An arc grounding fault identification method based on sound line feature quantization

Abstract

The present invention belongs to the field of relay protection for power systems and discloses a method for identifying arc ground faults based on the quantification of healthy line characteristics. The method comprises the following steps: using a relay protection device to detect in real time whether a ground fault has occurred in the power grid; if a fault is detected, filtering and performing computational processing on the incoming line zero-sequence voltage signal sampling data to obtain an operational array; processing the operational array using the least squares method, fitting the operational array, and constructing a reference array; performing gradient product calculations on the reference array and the healthy line zero-sequence current array to obtain a gradient array; simultaneously calculating the correlation coefficient between the reference array and the healthy line zero-sequence current array to obtain a correlation coefficient array; and comprehensively determining whether an arc fault has occurred in the power grid based on this calculation. The present invention utilizes the characteristics of the healthy line waveform for detection, enabling accurate identification of non-arcing ground faults and arc ground faults. The method has the advantages of wide applicability, high sensitivity, and good accuracy.

Description

Technical Field

[0001] The present invention belongs to the technical field of power system relay protection, and in particular relates to an arc grounding fault identification method based on quantification of sound line characteristics. Background Art

[0002] With the increasing development of power system technology and scale, the safe operation of distribution networks has become an increasingly critical factor in evaluating the power supply capacity of power systems. Various faults often occur during distribution network operation, and how to quickly identify the fault type and accurately locate the fault has become a top priority for current research.

[0003] Single-phase grounding faults are the most common fault type in distribution networks, and arc grounding faults are a special type of single-phase grounding fault. Arc faults are accompanied by strong arcing light and arcing noise. Prolonged presence in the network can cause immeasurable damage to the entire distribution network. Currently, arc fault detection primarily focuses on two aspects. First, focusing on the physical characteristics of arc faults, temperature and light sensors are used to detect arc faults in the line. However, this approach is often limited by environmental conditions, and sensors and other components are typically located only in switchgear. Second, focusing on the unique "saddle-shaped" current waveform and the voltage waveform with "zero-off periods" during arc faults, existing methods rely on detecting characteristic waveforms such as zero-sequence current and zero-sequence voltage. Typical methods include detecting zero-sequence current concavity and zero-crossing detection to identify arc faults. However, these methods, which utilize arc fault waveforms, suffer from randomness and idealization, and ignore the rich fault characteristics that can be used to identify arc faults on healthy lines. In fact, sound line fault characteristics are very helpful for identifying fault arcs, especially when the zero-off period of the zero-sequence current signal of the arc fault circuit is not obvious or difficult to distinguish, while the accuracy of fault identification using common fault identification methods cannot meet the expected goals.

[0004] Therefore, researchers in this field urgently need to develop an arc grounding fault identification method based on the quantification of sound line characteristics. Summary of the Invention

[0005] The purpose of the present invention is to provide an arc grounding fault detection method with strong resolution, fast recognition speed and high detection efficiency. The method is applicable to a power grid in which the neutral point is effectively grounded via a low resistance, and can detect and distinguish between non-arcing grounding faults and arc grounding faults, thereby solving the problems of low fault detection efficiency and slow protection speed caused by the complexity of current power grid operation.

[0006] To achieve the above-mentioned object, the technical solution adopted by the present invention is a method for identifying arc grounding faults based on quantification of sound line characteristics, comprising the following steps:

[0007] Step S1: Real-time detection of busbar zero-sequence voltage signal and each line zero-sequence current signal. When a ground fault occurs in the line, the fault line and the healthy line are determined by the relay protection device, and the healthy line zero-sequence current array is collected by the sampler. ,in, N The number of samples in the last 20 milliseconds;

[0008] Step S2: filtering the collected bus zero-sequence voltage signal and each line zero-sequence current signal using a filtering method;

[0009] Step S3: The filtered bus zero-sequence voltage signal is processed according to the formula (1) to obtain a set of operation arrays: ;

[0010] (1)

[0011] In formula (1), U m is the sampling data of the bus zero-sequence voltage signal after filtering. m is the data sequence number, U max and U min They are the maximum and minimum values of the busbar zero-sequence voltage signal sampling data in the last 20 milliseconds respectively;

[0012] Step S4: Operation array Substitute the fitting target set by the following formula (2) and use the least square method to calculate the array Perform fitting; find the fitting parameters that make the following formula (3) reach the minimum value 、 、 The value of

[0013] (2)

[0014] (3)

[0015] In formula (2), X M is the amplitude variable, θ is the initial phase angle; k= 1 , 2 , … ,N , k Represents the data sequence number; f (k ) is the fitting objective function; in formula (3), R 2 is the residual value;

[0016] Step S5: Fitting parameters 、 、 Substitute the value of into the construction target set by the following formula (4) to obtain a set of control arrays ;

[0017] (4)

[0018] Step S6: compare the array Perform gradient product calculation to obtain the gradient array ;

[0019] Step S7, compare the array The collected sound line zero sequence current array Calculate the correlation coefficients together and get the correlation coefficient array ;

[0020] Step S8: Determine the gradient array and correlation coefficient arrays Whether the following formula (5) is satisfied at the same time;

[0021] (5)

[0022] In formula (5), λ set is the threshold;

[0023] If equation (5) is satisfied, the fault is determined to be a non-arcing ground fault; otherwise, it is determined to be an arcing ground fault.

[0024] Preferably, the filtering method in step S2 is Butterworth filtering. The Butterworth filtering method is used to filter the collected signal. Its advantage is that the frequency response curve within the passband is extremely flat, without fluctuations, and gradually decreases to zero in the stopband, which can effectively filter out interference noise in the network.

[0025] Preferably, the gradient array in step S6 is obtained by the following method:

[0026] According to the set step size, the comparison array Substitute into the following formula (6), calculate the gradient product, and get the gradient array ;

[0027] (6)

[0028] Where, bc Indicates the setting step size, L k Indicates the setting step size k The left gradient product of the group, R k Indicates the setting step size k The right gradient product of the group, d k Indicates the k The gradient product of the group.

[0029] Preferably, the correlation coefficient array in step S7 It is obtained by the following formula:

[0030] (7).

[0031] Preferably, in step S8, the threshold λ set The value range is 0.2~0.5.

[0032] The working principle of the present invention is as follows: For arc grounding faults, the amplitude of the zero-sequence current signal is generally too small to be easily detected, and its prolonged existence poses a serious threat to the safety of the power grid. When an arc grounding fault occurs, the fault line's unique "saddle-shaped" current waveform and the voltage and current waveforms with a "zero off period" become important bases for detection and identification, accurately distinguishing them from non-arcing grounding faults. However, when the dissipated power is low, these waveform characteristics become less obvious, and the accuracy of previous detection methods using arc current characteristics of the fault line will be greatly reduced. The rich fault characteristics on healthy lines that can be used for arc fault identification are often overlooked. To address this special situation, the present invention considers starting with the quantification of the characteristics of healthy lines for detection and identification, using these as characteristic points to distinguish non-arcing grounding faults from arc grounding faults.

[0033] Relay protection devices are installed to monitor busbar zero-sequence voltage signals and zero-sequence current signals on each line. When a ground fault occurs, the relay protection device identifies the faulty line. A filtering method is used to filter the collected signals to remove interfering noise in the network. When a single-phase ground fault occurs in the system, the neutral current flows through the fault point, forming a loop with the ground. At this time, the busbar voltage at the neutral point is the phase voltage of the faulted line. Therefore, the magnitude of the intact line zero-sequence current has a derivative relationship with the faulted line phase voltage. By selecting an appropriate fitting target, the filtered voltage calculation array is processed. By substituting the calculated fitting parameters into the target, a comparison array is generated. Because the non-arcing ground fault current waveform and the intact line zero-sequence current waveform exhibit sinusoidal characteristics, and the comparison array shares distinct characteristics with the collected intact line zero-sequence current array, feature quantification can effectively distinguish between the two fault types.

[0034] Using an appropriate step size, the gradient product calculation is performed on the control array and the healthy line zero-sequence current array. The special waveform characteristics of the arc healthy zero-sequence current are used to effectively distinguish it from the healthy zero-sequence current in the non-arcing grounding fault (the waveform is a standard sine wave). After the gradient product calculation, the gradient array obtained in the arc grounding fault is greater than the threshold λ set The gradient array obtained in the non-arcing ground fault is less than the threshold λ set .

[0035] The correlation coefficient is generally used to reflect the closeness of the correlation between variables. The closer the correlation coefficient between two variables is to 1, the closer the correlation between the two variables is. By using the correlation coefficient, it is possible to accurately distinguish between non-arcing grounding faults and arcing grounding faults by analyzing the difference between the zero-sequence current waveform of the healthy line and the reference array in the arc fault network, and the high consistency between the zero-sequence current waveform of the healthy line and the reference array in the non-arcing grounding fault network. By grouping the sampled healthy line zero-sequence current array and the reference array according to a selected step size and calculating the correlation coefficient for the two arrays obtained after grouping, the results can effectively eliminate the influence of interference and significantly improve the accuracy of fault identification. The correlation coefficient arrays for non-arcing grounding faults are all greater than zero, while the correlation coefficient arrays for arcing grounding faults are all less than zero.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] By combining these two criteria, the fault type can be determined more accurately. This is especially true when the characteristic waveform of an arc-to-ground fault line is unclear or difficult to distinguish. Existing methods, such as determining concavity or zero-crossing detection, may fail. However, this method fully utilizes the rich fault characteristics of a healthy line and flexibly applies feature quantification to make accurate judgments.

[0038] The present invention proposes an arc grounding fault identification method based on the quantification of sound line characteristics. Its identification effect has excellent anti-interference ability in identifying extreme situations, can greatly improve the ability of the power grid to correctly identify faults, ensure the safe operation of the distribution network, and has the advantages of wide applicability, high sensitivity and good accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The present invention will be further described below with reference to the accompanying drawings and examples.

[0040] Figure 1 This is a flow chart of identifying arc grounding faults in the present invention in Example 1 and Example 2;

[0041] Figure 2This is a waveform diagram of the zero-sequence current of the arc fault line after filtering in Example 1 of the present invention;

[0042] Figure 3 This is a diagram showing the change of the gradient product of the arc fault line in Example 1 of the present invention.

[0043] Figure 4 This is a diagram showing the change of the correlation coefficient of the arc fault line in Example 1 of the present invention;

[0044] Figure 5 This is a waveform diagram of the zero-sequence current of the non-arcing fault line after filtering in Example 2 of the present invention;

[0045] Figure 6 This is a diagram showing the gradient product change of the non-arcing fault line in Example 2 of the present invention;

[0046] Figure 7 This is a graph showing the change in correlation coefficient of the non-arcing fault line in Example 2 of the present invention. DETAILED DESCRIPTION

[0047] The following is a detailed explanation of how to implement the present invention in conjunction with the accompanying drawings and embodiments of the present invention. The present invention is described herein only with reference to the given embodiments and is not limited to the description herein.

[0048] Example 1

[0049] The simulation model of Example 1 was built in the visual simulation tool Simulink. The system is a 10kV distribution network with a sampling frequency of 6000Hz and a system frequency of 50Hz. Three lines were set up, corresponding to line 1 of 6km, line 2 of 8km, and line 3 of 10km. The neutral point was effectively grounded. The arc fault model was selected as the improved Mayr arc model. The experimental data was based on the experimental data obtained by connecting the arc model to the 10kV distribution network. An arc grounding fault was set up for simulation experiments. This embodiment proposed an arc grounding fault identification method based on the quantification of sound line characteristics. The steps are as follows: Figure 1 As shown in , follow the steps below to identify:

[0050] Step S1: Real-time detection of busbar zero-sequence voltage signal and each line zero-sequence current signal. When a ground fault occurs in the line, the relay protection device issues a warning and determines that the faulty line is line 1 and the healthy lines are lines 2 and 3. The zero-sequence current array of the healthy lines is collected by the sampler. ,in, N The number of samples in the last 20 milliseconds;

[0051] Step S2: The collected bus zero-sequence voltage signal and the zero-sequence current signal in line 1 are filtered using the Butterworth filtering method; the waveform of the zero-sequence current signal in line 1 after filtering is as shown in FIG. Figure 2 As shown;

[0052] Step S3: The filtered bus zero-sequence voltage signal is processed according to the formula (1) to obtain a set of operation arrays: ;

[0053] (1)

[0054] In formula (1), U m is the sampling data of the bus zero-sequence voltage signal after filtering. m is the data sequence number, U max and U min They are the maximum and minimum values of the busbar zero-sequence voltage signal sampling data in the last 20 milliseconds respectively;

[0055] Step S4: The above operation array Substitute the fitting target set by the following formula (2) and use the least square method to calculate the array Perform fitting; find the fitting parameters that make the following formula (3) reach the minimum value 、 、 The value of

[0056] (2)

[0057] (3)

[0058] In formula (2), X M is the amplitude variable, θ is the initial phase angle; k= 1 , 2 , … ,N , k Represents the data sequence number; f ( k ) is the fitting objective function; in formula (3), R 2 is the residual value;

[0059] Step S5: Fitting parameters 、 、 Substitute the value of into the construction target set by the following formula (4) to obtain a set of control arrays ;

[0060] (4)

[0061] Step S6: Set the comparison array Perform gradient product calculation to obtain a set of gradient arrays , the data display results are as follows Figure 3 As shown;

[0062] Step S7, compare the array The collected sound line zero sequence current array Calculate the correlation coefficients together to get a set of correlation coefficient arrays , the data display results are as follows Figure 4 As shown;

[0063] Step S8: Determine the gradient array and correlation coefficient arrays Whether the following formula (5) is satisfied at the same time;

[0064] (5)

[0065] In formula (5), λ set is the threshold value. In this embodiment, λ set The value of is 0.3;

[0066] It has been verified that the gradient array and the correlation coefficient array obtained in this embodiment do not satisfy formula (5) at the same time, thereby judging that an arc grounding fault has occurred in the power grid. This is consistent with the preset situation, indicating that the method of the present invention can accurately judge that the fault type is an arc grounding fault.

[0067] Example 2

[0068] This embodiment proposes an arc grounding fault identification method based on the quantification of sound line characteristics. On the 10kV distribution network simulation model set in Example 1, a single-phase non-arcing grounding fault is established. The identification steps are also as follows: Figure 1 As described in , the detailed steps are as follows:

[0069] Step S1: Real-time detection of busbar zero-sequence voltage signal and each line zero-sequence current signal. When a ground fault occurs in the line, the relay protection device issues a warning and determines that the faulty line is line 1 and the healthy lines are lines 2 and 3. The zero-sequence current array of the healthy lines is collected by the sampler. ,in, N The number of samples in the last 20 milliseconds;

[0070] Step S2: The collected bus zero-sequence voltage signal and the zero-sequence current signal in line 1 are filtered using the Butterworth filtering method; the waveform of the zero-sequence current signal in line 1 after filtering is as shown in FIG. Figure 5 As shown;

[0071] Step S3: The filtered bus zero-sequence voltage signal is processed according to the formula (1) to obtain a set of operation arrays: ;

[0072] (1)

[0073] In formula (1), U m is the sampling data of the bus zero-sequence voltage signal after filtering. m is the data sequence number, U max and U min They are the maximum and minimum values of the busbar zero-sequence voltage signal sampling data in the last 20 milliseconds respectively;

[0074] Step S4: Operation array Substitute the fitting target set by the following formula (2) and use the least square method to calculate the array Perform fitting; find the fitting parameters that make the following formula (3) reach the minimum value 、 、 The value of

[0075] (2)

[0076] (3)

[0077] In formula (2), X M is the amplitude variable, θ is the initial phase angle; k= 1 , 2 , … ,N , k Represents the data sequence number; f ( k ) is the fitting objective function; in formula (3), R 2 is the residual value;

[0078] Step S5: Fitting parameters 、 、 Substitute the value of into the construction target set by the following formula (4) to obtain a set of control arrays ;

[0079] (4)

[0080] Step S6: Calculate the gradient product of the reference array to obtain the gradient array , the data display results are as follows Figure 6 As shown;

[0081] Step S7, compare the array The collected sound line zero sequence current array Calculate the correlation coefficients together and get the correlation coefficient array , the data display results are as follows Figure 7 As shown;

[0082] Step S8: Determine the gradient array and correlation coefficient arrays Whether the following formula (5) is satisfied at the same time;

[0083] (5)

[0084] In formula (5), λ set is the threshold value. In this embodiment, λ set The value of is 0.3;

[0085] It has been verified that the gradient array and the correlation coefficient array obtained in this embodiment both satisfy equation (5), thereby determining that a non-arcing ground fault has occurred in the power grid. This is also consistent with the preset situation, indicating that the method of the present invention can accurately determine that the fault type is a non-arcing ground fault.

[0086] The above two embodiments show that the method of the present invention is effective in identifying arcing grounding faults and non-arcing grounding faults.

[0087] It should be noted that the embodiments described above are only preferred embodiments of the present invention and are used to illustrate the technical solutions of the present invention. However, the scope of protection of the present invention is not limited by the embodiments described above. Finally, it should be noted that within the scope of this technical field, simple modifications, improvements, and equivalent substitutions made by other persons in the relevant field without departing from the technical solutions of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for identifying arc grounding faults based on quantification of sound line characteristics, characterized in that: The following steps are involved: Step S1: Real-time detection of busbar zero-sequence voltage signal and each line zero-sequence current signal. When a ground fault occurs in the line, the fault line and the healthy line are determined by the relay protection device, and the healthy line zero-sequence current array is collected by the sampler. ,in, N The number of samples in the last 20 milliseconds; Step S2: filtering the collected bus zero-sequence voltage signal and each line zero-sequence current signal using a filtering method; Step S3: The filtered bus zero-sequence voltage signal is processed according to the formula (1) to obtain a set of operation arrays: ; (1) In formula (1), U m is the sampling data of the bus zero-sequence voltage signal after filtering. m is the data sequence number, U max and U min They are the maximum and minimum values of the busbar zero-sequence voltage signal sampling data in the last 20 milliseconds respectively; Step S4: Operation array Substitute the fitting target set by the following formula (2) and use the least square method to calculate the array Perform fitting; find the fitting parameters that make the following formula (3) reach the minimum value 、 、 The value of (2) (3) In formula (2), X M is the amplitude variable, θ is the initial phase angle; k= 1 , 2 , … ,N , k Represents the data sequence number; f ( k ) is the fitting objective function; in formula (3), R 2 is the residual value; Step S5: Fitting parameters 、 、 Substitute the value of into the construction target set by the following formula (4) to obtain a set of control arrays ; (4) Step S6: compare the array Perform gradient product calculation to obtain the gradient array ; Step S7, compare the array The collected sound line zero sequence current array Calculate the correlation coefficients together and get the correlation coefficient array ; Step S8: Determine the gradient array and correlation coefficient arrays Whether the following formula (5) is satisfied at the same time; (5) In formula (5), λ set is the threshold; If equation (5) is satisfied, the fault is determined to be a non-arcing ground fault; otherwise, it is determined to be an arcing ground fault.

2. The arc grounding fault identification method based on quantification of sound line characteristics according to claim 1 is characterized in that: The filtering method in step S2 adopts Butterworth filtering method.

3. The arc grounding fault identification method based on sound line feature quantification according to claim 1, characterized in that: The gradient array in step S6 is obtained specifically by the following method: According to the set step size, the comparison array Substitute into the following formula (6), calculate the gradient product, and get the gradient array ; (6) Where, bc Indicates the setting step size, L k Indicates the setting step size k The left gradient product of the group, R k Indicates the setting step size k The right gradient product of the group, d k Indicates the k The gradient product of the group.

4. The arc grounding fault identification method based on quantification of sound line characteristics according to claim 1, characterized in that: The correlation coefficient array in step S7 It is obtained by the following formula: (7)。 5. The arc grounding fault identification method based on sound line feature quantification according to claim 1, characterized in that: In step S8, the threshold λ set The value range is 0.2~0.5.

Images