Method for judging series arc fault of low-voltage distribution cable based on disturbance current characteristics

Abstract

The invention belongs to the technical field of low-voltage distribution cable fault identification, and particularly relates to a method for judging a series arc fault of a low-voltage distribution cable based on disturbance current characteristics, which comprises the following steps of: S1, acquiring current data of three-phase current of the cable; s2, analyzing the similarity of the current waveforms of the adjacent current periods by adopting a Pearson correlation coefficient PCC, and judging whether the cable has current disturbance or not; s3, when it is judged that current disturbance occurs, intercepting current data of a disturbance starting point and a plurality of current periods after the disturbance starting point, and recording the current data as disturbance current; calculating the time-frequency domain characteristic quantity of the intercepted disturbance current; and S4, inputting the time-frequency domain characteristic quantity of the disturbance current into a preset fault identification model, carrying out detection and identification on a cable series arc fault, and judging whether the fault causing current disturbance is the series arc fault or not. According to the method, the low-voltage distribution cable series arc fault can be efficiently, accurately and timely monitored and identified, and the safety and stability of a power system are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of low-voltage distribution cable fault identification, and in particular relates to a method for judging a series arc fault in a low-voltage distribution cable based on disturbance current characteristics. Background Art

[0002] In the power system, the safe operation of low-voltage distribution cables is of vital importance. However, in actual operation, cables are often affected by various factors, resulting in faults, among which series arc faults are a more common type of fault. Long-term low-voltage cables show obvious signs of aging. Cable insulation degradation and poor contact can easily lead to series arc faults. When a series arc fault occurs, a large amount of heat is released, which may ignite nearby combustibles. The resulting electrical fire will pose a serious threat to the power system, equipment and personnel safety. If this fault is not discovered and handled in time, it may pose a serious threat to the stable operation of the power system, and even cause serious consequences such as equipment damage or fire. Therefore, once a series arc fault occurs in a low-voltage cable, staff are required to discover and troubleshoot the fault in time to avoid greater losses.

[0003] At present, arc and ultrasonic are the two main methods used for online monitoring of arc in distribution network cables. The arc monitoring method determines whether arc partial discharge occurs by detecting whether the cable emits abnormal arc light, while the ultrasonic monitoring method determines by detecting the ultrasonic signal generated by arc discharge. However, these two methods have some obvious problems in practical application.

[0004] First, the coverage of arc and ultrasonic monitoring methods is relatively small, and it is difficult to achieve comprehensive monitoring of the entire distribution network cable. This may cause some potential arc faults to be missed, thereby increasing the operation risk of the power system. Secondly, the judgment accuracy of these two methods is low. Since arc and ultrasonic signals are easily interfered by the external environment, such as electromagnetic noise, mechanical vibration, etc., false alarms or missed alarms often occur in practical applications. This will not only reduce the reliability of the monitoring system, but also increase the workload of operation and maintenance personnel. In addition, the detection timeliness of arc and ultrasonic monitoring methods is low, and they can only be detected when the series arc fault of the cable is already very obvious. This means that when the monitoring system issues an alarm, the fault has often developed to a more serious level, which is easy to cause the fault to further expand and the harm to increase.

[0005] Therefore, the existing online monitoring methods for distribution network cable arcs have problems such as small coverage, low judgment accuracy and low detection timeliness. How to efficiently, accurately and timely monitor and identify series arc faults in low-voltage distribution cables and improve the safety and stability of the power system has become a problem that needs to be solved urgently. Summary of the invention

[0006] In view of the deficiencies of the above-mentioned prior art, the present invention provides a method for judging series arc faults in low-voltage distribution cables based on disturbance current characteristics, which can efficiently, accurately and timely monitor and identify series arc faults in low-voltage distribution cables, thereby improving the safety and stability of the power system.

[0007] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0008] The method for judging a series arc fault in a low-voltage distribution cable based on disturbance current characteristics comprises the following steps:

[0009] S1, real-time acquisition of current data of three-phase current of low-voltage distribution network cables;

[0010] S2. For the acquired current data, the Pearson correlation coefficient PCC is used to analyze the similarity of the current waveforms of each adjacent current cycle, and determine whether the cable has a current disturbance;

[0011] S3. When it is determined that a current disturbance occurs, the current cycle where the current disturbance occurs is taken as the disturbance starting point, and the current data of the disturbance starting point and multiple current cycles thereafter are intercepted and recorded as the disturbance current; and the time-frequency domain characteristic quantity of the intercepted disturbance current is calculated;

[0012] S4. Input the time-frequency domain characteristic quantity of the disturbance current into a preset fault identification model to detect and identify the cable series arc fault, and determine whether the fault causing the current disturbance is a series arc fault.

[0013] Compared with the prior art, the present invention has the following beneficial effects:

[0014] 1. Improve the accuracy of monitoring and identification. By using the Pearson correlation coefficient PCC to analyze the similarity of current waveforms in each adjacent current cycle, this method can preliminarily determine whether a current disturbance has occurred in the cable. This method effectively captures the slight changes in the current waveform, so that it can accurately screen out the current suspected of causing a cable series arc fault. Compared with traditional methods, this solution improves the accuracy of monitoring and identifying cable fault arcs in the initial stage, laying a solid foundation for subsequent analysis and judgment.

[0015] 2. Ensure the timeliness of series arc fault identification. This method analyzes the current data of the starting point of the disturbance current and multiple cycles thereafter, and can quickly capture the initial characteristics of the arc fault. Since arc faults are often weak in the early stages, but gradually intensify over time, timely identification and measures are crucial to prevent the expansion of faults and protect equipment and personnel safety. Through this method, accurate identification can be performed in the early stages of the development of arc faults, thereby ensuring the timeliness of series arc fault identification and providing a valuable time window for subsequent fault handling.

[0016] 3. Reduce the amount of current data to be identified and reduce monitoring delay. After confirming that a current disturbance has occurred, the solution only captures data from multiple current cycles after the starting point of the disturbance for further analysis. This approach avoids the processing of a large amount of irrelevant data and significantly reduces the amount of current data that needs to be calculated. At the same time, since only the data after the disturbance is processed, the time of the entire monitoring and identification process is greatly shortened, thereby reducing the monitoring delay and improving the real-time response capability of the system.

[0017] 4. Ensure the accuracy of identification. When a series arc fault occurs in the low-voltage distribution network cable, the current data will show obvious time-frequency domain feature changes such as an increase in high-frequency components and a decrease in periodic characteristics. This method calculates the time-frequency domain feature quantities of the disturbance current and inputs them into the preset fault identification model, which can make full use of these feature changes to accurately determine the fault type. This method not only takes into account the time domain characteristics of the current, but also combines the frequency domain characteristics, thereby ensuring the accuracy and reliability of cable series arc fault identification.

[0018] This method realizes efficient, accurate and timely monitoring and identification of series arc faults in low-voltage distribution cables by comprehensively using Pearson correlation coefficient analysis, time-frequency domain feature quantity calculation and fault identification model, effectively improving the safety and stability of the power system.

[0019] Preferably, in S2, when judging whether the cable has a current disturbance, the current disturbance is judged in combination with the cumulative sum algorithm CUSUM based on the double sliding windows W1 and W2; wherein W1 is a front window, used to obtain the distribution expected value E and determine the upper threshold UB and the lower threshold LB; the expected value E is the average value calculated based on the current waveform data distribution in the front window W1; the values ​​of the upper threshold UB and the lower threshold LB are set based on historical data; W2 is a rear window, used to calculate the offset accumulation value C of the data sequence in the window; the window lengths of the two windows are both M current cycles;

[0020] Set the initial accumulation value of the positive offset and negative offset initial accumulation value are all 0; after each sliding displacement of the window, based on the expected value E and the thresholds UB and LB, the offset of each point data in W2 is calculated in turn, and the accumulated value of the positive offset is and negative offset initial accumulation value Iterate; the iterative process is as follows:

[0021]

[0022] In the formula, x k is the value of the Pearson correlation coefficient of the kth current cycle in W2; when x k When it is greater than the upper threshold E+UB, the positive offset accumulation value Increase; when x k When it is less than the lower threshold E-LB, the negative offset accumulation value reduce;

[0023] If satisfied or It is determined that a disturbance has occurred in the current in the window; wherein DL is a preset offset limit.

[0024] Such a setting can 1. improve the sensitivity of disturbance detection. By setting the front window, the technology can obtain the expected value of the distribution of the current waveform data, and determine the upper and lower thresholds based on this. These thresholds are set based on historical data and can reflect the fluctuation range of the normal current waveform. When the data in the rear window deviates from these thresholds, it indicates that the current may have been disturbed. Compared with a single threshold judgment, this method can capture more subtle current changes, thereby improving the sensitivity of disturbance detection.

[0025] 2. It can enhance the accuracy of disturbance identification. The post-window is used to calculate the offset accumulation value of the data sequence in the window. By iterating the positive offset accumulation value and the negative offset accumulation value, the offset of the current waveform data relative to the expected value can be accumulated. When the offset accumulation value reaches the preset offset limit, it can be determined that the current has undergone a disturbance change. This method can more accurately reflect the overall change trend of the current waveform by accumulating the offset, thereby enhancing the accuracy of disturbance identification.

[0026] 3. Strong adaptability, suitable for a variety of scenarios. By adjusting the window length, threshold, offset limit and other parameters of the front window and the rear window, it can adapt to the current disturbance detection needs under different cable types and different operating conditions. This makes the technology highly adaptable and can be widely used in various power system monitoring and fault diagnosis scenarios.

[0027] Preferably, in S2, when the Pearson correlation coefficient PCC is used to analyze the similarity of the current waveforms of each adjacent current cycle, the calculation formula of the Pearson correlation coefficient of each current cycle is:

[0028]

[0029] Where PCC m represents the Pearson correlation coefficient of the mth current cycle; represents the kth instantaneous current value in the mth current cycle, represents the current mean value of the mth cycle; K is the number of instantaneous current values ​​in one current cycle.

[0030] With this setting, the Pearson correlation coefficient is a statistical indicator that measures the degree of linear correlation between two variables, and its value ranges from -1 to 1. When it is applied to analyze the similarity of current waveforms in adjacent current cycles, it can accurately capture the subtle changes in the current waveform between cycles. This change may reflect potential faults or abnormal conditions in the cable, such as series arc faults. When the current waveform changes significantly, the Pearson correlation coefficient will also change significantly accordingly. This change provides a sensitive indicator for fault detection, allowing the system to detect potential faults earlier, thereby improving the sensitivity of fault detection. In subsequent analysis, the Pearson correlation coefficient can be used as an important basis for determining whether a current disturbance occurs in the cable. When the Pearson correlation coefficient of adjacent current cycles is lower than a preset threshold, it can be considered that the current waveform has changed significantly, and then it can be determined that a fault may have occurred in the cable. This provides strong data support for fault identification.

[0031] Preferably, in S3, when calculating the time-frequency domain characteristics of the intercepted disturbance current, the calculated time-domain characteristics include zero-rest time, peak-to-peak value, standard deviation and kurtosis; the calculated frequency-domain characteristics include harmonic amplitude, spectrum centroid and spectrum standard deviation.

[0032] This setting can fully reflect the characteristics of series arc faults in cables by calculating the time domain characteristics (zero-off time, peak-to-peak value, standard deviation, kurtosis) and frequency domain characteristics (harmonic amplitude, spectrum centroid, spectrum standard deviation) of the disturbance current, and provide strong data support for the detection and diagnosis of series arc faults in cables. These characteristics are not only intuitive, but also easy to calculate and extract, and have broad application prospects in power system monitoring and fault diagnosis.

[0033] Preferably, when calculating the time domain feature quantity, the peak-to-peak value is calculated as follows:

[0034]

[0035] PPV k represents the normalized peak-to-peak value of the kth current cycle, and Indicates the maximum and minimum values ​​of the current collected in the kth current cycle.

[0036] With such a setting, when an arc fault occurs, due to the nonlinear characteristics and instability of the arc, the current waveform tends to fluctuate greatly, resulting in a significant increase in the peak-to-peak value. Therefore, by calculating the peak-to-peak value, this abnormal change in the current waveform can be intuitively captured, providing a strong basis for the detection of arc faults.

[0037] Preferably, when calculating the time domain feature quantity, the calculation formula of the standard deviation and the kurtosis is:

[0038]

[0039] In the formula, Std rms Indicates the standard deviation of the effective value of the current, Kurt rms represents the kurtosis of the effective value of the current, N represents the total number of current cycles intercepted, k represents the ordinal number of the cycle, Indicates the mean effective value of the current intercepted, I m It represents the mth current sampling value in a cycle, and n represents the total number of current value sampling in a cycle.

[0040] With such a setting, when an arc fault occurs, the current waveform becomes irregular, the discreteness of the current value increases, and the standard deviation increases. Therefore, by calculating the standard deviation, the discreteness of the current waveform can be quantified, and then the presence and severity of the arc fault can be determined. In addition, in current waveform analysis, kurtosis can be used to detect extreme values ​​or abnormal fluctuations in the current waveform. When an arc fault occurs, large instantaneous current values ​​or abnormal fluctuation patterns may appear in the current waveform, causing the kurtosis to change. Therefore, by calculating the kurtosis, these abnormal features in the current waveform can be captured, providing additional information for the detection and diagnosis of arc faults.

[0041] Preferably, when calculating the frequency domain feature quantity, the calculation process of the harmonic amplitude includes:

[0042] First, the current signal of the disturbance current is filtered and preprocessed;

[0043] Then, the preprocessed current signal is subjected to short-time Fourier transform STFT;

[0044] Then, the average value of the results obtained by the short-time Fourier transform is calculated to obtain the harmonic amplitude of the disturbance current; wherein the Hanning window is selected as the window function of the short-time Fourier transform.

[0045] With such settings, filtering preprocessing can remove high-frequency noise and interference in the current signal and improve the quality of the signal. This is crucial for the subsequent short-time Fourier transform and accurate calculation of harmonic amplitudes. STFT can jointly analyze the signal in the time domain and frequency domain, and provide frequency component information of the signal in different time periods. This is particularly useful for analyzing non-stationary signals (such as disturbed currents). By selecting a suitable window function (Hanning window) and parameter settings, STFT can achieve higher time resolution and frequency resolution, thereby more accurately capturing the detailed characteristics of the signal. By averaging the STFT results, the influence of random errors and measurement noise can be reduced, and the calculation accuracy of the harmonic amplitude can be improved; the stability of the results can also be enhanced, so that the calculation results of the harmonic amplitude remain consistent in different time periods. This method can accurately calculate the harmonic amplitude of the disturbed current, providing strong support for power system analysis, signal processing, and fault diagnosis.

[0046] Preferably, when calculating the frequency domain feature quantity, the spectrum centroid f c The calculation formula is:

[0047]

[0048] Spectral standard deviation Std f The calculation formula is:

[0049]

[0050] In the formula, f k represents the frequency value at the kth position in the frequency sequence, A k Indicates frequency f k The corresponding amplitude is at , and N is the limited spectrum width.

[0051] With this setup, the spectrum centroid can be used to detect spectrum changes in current or voltage signals. When there is a fault or anomaly in the power system, such as a series arc fault, the spectrum distribution of the current or voltage signal may change, resulting in a shift in the spectrum centroid. Therefore, by calculating the spectrum centroid, this change in spectrum distribution can be captured, providing a powerful basis for fault detection and diagnosis.

[0052] The spectrum standard deviation can be used to evaluate the spectrum stability of current or voltage signals. When there are unstable factors or faults in the power system, the spectrum distribution of the current or voltage signal may become more dispersed, resulting in an increase in the spectrum standard deviation. Therefore, by calculating the spectrum standard deviation, the degree of dispersion of this spectrum distribution can be quantified, providing important reference information for the stability and reliability assessment of the power system.

[0053] Preferably, in S4, the preset fault identification model is a Stacking model composed of a decision tree DT, a support vector machine SVM, a K nearest neighbor KNN and a logistic regression LR built based on the logic of an integrated learning algorithm; and the Stacking model is trained using historical data of current time-frequency domain characteristics of actual cable series arc faults that occurred in the past.

[0054] Such a setting can, 1. improve the accuracy of fault identification. The Stacking model can fully utilize the advantages of each base learner by integrating multiple different base learners (such as decision trees, support vector machines, K nearest neighbors, and logistic regression), thereby improving the accuracy of fault identification. Each base learner can extract different information from the current time-frequency domain feature quantity, and the Stacking model can effectively combine this information to form a more comprehensive fault identification capability.

[0055] 2. It can enhance the generalization ability of the model. Using historical data of the current time-frequency domain characteristics of real cable series arc faults in the past to train the Stacking model can enhance the generalization ability of the model. This means that the model can not only accurately identify the fault modes in the training data, but also effectively predict and identify the unseen fault data. This is crucial for fault detection and diagnosis in practical applications.

[0056] Preferably, in S4, if it is determined that a series arc fault occurs, the disturbance current waveform of the cable fault and the location of the faulty cable joint are sent to the back-end, and the back-end sends a fault warning message to the management end and / or the maintenance end.

[0057] With this setup, the system can respond quickly after detecting a series arc fault by automatically sending fault information. This helps management and maintenance personnel to understand the fault situation in a timely manner, so that they can quickly take necessary intervention measures to prevent the fault from further deteriorating or causing a larger safety accident.

[0058] Furthermore, the system can accurately send the location of the faulty cable joint to the backend, which greatly simplifies the troubleshooting process. Maintenance personnel can go directly to the designated location for inspection without extensive searching and troubleshooting, thus improving maintenance efficiency and reducing maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to make the purpose, technical solution and advantages of the invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings, in which:

[0060] Figure 1 A flowchart of the method is shown in FIG.

[0061] Figure 2is a graph showing the change of the Pearson correlation coefficient after the current disturbance occurs in the cable in the embodiment;

[0062] Figure 3 It is a schematic diagram of the sliding window displacement logic in the embodiment;

[0063] Figure 4 Schematic diagram of the training process of the Stacking model in the embodiment. DETAILED DESCRIPTION

[0064] The following is a further detailed description through specific implementation methods:

[0065] Example:

[0066] like Figure 1 As shown, this embodiment discloses a method for judging a series arc fault in a low-voltage distribution cable based on a disturbance current characteristic, comprising the following steps:

[0067] S1. Real-time acquisition of current data of three-phase current of low-voltage distribution network cables.

[0068] In specific implementation, the three-phase current data of the cable intermediate joint can be collected in real time through the cable intermediate joint current carrier and stored in the data storage module. The current sensor and the cable joint current carrier are integrated to collect current signals, and the Hall sensor in the cable intermediate joint current carrier and sensor integration module stores the collected real-time three-phase current data in the data storage module, and transmits the data through the communication module.

[0069] S2. The Pearson correlation coefficient PCC is used to analyze the similarity of the current waveforms of adjacent current cycles for the acquired current data, and to determine whether current disturbance occurs in the cable.

[0070] Among them, when the Pearson correlation coefficient PCC is used to analyze the similarity of the current waveforms of each adjacent current cycle, the calculation formula of the Pearson correlation coefficient of each current cycle is:

[0071]

[0072] Where, PCC m represents the Pearson correlation coefficient of the mth current cycle; represents the kth instantaneous current value in the mth current cycle, represents the current mean value of the mth cycle; K is the number of instantaneous current values ​​in one current cycle.

[0073] The change of Pearson correlation coefficient after the current disturbance in the cable is as follows: Figure 2 shown.

[0074] The Pearson correlation coefficient is a statistical indicator that measures the degree of linear correlation between two variables, and its value ranges from -1 to 1. When it is applied to analyze the similarity of current waveforms in adjacent current cycles, it can accurately capture the subtle changes in the current waveform between cycles. This change may reflect potential faults or abnormal conditions in the cable, such as series arc faults. When the current waveform changes significantly, the Pearson correlation coefficient will also change significantly accordingly. This change provides a sensitive indicator for fault detection, allowing the system to detect potential faults earlier, thereby improving the sensitivity of fault detection. In subsequent analysis, the Pearson correlation coefficient can be used as an important basis for determining whether a current disturbance occurs in the cable. When the Pearson correlation coefficient of adjacent current cycles is lower than a preset threshold, it can be considered that the current waveform has changed significantly, and then it can be determined that a fault may have occurred in the cable. This provides strong data support for fault identification.

[0075] like Figure 3 In specific implementation, when judging whether the cable has current disturbance, it is assumed that the obtained continuous Pearson correlation coefficient PCC data sequence X = {x1, x2, ..., x K}, Pearson correlation coefficient indicator processing uses the CUSUM algorithm based on double sliding windows W1 and W2 to judge the current disturbance.

[0076] Among them, W1 is the front window, which is used to obtain the expected value E of the distribution and determine the upper threshold UB and the lower threshold LB; the expected value E is the average value calculated based on the current waveform data distribution in the front window W1; the values ​​of the upper threshold UB and the lower threshold LB are set based on historical data, and can usually be set to μ±kσ, where σ is the standard deviation and k is the preset coefficient. W2 is the rear window, which is used to calculate the offset accumulation value C of the data sequence in the window; the window length of the two windows is M current cycles.

[0077] Set the initial accumulation value of the positive offset and negative offset initial accumulation value are all 0; after each sliding displacement of the window, based on the expected value E and the thresholds UB and LB, the offset of each point data in W2 is calculated in turn, and the accumulated value of the positive offset is and negative offset initial accumulation value Iterate; the iterative process is as follows:

[0078]

[0079] In the formula, x k is the value of the Pearson correlation coefficient of the kth current cycle in W2; when x k When it is greater than the upper threshold E+UB, the positive offset accumulation value Increase; when xk When it is less than the lower threshold E-LB, the negative offset accumulation value reduce;

[0080] If satisfied or It is determined that a disturbance has occurred in the current in the window; wherein DL is a preset offset limit.

[0081] Such a setting can improve the sensitivity of disturbance detection. By setting the front window, the technology can obtain the expected value of the distribution of the current waveform data, and determine the upper and lower thresholds based on this. These thresholds are set based on historical data and can reflect the fluctuation range of the normal current waveform. When the data in the rear window deviates from these thresholds, it indicates that the current may be disturbed. Compared with a single threshold judgment, this method can better capture small current changes, thereby improving the sensitivity of disturbance detection. In addition, the accuracy of disturbance identification can be enhanced. The rear window is used to calculate the offset accumulation value of the data sequence in the window. By iterating the positive offset accumulation value and the negative offset accumulation value, the offset of the current waveform data relative to the expected value can be accumulated. When the offset accumulation value reaches the preset offset limit, it can be judged that the current has undergone a disturbance change. This method can more accurately reflect the overall change trend of the current waveform by accumulating the offset, and enhance the accuracy of disturbance identification. And it is highly adaptable and suitable for a variety of scenarios. By adjusting the window length, threshold, and offset limit parameters of the front window and the rear window, it can meet the current disturbance detection requirements under different cable types and different operating conditions. This makes the technology highly adaptable and can be widely used in various power system monitoring and fault diagnosis scenarios.

[0082] S3. When it is determined that a current disturbance occurs, the current period in which the current disturbance occurs is taken as the disturbance starting point, and the current data of the disturbance starting point and multiple current periods thereafter are intercepted and recorded as the disturbance current; and the time-frequency domain characteristic quantity of the intercepted disturbance current is calculated.

[0083] Among them, when calculating the time-frequency domain characteristics of the intercepted disturbance current, the calculated time-domain characteristics include zero-off time, peak-to-peak value, standard deviation and kurtosis; the calculated frequency-domain characteristics include harmonic amplitude, spectrum centroid and spectrum standard deviation. In this way, by calculating the time-domain characteristics (zero-off time, peak-to-peak value, standard deviation, kurtosis) and frequency-domain characteristics (harmonic amplitude, spectrum centroid, spectrum standard deviation) of the disturbance current, the characteristics of the series arc fault in the cable can be fully reflected, providing strong data support for the detection and diagnosis of the series arc fault in the cable. These characteristics are not only intuitive, but also easy to calculate and extract, and have broad application prospects in power system monitoring and fault diagnosis.

[0084] In specific implementation, when calculating the time domain feature quantity, the zero rest time is calculated by the following formula:

[0085]

[0086] Where, T zero Indicates zero rest time in a single cycle, i k Indicates the amplitude of the kth sampled current in the cycle, I rms Indicates 5% of the effective value of the current in this cycle, f s Indicates the corresponding acquisition frequency; N indicates the total number of current acquisitions within a cycle; α k (i k <I rms ) is whether the condition i is met k <I rms The statistical parameters of k <I rms When α k (i k <I rms )=1, when condition i is not met k <I rms When α k (i k <I rms )=0.

[0087] Zero-off time is the specific period of arc extinguishing and reigniting in the periodic change of alternating current, which is manifested as a period of time when the current value is close to zero. This feature can intuitively reflect the periodicity of arc activity and is crucial for judging the existence of arc faults and their activity patterns.

[0088] The peak-to-peak value is calculated as:

[0089]

[0090] PPV k represents the normalized peak-to-peak value of the kth current cycle, and Indicates the maximum and minimum values ​​of the current collected in the kth current cycle.

[0091] When an arc fault occurs, due to the nonlinear characteristics and instability of the arc, the current waveform tends to fluctuate greatly, resulting in a significant increase in the peak-to-peak value. Therefore, by calculating the peak-to-peak value, this abnormal change in the current waveform can be intuitively captured, providing a strong basis for the detection of arc faults.

[0092] The calculation formulas for standard deviation and kurtosis are:

[0093]

[0094] In the formula, Stdrms Indicates the standard deviation of the effective value of the current, Kurt rms represents the kurtosis of the effective value of the current, N represents the total number of current cycles intercepted, k represents the ordinal number of the cycle, Indicates the mean effective value of the current intercepted, I m It represents the mth current sampling value in a cycle, and n represents the total number of current value sampling in a cycle.

[0095] When an arc fault occurs, the current waveform becomes irregular, and the degree of dispersion of the current value increases, resulting in an increase in the standard deviation. Therefore, by calculating the standard deviation, the degree of dispersion of the current waveform can be quantified, and the presence and severity of the arc fault can be determined. In addition, in current waveform analysis, kurtosis can be used to detect extreme values ​​or abnormal fluctuations in the current waveform. When an arc fault occurs, large instantaneous current values ​​or abnormal fluctuation patterns may appear in the current waveform, causing the kurtosis to change. Therefore, by calculating the kurtosis, these abnormal features in the current waveform can be captured, providing additional information for the detection and diagnosis of arc faults.

[0096] When calculating the frequency domain characteristic quantity, the calculation process of the harmonic amplitude includes:

[0097] First, the current signal of the disturbance current is filtered and preprocessed;

[0098] Then, the preprocessed current signal is subjected to short-time Fourier transform STFT;

[0099] Then, the results obtained by the short-time Fourier transform are averaged to obtain the harmonic amplitude of the disturbance current; wherein the Hanning window is selected as the window function of the short-time Fourier transform. In specific implementation, before averaging, the high-order harmonics in the embodiment with extremely small amplitude can be screened out, and the harmonics from the 2nd to the 21st can be retained.

[0100] In this way, filtering preprocessing can remove high-frequency noise and interference in the current signal and improve the quality of the signal. This is crucial for the subsequent short-time Fourier transform and accurate calculation of harmonic amplitudes. STFT can jointly analyze the signal in the time domain and frequency domain, and provide frequency component information of the signal in different time periods. This is particularly useful for analyzing non-stationary signals (such as disturbed currents). By selecting a suitable window function (Hanning window) and parameter settings, STFT can achieve higher time resolution and frequency resolution, thereby more accurately capturing the detailed characteristics of the signal. By averaging the STFT results, the influence of random errors and measurement noise can be reduced, and the calculation accuracy of the harmonic amplitude can be improved; the stability of the results can also be enhanced, so that the calculation results of the harmonic amplitude remain consistent in different time periods. This method can accurately calculate the harmonic amplitude of the disturbed current, providing strong support for power system analysis, signal processing, and fault diagnosis.

[0101] Spectral centroid f c The calculation formula is:

[0102]

[0103] Spectral standard deviation Std f The calculation formula is:

[0104]

[0105] In the formula, f k represents the frequency value at the kth position in the frequency sequence, A k Indicates frequency f k The corresponding amplitude is at , and N is the limited spectrum width.

[0106] The spectrum centroid can be used to detect spectrum changes of current or voltage signals. When there is a fault or anomaly in the power system, such as a series arc fault, the spectrum distribution of the current or voltage signal may change, resulting in a shift in the spectrum centroid. Therefore, by calculating the spectrum centroid, this change in spectrum distribution can be captured, providing a powerful basis for fault detection and diagnosis.

[0107] The spectrum standard deviation can be used to evaluate the spectrum stability of current or voltage signals. When there are unstable factors or faults in the power system, the spectrum distribution of the current or voltage signal may become more dispersed, resulting in an increase in the spectrum standard deviation. Therefore, by calculating the spectrum standard deviation, the degree of dispersion of this spectrum distribution can be quantified, providing important reference information for the stability and reliability assessment of the power system.

[0108] S4. Input the time-frequency domain characteristic quantity of the disturbance current into a preset fault identification model to detect and identify the cable series arc fault, and determine whether the fault causing the current disturbance is a series arc fault.

[0109] The preset fault identification model is a Stacking model composed of decision tree DT, support vector machine SVM, K nearest neighbor KNN and logistic regression LR based on the logic of integrated learning algorithm; and the Stacking model is trained using the historical data of the current time-frequency domain characteristics of the cable series arc fault that actually occurred in the past. The training process of the Stacking model is as follows: Figure 4 shown.

[0110] In this way, the accuracy of fault identification can be improved. The Stacking model can fully utilize the advantages of each base learner by integrating multiple different base learners (such as decision trees, support vector machines, K nearest neighbors, and logistic regression), thereby improving the accuracy of fault identification. Each base learner can extract different information from the current time-frequency domain features, and the Stacking model can effectively combine this information to form a more comprehensive fault identification capability. In addition, the generalization ability of the model can also be enhanced. Using historical data of current time-frequency domain features of cable series arc faults that actually occurred in the past to train the Stacking model can enhance the generalization ability of the model. This means that the model can not only accurately identify fault modes in training data, but also effectively predict and identify unseen fault data. This is crucial for fault detection and diagnosis in practical applications.

[0111] In specific implementation, if it is determined that a series arc fault occurs, the disturbance current waveform of the cable fault and the location of the faulty cable joint are sent to the back-end, and the back-end sends a fault warning message to the management end and / or maintenance end.

[0112] In this way, by automatically sending fault information, the system can respond quickly after detecting a series arc fault. This helps management and maintenance personnel to understand the fault situation in a timely manner, so that they can quickly take necessary intervention measures to prevent the fault from further deteriorating or causing a larger safety accident. In addition, the system can accurately send the location of the faulty cable joint to the back-end, which greatly simplifies the troubleshooting process. Maintenance personnel can go directly to the designated location for maintenance without extensive searching and troubleshooting, thereby improving maintenance efficiency and reducing maintenance costs.

[0113] Compared with the prior art, the present invention can improve the accuracy of monitoring and identification. By using the Pearson correlation coefficient PCC to analyze the similarity of the current waveforms of each adjacent current cycle, the present method can preliminarily determine whether a current disturbance has occurred in the cable. This method effectively captures the slight changes in the current waveform, so that the current suspected of causing a cable series arc fault can be accurately screened out. Compared with the traditional method, the scheme improves the accuracy of monitoring and identification of cable fault arcs in the initial stage, laying a solid foundation for subsequent analysis and judgment. The timeliness of the identification of series arc faults can also be guaranteed. The present method analyzes the current data of the starting point of the disturbance current and multiple cycles thereafter, and can quickly capture the initial characteristics of the arc fault. Since arc faults are often weak in the early stage, but will gradually intensify over time, timely identification and taking measures are crucial to preventing the expansion of faults and protecting equipment and personnel safety. Through this method, accurate identification can be performed in the early stages of the development of arc faults, thereby ensuring the timeliness of the identification of series arc faults and providing a valuable time window for subsequent fault handling.

[0114] In addition, this method can reduce the amount of current data to be identified and reduce the monitoring delay. After confirming that a current disturbance has occurred, the scheme only intercepts the data of multiple current cycles after the starting point of the disturbance for further analysis. This approach avoids the processing of a large amount of irrelevant data and significantly reduces the amount of current data that needs to be calculated. At the same time, since only the data after the disturbance is processed, the time of the entire monitoring and identification process is greatly shortened, thereby reducing the monitoring delay and improving the real-time response capability of the system. The accuracy of identification can also be ensured. When a series arc fault occurs in the low-voltage distribution network cable, the current data will show obvious time-frequency domain feature changes such as an increase in high-frequency components and a decrease in periodic characteristics. This method calculates the time-frequency domain feature quantity of the disturbance current and inputs it into the preset fault identification model, which can make full use of these feature changes to accurately determine the fault type. This method not only takes into account the time domain characteristics of the current, but also combines the frequency domain characteristics, thereby ensuring the accuracy and reliability of the cable series arc fault identification.

[0115] This method realizes efficient, accurate and timely monitoring and identification of series arc faults in low-voltage distribution cables by comprehensively using Pearson correlation coefficient analysis, time-frequency domain feature quantity calculation and fault identification model, effectively improving the safety and stability of the power system.

[0116] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit the technical solution. Those skilled in the art should understand that those modifications or equivalent substitutions of the technical solution of the present invention that do not depart from the purpose and scope of the technical solution should be included in the scope of the claims of the present invention.

Claims

1. A method for judging a series arc fault in a low-voltage distribution cable based on disturbance current characteristics, characterized in that: The following steps are involved: S1, real-time acquisition of current data of three-phase current of low-voltage distribution network cables; S2. For the acquired current data, the Pearson correlation coefficient PCC is used to analyze the similarity of the current waveforms of each adjacent current cycle, and determine whether the cable has a current disturbance; S3, when it is determined that a current disturbance occurs, the current cycle where the current disturbance occurs is taken as the disturbance starting point, and the current data of the disturbance starting point and multiple current cycles thereafter are intercepted and recorded as the disturbance current; Calculate the time-frequency domain characteristic quantities of the intercepted disturbance current; S4. Input the time-frequency domain characteristic quantity of the disturbance current into a preset fault identification model to detect and identify the cable series arc fault, and determine whether the fault causing the current disturbance is a series arc fault.

2. The method for judging a series arc fault in a low-voltage distribution cable based on disturbance current characteristics according to claim 1, characterized in that: In S2, when judging whether the cable has current disturbance, the current disturbance is judged by combining the cumulative sum algorithm CUSUM based on the double sliding windows W1 and W2; Wherein, W1 is the front window, which is used to obtain the expected value E of the distribution and determine the upper threshold UB and the lower threshold LB; the expected value E is the average value calculated based on the current waveform data distribution in the front window W1; the values ​​of the upper threshold UB and the lower threshold LB are set based on historical data; W2 is the rear window, which is used to calculate the offset accumulation value C of the data sequence in the window; the window lengths of the two windows are both M current cycles; Set the initial accumulation value of the positive offset and negative offset initial accumulation value are all 0; after each sliding displacement of the window, based on the expected value E and the thresholds UB and LB, the offset of each point data in W2 is calculated in turn, and the accumulated value of the positive offset is and negative offset initial accumulation value Iterate; the iterative process is as follows: In the formula, x k is the value of the Pearson correlation coefficient of the kth current cycle in W2; when x k When it is greater than the upper threshold E+UB, the positive offset accumulation value Increase; when x k When it is less than the lower threshold E-LB, the negative offset accumulation value reduce; If satisfied or It is determined that a disturbance has occurred in the current in the window; wherein DL is a preset offset limit.

3. The method for judging a series arc fault in a low-voltage distribution cable based on disturbance current characteristics according to claim 2, characterized in that: In S2, when the Pearson correlation coefficient PCC is used to analyze the similarity of the current waveforms of each adjacent current cycle, the calculation formula of the Pearson correlation coefficient of each current cycle is: Where, PCC m represents the Pearson correlation coefficient of the mth current cycle; represents the kth instantaneous current value in the mth current cycle, represents the current mean value of the mth cycle; K is the number of instantaneous current values ​​in one current cycle.

4. The method for judging a series arc fault in a low-voltage distribution cable based on disturbance current characteristics according to claim 1, characterized in that: In S3, when calculating the time-frequency domain characteristic quantities of the intercepted disturbance current, the calculated time-domain characteristic quantities include zero-off time, peak-to-peak value, standard deviation and kurtosis; the calculated frequency-domain characteristic quantities include harmonic amplitude, spectrum centroid and spectrum standard deviation.

5. The method for judging a series arc fault in a low-voltage distribution cable based on disturbance current characteristics according to claim 4, characterized in that: When calculating the time domain feature quantity, the peak-to-peak value is calculated as follows: PPV k represents the normalized peak-to-peak value of the kth current cycle, and Indicates the maximum and minimum values ​​of the current collected in the kth current cycle.

6. The method for judging a series arc fault in a low-voltage distribution cable based on disturbance current characteristics according to claim 4, characterized in that: When calculating the time domain feature quantity, the calculation formula for standard deviation and kurtosis is: In the formula, Std rms Indicates the standard deviation of the effective value of the current, Kurt rms represents the kurtosis of the effective value of the current, N represents the total number of current cycles intercepted, k represents the ordinal number of the cycle, Represents the mean effective value of the current intercepted, I m It represents the mth current sampling value in a cycle, and n represents the total number of current value sampling in a cycle.

7. The method for judging a series arc fault in a low-voltage distribution cable based on disturbance current characteristics according to claim 4, characterized in that: When calculating the frequency domain characteristic quantity, the calculation process of the harmonic amplitude includes: First, the current signal of the disturbance current is filtered and preprocessed; Then, the preprocessed current signal is subjected to short-time Fourier transform STFT; Then, the average value of the results obtained by the short-time Fourier transform is calculated to obtain the harmonic amplitude of the disturbance current; wherein the Hanning window is selected as the window function of the short-time Fourier transform.

8. The method for judging a series arc fault in a low-voltage distribution cable based on disturbance current characteristics according to claim 4, characterized in that: When calculating the frequency domain feature quantity, the spectrum centroid f c The calculation formula is: Spectral standard deviation Std f The calculation formula is: In the formula, f k represents the frequency value at the kth position in the frequency sequence, A k Indicates frequency f k The corresponding amplitude is at , and N is the limited spectrum width.

9. The method for determining a series arc fault in a low-voltage distribution cable based on disturbance current characteristics according to claim 1, characterized in that: In S4, the preset fault identification model is a Stacking model composed of a decision tree DT, a support vector machine SVM, a K-nearest neighbor KNN and a logistic regression LR based on the logic of an integrated learning algorithm; and the Stacking model is trained using historical data of current time-frequency domain characteristics of cable series arc faults that actually occurred in the past.

10. The method for judging a series arc fault in a low-voltage distribution cable based on disturbance current characteristics according to claim 1, characterized in that: In S4, if it is determined that a series arc fault occurs, the disturbance current waveform of the cable fault and the location of the faulty cable joint are sent to the back-end, and the back-end sends a fault warning message to the management end and / or the maintenance end.

Images