high-frequency electromagnetic interference, sampling noise and non-fault transient components. Directly using it for analysis will lead to feature extraction distortion. Therefore, in this embodiment, the three-phase current waveform after the fault of the transmission line in the plateau mountain area is first collected, and the current waveform is preprocessed to suppress noise and interference to obtain the denoised time-domain current signal, that is, the denoised time-domain current signal.

[0063] S20, multiple time windows are slidably extracted on the denoised time-domain current signal with a preset analysis time window and sliding step size. Candidate fault event segments are obtained by first-level merging based on the time interval of adjacent time windows and waveform similarity, and target fault event segments are obtained by second-level aggregation based on the time interval of adjacent candidate fault event segments.

[0064] After obtaining the denoised time-domain current signal, this step uses a two-level aggregation strategy to locate the fault boundary.

[0065] In the first-level merging, real fault pulses are distinguished from random noise spikes by time intervals and waveform morphology similarity. Pulses identified as real fault pulses are selected as candidate segments to prevent noise interference from contaminating the merging process. In the second-level merging, considering the common multi-impact lightning strikes or intermittent arc faults in high-altitude mountainous environments, multiple candidate segments belonging to the same physical fault process but potentially experiencing brief interruptions are merged into a single comprehensive target fault event segment based on time intervals. This avoids misclassifying independent events as single events and also prevents the missegmentation of multiple components of a single event into multiple independent events.

[0066] S30, extract the fault component current in the target fault event segment, and calculate the average power ratio of the fault duration stage to the decay stage in the fault component current.

[0067] In this embodiment, the fault component current refers to the pure fault information component obtained after removing the normal load component from the total current. Since lightning strikes typically manifest as high-energy transient impacts with energy concentrated in an extremely short duration and rapidly decaying, while other non-lightning short-circuit faults mostly exhibit steady-state continuous discharges with relatively balanced energy distribution across the two stages, this embodiment distinguishes between the fault duration stage and the decay stage from the fault component current. The fault duration stage is self-explanatory, while the decay stage refers to the dissipation and transition process of fault energy. Using the average power ratio of these two types of faults can quantify and distinguish their physical differences. Experiments have shown that this method has stronger anti-interference capabilities compared to simply relying on amplitude or frequency characteristics.

[0068] S40, determine the fault type based on the relationship between the average power ratio and the preset power ratio threshold, wherein the fault type includes lightning strike faults and non-lightning strike faults.

[0069] In some optional implementations, when the average power ratio is greater than a preset power ratio threshold, it means that the power difference between the fault duration stage and the attenuation stage is large, and it is determined to be a lightning strike fault. If the average power ratio is less than or equal to the preset power ratio threshold, it means that the power difference between the fault duration stage and the attenuation stage is small, and it is determined to be a non-lightning strike short circuit fault.

[0070] It should be noted that the preset power ratio threshold needs to be greater than 1.

[0071] Reference Figure 2 and Figure 3 The current waveforms of transmission lines in high-altitude mountainous areas under lightning strike faults and without lightning strike faults are shown respectively. It can be seen that the current waveform of a lightning strike fault exhibits typical high-energy transient impulses, with its amplitude reaching its peak within a very short fault duration and then rapidly decaying. There is a significant power difference between the fault duration and decay stages. In contrast, the current waveform of a non-lightning short-circuit fault mostly exhibits steady-state continuous discharge or approximate power frequency oscillations. Its amplitude changes are relatively gradual between the fault duration and decay stages, and the energy distribution is relatively balanced. These differences in the characteristics of the two types of faults during the duration and decay stages provide an intuitive physical basis for reliable identification using the average power ratio.

[0072] In the technical solution provided in this embodiment, a two-level aggregation strategy is used to extract fault event segments from the denoised fault current signal. For each fault event segment, the fault component current is extracted and the average power ratio between the fault duration stage and the attenuation stage is calculated. The fault type is determined by comparing the average power ratio with the discrimination threshold, thereby realizing reliable fault type identification under high-altitude mountain transmission lines.

[0073] Second Embodiment

[0074] Based on any embodiment, in this embodiment, in S20, candidate fault event segments are obtained by first-level merging based on the time interval of adjacent time windows and waveform similarity, and target fault event segments are obtained by second-level aggregation based on the time interval of adjacent candidate fault event segments, including:

[0075] S21, if the time interval between adjacent time windows is less than the first threshold and the waveform similarity is greater than the preset similarity threshold, then the two adjacent time windows are merged into one candidate fault event segment.

[0076] S22, if the time interval between any two adjacent candidate fault event segments is less than the second threshold, then the adjacent candidate fault event segments are aggregated into one target fault event segment;

[0077] The first threshold and the second threshold are related to the altitude of the transmission line in the plateau and mountainous area, and the second threshold is greater than the first threshold.

[0078] In this embodiment, in the fault waveform recording data of transmission lines in high-altitude mountainous areas, in addition to the actual fault transient pulses, there are often random noise spikes caused by corona discharge, switching operations, or electromagnetic interference. Although these noise spikes may be temporally adjacent to the fault pulses, their waveforms are random and non-periodic, showing a significant difference from the inherent oscillation mode of the fault current. Therefore, if merging is based solely on time intervals, noise spikes are easily mistakenly included in the fault event segment, leading to contamination of subsequent feature extraction. Therefore, in this embodiment, by simultaneously limiting the time interval to less than a first threshold and the waveform similarity to a preset similarity threshold, it is ensured that the merged time windows are continuous in time and consistent in waveform morphology, thereby eliminating the interference of random noise.

[0079] In some alternative implementations, the first threshold τ1 is set to a range of 1ms to 10ms, which is determined based on the minimum typical interval between adjacent impact events in the fault current waveform of the transmission line in the plateau mountainous area; the similarity threshold ρ is set to 0.8 to ensure that only adjacent time windows with highly similar waveform morphology are merged.

[0080] In some alternative implementations, the second threshold τ2 ranges from 5ms to 50ms, which is set based on the statistical characteristics of typical time intervals between multiple lightning strikes or consecutive fault events in high-altitude mountain transmission lines.

[0081] In this embodiment, the first threshold and the second threshold are related to the altitude of the transmission line in the high-altitude mountainous area. In some alternative embodiments, for areas with an altitude higher than 3000 meters, τ2 takes a smaller value within the range, which can be a lower limit value, to adapt to the more frequent lightning strikes and faster waveform attenuation characteristics in high-altitude areas.

[0082] Third Embodiment

[0083] Based on any embodiment, in this embodiment, the duration of the attenuation phase is directly proportional to the duration of the fault duration phase, and the direct proportionality is adjusted by a proportional coefficient that can adapt to the attenuation rate of the fault current.

[0084] Specifically, the fault current is not a homogeneous process from occurrence to termination, but rather includes a sustained fault phase characterized by a rapid release of energy, and a decay phase characterized by the gradual dissipation of energy. It's important to note that the duration of these two phases is not fixed, but depends on the fault type and the environmental conditions of the line. For example, the low air density and poor heat dissipation in high-altitude mountainous environments can lead to differences in the decay characteristics of the fault current compared to plains areas. Using a fixed time window to capture these two phases can easily result in a mismatch between the captured range and the actual physical process, thus causing the calculated average power ratio to lose its discriminative power.

[0085] Therefore, this embodiment establishes a dynamic proportional relationship between the duration of the decay phase and the duration of the fault duration by introducing an adjustable proportional coefficient k.

[0086] In some alternative implementations, the proportionality coefficient k is preferably in the range of 0.8 to 1.4. When the identified object tends to be a rapidly decaying lightning fault, k can take a smaller value; when the identified object tends to be a slowly decaying short-circuit fault or is in a special environment, k can take a larger value.

[0087] For example, the duration T of the fault duration phase d Duration T of the decay phase e The following relationship must be satisfied:

[0088]

[0089] In the formula, k is an adjustable proportional coefficient in the range of 0.8 to 1.4, which is used to adaptively adjust the observation window of the attenuation stage according to the specific electrical environment and historical fault data of the plateau and mountainous areas where the line is located, so as to match the attenuation rate of the local fault current.

[0090] Furthermore, the average power P during the fault duration phase d and the average power P during the decay phase e The specific calculation formula is as follows:

[0091]

[0092]

[0093] Among them, T e T represents the duration of the decay phase. d R is the duration of the fault duration, R is the equivalent positive sequence resistance calculated based on the line parameters, t0 is the fault start time within the fault event segment, and i f (t) represents the fault component current.

[0094] It should be noted that for lightning strike faults, the physical process is that the lightning wave propagates along the line, causing instantaneous overvoltage breakdown. The energy is concentrated in the microsecond to millisecond range and is released explosively. Therefore, the average power value is relatively large during the fault duration phase. In the subsequent phase, due to the rapid attenuation of the lightning wave and the possibility of insulation recovery, the current drops sharply, resulting in a smaller average power during the attenuation phase. Therefore, under ideal conditions, the final calculated average power ratio under lightning strike faults is greater than 1.

[0095] For non-lightning short-circuit faults, the physical process is usually continuous contact or arc discharge between conductors. The energy is relatively evenly distributed in the two stages. In some metallic short-circuit cases, the current amplitude in stage one is the same as that in stage two. Therefore, the average power values ​​of the two are close, and the calculated average power ratio will approach 1.

[0096] Fourth embodiment

[0097] Based on any embodiment, in this embodiment, the preset power ratio threshold is determined based on the altitude of the location of the plateau mountain transmission line and the average number of thunderstorm days per year. The calculation expression for the preset power ratio threshold is:

[0098]

[0099] Where h is the altitude of the location of the route; ρ storm ρ represents the average number of thunderstorm days per year in the region; γ is a correction factor greater than 1, whose value varies with h and ρ. storm It increases with the increase of .

[0100] In this embodiment, the baseline value for the preset power ratio threshold is 2, which is an empirical value. Based on this, a correction coefficient γ greater than 1 is introduced, the value of which varies with h and ρ. storm It increases with the increase of .

[0101] It should be noted that in high-altitude and mountainous areas, air density and air pressure decrease significantly with increasing altitude, leading to a decrease in the external insulation strength of transmission lines and a reduction in lightning impulse discharge voltage. This electrical characteristic causes the fault current waveform to exhibit distinctly different features compared to plains areas: the same lightning overvoltage can easily cause a shift in the transient energy distribution of lightning-induced faults at high altitudes. In this case, if the fixed benchmark value from plains areas is continued to be used as the discrimination threshold, non-lightning faults with high transient components are easily misclassified as lightning faults, or faults with relatively weak energy but definitely caused by lightning are missed as non-lightning faults. Therefore, in this embodiment, a correction factor related to altitude and the average number of thunderstorm days per year is set for this scenario. When the line is located at high altitude or in areas with frequent thunderstorms, the adaptability of the correction factor increases, thereby raising the discrimination threshold; conversely, the correction factor is reduced to the conventional benchmark value.

[0102] Fifth Embodiment

[0103] Based on any embodiment, in this embodiment, the denoising step of the denoised time-domain current signal includes:

[0104] S11, a low-pass filter is used to filter the acquired raw current waveform;

[0105] S12, a second denoising process is performed on the filtered signal using a threshold denoising method based on discrete wavelet transform, wherein the threshold function in the threshold denoising method is:

[0106]

[0107] In the formula, w represents the wavelet decomposition coefficients, sign() is the sign function, and max() is the maximum value function; λ The threshold for wavelet denoising.

[0108] In some alternative implementations, the low-pass filter is a Butterworth low-pass filter, which has the largest flat passband amplitude response characteristics. It can effectively filter out high-frequency electromagnetic interference, carrier communication signals and high-frequency noise generated by switching operations introduced during the sampling process, while preserving the fundamental frequency and main harmonic components of the fault current to the greatest extent and preventing waveform distortion caused by the phase nonlinearity of the filter.

[0109] In some optional implementations, the db4 wavelet basis is further selected, and the decomposition layer is 5 layers for secondary denoising in step S12. The db4 wavelet basis can match the oscillation attenuation characteristics of power system transient signals and can efficiently and sparsely represent fault pulses in the time and frequency domain; while the 5-layer decomposition divides the signal into multiple frequency bands, so that the noise energy is mainly concentrated in the high-level detail coefficients, while the fault features are distributed in coefficients of a specific scale.

[0110] Compared to the traditional VisuShrink or HeurSure criteria, the improved Stein unbiased risk estimation criterion in this embodiment searches for the optimal threshold by minimizing the mean square error risk, thus avoiding signal distortion or noise interference caused by excessive noise reduction.

[0111] Sixth Embodiment

[0112] Based on any embodiment, in this embodiment, after step S40, the method further includes:

[0113] S50, determine the coefficient of variation of the power sequence during the fault duration phase, and calculate the identification confidence level based on the coefficient of variation;

[0114] S60, when the identification confidence level is lower than the preset confidence threshold, adjust the analysis time window or the sliding step size, and return to step S20.

[0115] This embodiment further includes a reliability verification of the identification results. Specifically, the coefficient of variation reflects the dispersion of power fluctuations during the fault duration. Smaller fluctuations indicate a more stable signal, more reliable feature extraction, and higher confidence. When the confidence is below a preset threshold, it means that the current instantaneous window length or sliding step may not be able to effectively capture the features of the current fault signal. In this case, the signal features are re-adapted by changing the analysis window or sliding step, and the aggregation and feature extraction process in step S20 is re-executed to ensure that the output fault type determination result is reliable.

[0116] Further and optionally, the S50 includes:

[0117] S51, during the fault duration phase, multiple time periods are divided with a fixed sub-window length, and the average power of the fault component current in each time period is calculated to obtain the power sequence.

[0118] S52, calculate the coefficient of variation (CV) of the power sequence:

[0119]

[0120] In the formula, σ P and μ P These are the standard deviation and mean of the power sequence, respectively.

[0121] S53, Calculate the identification confidence level C based on the coefficient of variation:

[0122] .

[0123] It should be noted that the fault current typically exhibits a steady-state characteristic of the power frequency during the fault duration, at which point the power sequence fluctuation is relatively small, and the standard deviation σ is... P Relative to the mean μ P The power sequence is relatively small, hence the CV value is low; when the signal is subjected to strong random electromagnetic interference or noise pollution, the power sequence will exhibit abnormal and drastic jumps, leading to an increase in the standard deviation σ. P This increases significantly, leading to a sharp rise in the CV value.

[0124] In this embodiment, the closer the obtained identification confidence C value is to 1, the more stable the power sequence is, the less the extracted features are affected by interference, and the more reliable the identification result is; conversely, the lower the value, the worse the current signal quality is or the event segmentation may be biased, and the identification result is questionable. In some optional embodiments, the preset confidence threshold can be 0.75, and an identification confidence C value greater than or equal to 0.75 is regarded as a reliable result, while the execution condition of step S60 is met otherwise.

[0125] It should be noted that when the confidence level of a particular identification is lower than the confidence threshold, it means that the current time window and / or sliding step size are not suitable for the current signal characteristics. For example, if the analysis time window is set too short, the high-frequency lightning pulse may be truncated, resulting in an incomplete and highly fluctuating power sequence; if the sliding step size is set too large, it is easy to introduce noise from non-faulty sections.

[0126] Therefore, when the confidence level is too low, this embodiment attempts to increase / decrease the analysis window and / or sliding step size to capture a more complete transient process.

[0127] Further and optionally, in S60, adjusting the analysis window or the sliding step size includes:

[0128] Determine the fluctuation characteristics of the power sequence within the current analysis window;

[0129] If the fluctuation characteristics indicate that the power sequence is incomplete and the high-frequency components are truncated, then the length of the analysis time window is increased by a preset time window step value, while keeping the sliding step size unchanged;

[0130] If the fluctuation characteristics indicate that the fault component current includes non-fault section noise, then the sliding step size is reduced by a preset sliding step value, while keeping the analysis time window unchanged.

[0131] Verification Example 1

[0132] Based on the first to sixth embodiments, refer to Figure 4 In this embodiment, on a 110kV transmission line in a plateau region, a fault recording device recorded a segment of fault current data caused by a lightning strike, with a sampling frequency of 1MHz. This data segment captured a complex sequence of lightning strike events: the A-phase conductor was struck by lightning three times within a 380ms time window, and each lightning strike was an independent discharge process, forming multiple transient impacts. Identification was performed according to the method proposed in the aforementioned embodiment, with the specific steps as follows:

[0133] (1) Input the collected fault current data into the system. First, use a Butterworth low-pass filter with a cutoff frequency of 500Hz for preliminary filtering to remove high-frequency sampling noise. Then, use a 5-level discrete wavelet transform based on the db4 wavelet basis, calculate the adaptive threshold using the improved Stein unbiased risk estimation criterion, and apply a soft threshold function for fine denoising to finally obtain the denoised time-domain current signal s(t).

[0134] (2) First, set the core algorithm parameters: analysis window length W=5ms, sliding step Δ=1ms, first-level merging time threshold τ1=3ms, window waveform similarity threshold ρ=0.8, and second-level aggregation time threshold τ2=15ms.

[0135] Then, a sliding scan was performed from the fault initiation point, and three groups of transient pulses with sudden energy increases were captured successively on the A-phase current waveform. For each pulse group, the multiple continuous time windows contained within it were merged into three candidate fault event segments, denoted as E1, E2, and E3, because the time interval was less than τ1 and the waveforms were highly similar.

[0136] Finally, the time interval between adjacent candidate event segments is calculated to obtain the interval ΔT between E1 and E2. 12 =226.35ms and the time interval ΔT between E2 and E3 23 =108.66ms. Actual measurement data shows that ΔT 12 With ΔT 23 All three values ​​are significantly greater than the second-level aggregation threshold τ2. Therefore, the algorithm determines that E1, E2, and E3 are three temporally independent and complete lightning strike fault event segments, and does not perform any cross-event aggregation, thus successfully separating the three lightning strikes.

[0137] (3) Set the scaling factor k to 1.1, and perform the same feature extraction process in parallel for the three independent lightning strike event segments E1, E2, and E3 separated in step (2). For any event segment, first extract the fault component current i from its current data. f (t); subsequently, the duration T of the fault duration is automatically identified. d Duration T of the decay phase e Finally, the average power P of the two stages is calculated separately. e P d The average power ratio (APR) was calculated. The core characteristic quantities for the three time periods were APR1≈71.56, APR2≈362.54, and APR3≈109.25, respectively.

[0138] (4) Based on the altitude h=3.5km and the average number of thunderstorm days per year ρ at the location of the route. storm =65 days, and an altitude-climate correction factor γ≈1.3 was obtained by matching from a preset localized parameter library. The unified dynamic discrimination threshold Th in this embodiment was calculated. adjusted =2×γ=2×1.3=2.6. Comparing the APR values ​​of each event segment with this threshold, we find that: APR1, APR2, APR3 > Th. adjusted .

[0139] (5) For each event segment E1, E2, and E3, during its fault duration T, respectively e The coefficient of variation (CV) of the power sequence was calculated, and the identification confidence level (C) was further obtained. The calculation results were: C1=0.89, C2=0.97, and C3=0.85, all of which were higher than the preset reliability threshold C. th =0.75.

[0140] Conclusion: Based on the above analysis, the APR values ​​of the three independent event segments E1, E2, and E3 are all significantly higher than the dynamic discrimination threshold, and the confidence assessments all meet the reliability requirements. Therefore, the system accurately outputs the identification result of "lightning strike fault" for all three events. This embodiment demonstrates that the method proposed in this application can effectively separate, extract features, and reliably identify multiple independent lightning strike events occurring consecutively within a short period of time, possessing a powerful ability to handle complex fault sequences.

[0141] Verification Example 2

[0142] Based on the first to sixth embodiments, refer to Figure 5 On a 110kV transmission line in a plateau region, a fault recording device recorded a segment of short-circuit fault current data with a sampling frequency of 1MHz. This data segment captures the current waveform 300ms after the fault, exhibiting typical continuous short-circuit characteristics. Identification was performed according to the method proposed in the aforementioned embodiment, with the specific steps as follows:

[0143] (1) Input the collected fault current data into the system. First, use a Butterworth low-pass filter with a cutoff frequency of 500Hz for preliminary filtering to remove high-frequency sampling noise. Then, use a 5-level discrete wavelet transform based on the db4 wavelet basis, calculate the adaptive threshold using the improved Stein unbiased risk estimation criterion, and apply a soft threshold function for fine denoising to finally obtain the denoised time-domain current signal s(t).

[0144] (2) First, set the core algorithm parameters: analysis window length W=5ms, sliding step Δ=1ms, first-level merging time threshold τ1=3ms, window waveform similarity threshold ρ=0.8, and second-level aggregation time threshold τ2=15ms.

[0145] Then, a sliding scan is performed from the fault initiation point to capture a group of transient pulses with sudden energy increases on the A-phase current waveform. For each pulse group, the multiple consecutive time windows contained within it are merged into a candidate fault event segment, denoted as F1, because the time interval is less than τ1 and the waveforms are highly similar.

[0146] (3) Set the scaling factor k to 1.1, and perform the feature extraction process in parallel for the independent fault event segment F1 separated in step (2). For this event segment, first extract the fault component current i from its current data. f (t); subsequently, the duration T of the decay phase is automatically identified. e Duration T of the fault duration d Finally, the average power P of the two stages is calculated separately. e P dThe average power ratio (APR) was calculated, and the core characteristic parameters for this time period were APR≈1.12.

[0147] (4) Based on the altitude h=3.2km and the average number of thunderstorm days per year ρ at the location of the route. storm =65 days, and an altitude-climate correction factor γ≈1.25 was obtained by matching from a preset localized parameter library. The unified dynamic discrimination threshold Th in this embodiment was calculated. adjusted =2×γ=2×1.25=2.5. Comparing the APR value of this event segment with this threshold, we have: APR>Th adjusted .

[0148] (5) For event segment F1, during its fault duration T e The coefficient of variation (CV) of the power sequence was calculated, and the identification confidence level (C) was further obtained. The calculation results were: C = 0.95, which is significantly higher than the preset reliability threshold C. th =0.75.

[0149] Conclusion: Based on the above analysis, the average power ratio (APR) of short-circuit fault event segment F1 is approximately 1.12, significantly close to 1 and far below the dynamic discrimination threshold of 2.5. Furthermore, the confidence level of this identification is as high as 0.95, meeting the reliability requirements. Therefore, the system accurately outputs the identification result of "non-lightning strike short-circuit fault". This embodiment demonstrates that the method proposed in this application can accurately extract features and perform threshold discrimination for single persistent non-lightning strike short-circuit faults, and ensures the reliability of the results through confidence level assessment, showcasing a stable identification capability for typical short-circuit faults under complex operating conditions.

[0150] As one implementation scheme, Figure 6 This is a schematic diagram of the hardware operating environment of the computer system involved in the embodiments of this application.

[0151] like Figure 6 As shown, the computer system may include: a processor 1001, such as a CPU; a memory 1005; a user interface 1003; a network interface 1004; and a communication bus 1002. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display screen or an input unit such as a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be high-speed RAM or non-volatile memory, such as a disk drive. Optionally, the memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0152] Those skilled in the art will understand that Figure 6 The computer system architecture shown does not constitute a limitation on the computer system and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0153] like Figure 6 As shown, the memory 1005, as a storage medium, may include an operating system, a network communication module, a user interface module, and computer programs. The operating system is a program that manages and controls the hardware and software resources of the computer system, as well as the operation of the computer programs and other software or programs.

[0154] exist Figure 6 In the computer system shown, the user interface 1003 is mainly used to connect to the terminal and communicate with the terminal; the network interface 1004 is mainly used to communicate with the backend server; and the processor 1001 can be used to call the computer program stored in the memory 1005.

[0155] In this embodiment, the computer system includes: a memory 1005, a processor 1001, and a computer program stored in the memory and executable on the processor, wherein:

[0156] When the processor 1001 calls the computer program stored in the memory 1005, it implements each step of the fault type identification method for high-altitude mountain transmission lines as described in the above embodiment.

[0157] Furthermore, those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program includes program instructions and can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in a computer system to implement the process steps of the embodiments of the above methods.

[0158] Therefore, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the various steps of the fault type identification method for high-altitude mountain transmission lines as described in the above embodiments.

[0159] The computer-readable storage medium can be any computer-readable storage medium capable of storing program code, such as a USB flash drive, portable hard drive, read-only memory (ROM), magnetic disk, or optical disk.

[0160] It should be noted that, since the storage medium provided in the embodiments of this application is the storage medium used to implement the methods of the embodiments of this application, those skilled in the art can understand the specific structure and variations of the storage medium based on the methods described in the embodiments of this application, and therefore will not be repeated here. All storage media used in the methods of the embodiments of this application fall within the scope of protection of this application.

[0161] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0162] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0163] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0164] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0165] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0166] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for identifying fault types in high-altitude mountain transmission lines, characterized in that, The method includes: S10, acquire the denoised time-domain current signal after a fault occurs in a power transmission line in a high-altitude mountainous area; S20, multiple time windows are slidably extracted on the denoised time-domain current signal with a preset analysis time window and sliding step size. Candidate fault event segments are obtained by first-level merging based on the time interval of adjacent time windows and waveform similarity, and target fault event segments are obtained by second-level aggregation based on the time interval of adjacent candidate fault event segments. S30, extract the fault component current in the target fault event segment, and calculate the average power ratio of the fault duration stage to the decay stage in the fault component current. S40, determine the fault type based on the relationship between the average power ratio and the preset power ratio threshold, wherein the fault type includes lightning strike faults and non-lightning strike faults.

2. The fault type identification method for high-altitude mountain transmission lines as described in claim 1, characterized in that, In S20, candidate fault event segments are obtained through primary merging based on the time intervals of adjacent time windows and waveform similarity, and target fault event segments are obtained through secondary aggregation based on the time intervals of adjacent candidate fault event segments, including: If the time interval between adjacent time windows is less than the first threshold and the waveform similarity is greater than the preset similarity threshold, then the two adjacent time windows are merged into one candidate fault event segment. If the time interval between any two adjacent candidate fault event segments is less than the second threshold, then the adjacent candidate fault event segments are aggregated into one target fault event segment. The first threshold and the second threshold are related to the altitude of the transmission line in the plateau and mountainous area, and the second threshold is greater than the first threshold.

3. The fault type identification method for high-altitude mountain transmission lines as described in claim 1, characterized in that, The duration of the attenuation phase is directly proportional to the duration of the fault duration phase, and this proportionality is adjusted by a proportional coefficient that can adapt to the attenuation rate of the fault current.

4. The fault type identification method for high-altitude mountain transmission lines as described in claim 1 or 3, characterized in that, Average power P during the fault duration phase d and the average power P during the decay phase e The specific calculation formula is as follows: ； ； Among them, T e T represents the duration of the decay phase. d R is the duration of the fault duration, R is the equivalent positive sequence resistance calculated based on the line parameters, t0 is the fault start time within the fault event segment, and i f (t) represents the fault component current.

5. The fault type identification method for high-altitude mountain transmission lines as described in claim 1, characterized in that, The preset power ratio threshold is determined based on the altitude of the location of the plateau mountain transmission line and the average number of thunderstorm days per year. The calculation expression for the preset power ratio threshold is as follows: ； Where h is the altitude of the location of the route; ρ storm ρ represents the average number of thunderstorm days per year in the region; γ is a correction factor greater than 1, whose value varies with h and ρ. storm It increases with the increase of.

6. The fault type identification method for high-altitude mountain transmission lines as described in claim 1, characterized in that, In S10, the denoising step of the denoised time-domain current signal includes: The acquired raw current waveform is filtered using a low-pass filter; A threshold denoising method based on discrete wavelet transform is used to perform secondary denoising on the filtered signal, wherein the threshold function in the threshold denoising method is: ； In the formula, w represents the wavelet decomposition coefficients, sign() is the sign function, and max() is the maximum value function; λ The threshold for wavelet denoising.

7. The fault type identification method for high-altitude mountain transmission lines as described in claim 1, characterized in that, After step S40, the following is also included: S50, determine the coefficient of variation of the power sequence during the fault duration phase, and calculate the identification confidence level based on the coefficient of variation; S60, when the identification confidence level is lower than the preset confidence threshold, adjust the analysis time window or the sliding step size, and return to step S20.

8. The fault type identification method for high-altitude mountain transmission lines as described in claim 7, characterized in that, The S50 includes: During the fault duration phase, multiple time periods are divided with a fixed sub-window length. The average power of the fault component current in each time period is calculated to obtain the power sequence. Calculate the coefficient of variation (CV) of the power sequence: ； In the formula, σ P and μ P These are the standard deviation and mean of the power sequence, respectively. The identification confidence level C is calculated based on the coefficient of variation: 。 9. The fault type identification method for high-altitude mountain transmission lines according to claim 6 or 7, characterized in that, In S60, adjusting the analysis time window or the sliding step size includes: Determine the fluctuation characteristics of the power sequence within the current analysis window; If the fluctuation characteristics indicate that the power sequence is incomplete and the high-frequency components are truncated, then the length of the analysis time window is increased by a preset time window step value, while keeping the sliding step size unchanged; If the fluctuation characteristics indicate that the fault component current includes non-fault section noise, then the sliding step size is reduced by a preset sliding step value, while keeping the analysis time window unchanged.

10. A computer system, characterized in that, The computer system includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the steps of the fault type identification method for high-altitude mountain transmission lines as described in any one of claims 1 to 9.

Images