Tree line fault characteristic signal detection method based on zero sequence voltage

Abstract

The invention discloses a zero sequence voltage-based tree line fault characteristic signal detection method, which comprises the following steps of: acquiring an original discrete signal after the zero sequence voltage starts to increase and before the zero sequence voltage is stably reduced for the first time, smoothing the signal and calculating an effective value of a discrete zero sequence voltage signal in each period; fitting the discrete zero-sequence voltage effective value signal to obtain a continuous function of the discrete zero-sequence voltage effective value signal about time; solving a second-order derivative function for the continuous function, and judging whether a middle time period value of the second-order derivative function is positive or negative; calculating a second-order differential signal of the discrete zero-sequence voltage effective value; comparing the fluctuation amplitudes of the second-order differential signal in the first half time and the second half time; calculating the skewness of the second-order differential signal; comparing the Pearson correlation coefficients of the second-order differential signal in the first half time and the second half time; comprehensively analyzing whether a tree line fault occurs; the method can be used for distinguishing whether the distribution line fault is caused by tree line contact or not, and the purpose of rapidly and effectively identifying the fault is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of tree-line fault detection, and in particular to a tree-line fault characteristic signal detection method based on zero-sequence voltage. Background Art

[0002] Most distribution networks are systems with indirect neutral grounding. As a channel for power transmission, distribution lines are inevitably crossing forests and other areas with many trees. Tree line faults causing distribution line tripping and wildfires have become an important issue affecting the safety of the power grid. Tree line faults are generally caused by excessive growth of trees or contact with distribution lines under the action of wind deflection, causing tree line discharge and fire, thus causing distribution line failure. How to identify whether the line fault is a tree line fault by detecting changes in the electrical signals of the distribution line is a major technical problem that urgently needs to be studied and solved. In the detection of distribution line faults, when the tree line contacts and discharges, the zero-sequence voltage changes dramatically, and the fluctuation degree of the second-order derivative of the zero-sequence voltage is much greater than the fluctuation degree of the first-order derivative and the original data. By detecting and analyzing the zero-sequence voltage signal, it is possible to determine whether the line fault is a tree line fault based on the different characteristics of the zero-sequence voltage signal during tree line faults and other types of faults, providing a basis for accident cause analysis. Summary of the invention

[0003] In order to solve the problems existing in the prior art, the purpose of the present invention is to provide a tree-line fault characteristic signal detection method based on zero-sequence voltage. The present invention can be used to distinguish whether the distribution line fault is caused by tree-line contact, so as to achieve the purpose of quickly and effectively identifying the fault. Moreover, the present invention only needs to obtain the industrial frequency signal of the line zero-sequence voltage to effectively identify it, which is more conducive to on-site operation and maintenance personnel to make judgments on the fault type.

[0004] To achieve the above object, the technical solution adopted by the present invention is: a method for detecting characteristic signals of tree-line faults based on zero-sequence voltage, comprising the following steps:

[0005] Step 1, collecting the original discrete signal from the time when the zero-sequence voltage starts to increase to the time before the first stable decrease, smoothing the signal and calculating the effective value of the discrete zero-sequence voltage signal in each cycle;

[0006] Step 2: Fit the discrete zero-sequence voltage effective value signal to obtain its continuous function with respect to time;

[0007] Step 3, calculate the second-order derivative of the continuous function, and determine whether the value of the second-order derivative in the middle period is positive or negative;

[0008] Step 4, calculating the second-order differential signal of the discrete zero-sequence voltage effective value;

[0009] Step 5, comparing the fluctuation amplitude of the second-order difference signal in the first half of the time period and the second half of the time period;

[0010] Step 6: Calculate the skewness of the second-order difference signal;

[0011] Step 7, comparing the Pearson correlation coefficient of the second-order difference signal in the first half of the time period and the second half of the time period;

[0012] Step 8: Comprehensively analyze whether a tree line failure has occurred.

[0013] As a further improvement of the present invention, the step 1 is specifically as follows:

[0014] The library function smooth in MATLAB is used to smooth the collected original signal to reduce the noise of the original signal, and then the effective value of the discrete signal in each power frequency cycle is calculated;

[0015] The smoothing process is as follows:

[0016]

[0017] In the formula, A y [n] is the discrete signal after the nth smoothing process, N is the window size of the smoothing process, k is the window radius of the smoothing process, and y[i] is the discrete signal before the ith smoothing process;

[0018] The calculation of the effective value of the discrete signal in each power frequency cycle is as follows:

[0019]

[0020] Where U n is the effective value of the zero-sequence voltage in the nth power frequency cycle, M is the number of discrete signals in one power frequency cycle, u x It is the xth zero-sequence voltage discrete signal in one power frequency cycle.

[0021] As a further improvement of the present invention, the step 2 is specifically as follows:

[0022] The library functions polyfit and polyval in MATLAB are used to perform n-th order fitting on the discrete zero-sequence voltage effective value signal in step 1 to obtain its continuous function with respect to time, where n is a positive integer greater than 4.

[0023] As a further improvement of the present invention, the step 3 is specifically as follows:

[0024] Find the second-order derivative of the continuous function of the zero-sequence voltage effective value with respect to time obtained in step 2, and determine whether the second-order derivative is The positive and negative situation at t max is the maximum value of time.

[0025] As a further improvement of the present invention, the step 4 is specifically as follows:

[0026] The discrete zero-sequence voltage effective value signal obtained in step 1 is subjected to second-order difference with respect to time to obtain a quadratic approximation of the rate of change of the signal between adjacent points; wherein, the second-order difference is specifically performed as follows:

[0027]

[0028] Wherein, F(n) is the nth value after the second-order difference, f(n+2) is the n+2th value of the discrete zero-sequence voltage effective value signal, f(n+1) is the n+1th value of the discrete zero-sequence voltage effective value signal, f(n) is the nth value of the discrete zero-sequence voltage effective value signal, and T is the power frequency period.

[0029] As a further improvement of the present invention, the step 5 is specifically as follows:

[0030] Taking time as the scale, the discrete zero-sequence voltage effective value second-order differential signal obtained in step 4 is evenly divided into two sections, and the fluctuation amplitudes of the two sections of differential signals are compared;

[0031] The fluctuation range is as follows:

[0032]

[0033] Where S1 is the fluctuation amplitude of the second-order difference signal in the first half of the time, S2 is the fluctuation amplitude of the second-order difference signal in the second half of the time, and x i is the i-th second-order difference signal, μ1 is the average amplitude of the second-order difference signal in the first half of the time, μ2 is the average amplitude of the second-order difference signal in the second half of the time, and A is the number of discrete second-order difference signals.

[0034] As a further improvement of the present invention, the step 6 is specifically as follows:

[0035] Calculate the skewness of the discrete zero-sequence voltage effective value second-order difference signal obtained in step 4; the skewness is as follows:

[0036]

[0037] Where N is the number of discrete second-order difference signals, μ is the mean of the discrete second-order difference signals, σ is the standard deviation of the discrete second-order difference signals, and x i represents the i-th discrete second-order difference signal, and γ is the skewness.

[0038] As a further improvement of the present invention, the step 7 is specifically as follows:

[0039] Taking time as the scale, the discrete zero-sequence voltage effective value second-order differential signal obtained in step 4 is evenly divided into two sections, and the Pearson correlation coefficients of the two sections of differential signals are respectively calculated and compared;

[0040] The Pearson correlation coefficient is as follows:

[0041]

[0042] Where r1 is the Pearson correlation coefficient of the second-order difference signal in the first half of the time, r2 is the Pearson correlation coefficient of the second-order difference signal in the second half of the time, and t i is the time when the ith power frequency cycle ends, d i is the i-th second-order difference signal, is the mean value of the second-order difference signal in the first half of the period, is the mean value of the second-order difference signal in the second half of the period, is the average value of the first half of the period, is the average value of the second half of the period.

[0043] As a further improvement of the present invention, the step 8 is specifically as follows:

[0044] In step 3, the second-order derivative function is is always positive; in step 5, the fluctuation amplitude of the discrete zero-sequence voltage effective value second-order difference signal in the first half of the time is smaller than the fluctuation amplitude in the second half of the time; in step 6, the absolute value of the skewness of the discrete zero-sequence voltage effective value second-order difference signal is less than 0.5; in step 7, the Pearson correlation coefficient of the discrete zero-sequence voltage effective value second-order difference signal in the first half is greater than 0, and the Pearson correlation coefficient in the second half is greater than that in the first half and greater than 0.3; among the above conditions, if three or more are met, the probability that the line fault is a tree-line fault is greater, and if less than three are met, the probability that the line fault is a tree-line fault is smaller.

[0045] The present invention is used to identify whether a fault occurring in a line is a tree-line fault. The collected original zero-sequence voltage discrete signal is smoothed to reduce noise, and the second-order derivative function of the discrete zero-sequence voltage effective value signal and its fitting function and the second-order differential signal of the discrete zero-sequence voltage effective value signal are calculated at the same time; the fluctuation amplitude and the Pearson correlation coefficient of the first half of the second-order differential signal and the second half of the second-order differential signal are compared, and the skewness of the second-order differential signal is calculated; the four characteristic signals of the second-order derivative function, the fluctuation amplitude, the Pearson correlation coefficient and the skewness are comprehensively analyzed to determine whether a tree-line fault occurs, thereby providing a technical basis for the identification of the tree-line fault.

[0046] The beneficial effects of the present invention are:

[0047] The present invention takes into account that when the distribution line passes through the forest area, the tree-line contact failure may cause other types of faults and forest fires. Other types of faults and forest fires may also cause the distribution line passing through the forest area to fail. Different types of faults are of great significance to the division of responsibility for fire accidents. The operation and maintenance personnel also have different inspection and treatment methods for different fault types, but it is very difficult to judge the order of occurrence in engineering practice. The present invention smoothes the collected original zero-sequence voltage discrete signal to reduce its noise, and at the same time calculates the discrete zero-sequence voltage effective value signal and the second-order derivative of its fitting function and the second-order difference signal of the discrete zero-sequence voltage effective value signal, compares the fluctuation amplitude and Pearson correlation coefficient of the second-order difference signal in the first half and the second half, and calculates the skewness of the second-order difference signal. The zero-sequence voltage characteristic signals of different fault types are not exactly the same, so analyzing the differences in the four characteristic signals of the second-order derivative function, fluctuation amplitude, Pearson correlation coefficient, and skewness of the zero-sequence voltage can facilitate the operation and maintenance personnel to determine whether a tree-line fault has occurred. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 is a flow chart of an embodiment of the present invention. DETAILED DESCRIPTION

[0049] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0050] Example

[0051] like Figure 1 As shown, a method for detecting a characteristic signal of a tree-line fault based on zero-sequence voltage comprises the following steps:

[0052] Step 1, collecting the original discrete signal from the time when the zero-sequence voltage starts to increase to the time before the first stable decrease, smoothing the signal and calculating the effective value of the discrete zero-sequence voltage signal in each cycle;

[0053] The step 1 is specifically as follows:

[0054] The library function smooth in MATLAB is used to smooth the collected original signal to reduce the noise of the original signal, and then the effective value of the discrete signal in each power frequency cycle is calculated;

[0055] The smoothing process is specifically as follows:

[0056]

[0057] In the formula, A y [n] is the discrete signal after the nth smoothing process, N is the window size of the smoothing process, k is the window radius of the smoothing process, and y[i] is the discrete signal before the ith smoothing process;

[0058] The calculation of the effective value of the discrete signal in each power frequency cycle is as follows:

[0059]

[0060] Where U n is the effective value of the zero-sequence voltage in the nth power frequency cycle, M is the number of discrete signals in one power frequency cycle, u x is the xth zero-sequence voltage discrete signal in one power frequency cycle;

[0061] Step 2: Fit the discrete zero-sequence voltage effective value signal to obtain its continuous function with respect to time;

[0062] The step 2 is specifically as follows:

[0063] Use the library functions polyfit and polyval in MATLAB to perform n-order fitting on the discrete zero-sequence voltage effective value signal in step 1 to obtain its continuous function with respect to time. In order to ensure the fitting accuracy, n is a positive integer greater than 4.

[0064] Step 3, calculate the second-order derivative of the continuous function, and determine whether the value of the second-order derivative in the middle period is positive or negative;

[0065] The step 3 is as follows:

[0066] Find the second-order derivative of the continuous function of the zero-sequence voltage effective value with respect to time obtained in step 2, and determine whether the second-order derivative is The positive and negative situation at t max is the maximum value of time;

[0067] Step 4, calculating the second-order differential signal of the discrete zero-sequence voltage effective value;

[0068] The step 4 is specifically as follows:

[0069] Perform a second-order difference of the discrete zero-sequence voltage effective value signal obtained in step 1 with respect to time to obtain a quadratic approximation of the rate of change of the signal between adjacent points;

[0070] The second-order difference is specifically as follows:

[0071]

[0072] Wherein, F(n) is the nth value after the second-order difference, f(n+2) is the n+2th value of the discrete zero-sequence voltage effective value signal, f(n+1) is the n+1th value of the discrete zero-sequence voltage effective value signal, f(n) is the nth value of the discrete zero-sequence voltage effective value signal, and T is the power frequency period;

[0073] Step 5, comparing the fluctuation amplitude of the second-order difference signal in the first half of the time period and the second half of the time period;

[0074] The step 5 is specifically as follows:

[0075] Taking time as the scale, the discrete zero-sequence voltage effective value second-order differential signal obtained in step 4 is evenly divided into two sections, and the fluctuation amplitudes of the two sections of differential signals are compared;

[0076] The fluctuation amplitude is as follows:

[0077]

[0078] Where: S1 is the fluctuation amplitude of the second-order difference signal in the first half of the time, S2 is the fluctuation amplitude of the second-order difference signal in the second half of the time, x i is the i-th second-order difference signal, μ1 is the average amplitude of the second-order difference signal in the first half of the time, μ2 is the average amplitude of the second-order difference signal in the second half of the time, and A is the number of discrete second-order difference signals;

[0079] Step 6: Calculate the skewness of the second-order difference signal;

[0080] The step 6 is specifically as follows:

[0081] Calculate the skewness of the discrete zero-sequence voltage effective value second-order difference signal obtained in step 4;

[0082] The skewness is as follows:

[0083]

[0084] Where N is the number of discrete second-order difference signals, μ is the mean of the discrete second-order difference signals, σ is the standard deviation of the discrete second-order difference signals, and x i represents the i-th discrete second-order difference signal, γ is the skewness;

[0085] Step 7, comparing the Pearson correlation coefficient of the second-order difference signal in the first half of the time period and the second half of the time period;

[0086] The step 7 is specifically as follows:

[0087] Taking time as the scale, the discrete zero-sequence voltage effective value second-order differential signal obtained in step 4 is evenly divided into two sections, and the Pearson correlation coefficients of the two sections of differential signals are respectively calculated and compared;

[0088] The Pearson correlation coefficient is as follows:

[0089]

[0090] Where r1 is the Pearson correlation coefficient of the second-order difference signal in the first half of the time, r2 is the Pearson correlation coefficient of the second-order difference signal in the second half of the time, and t i is the time when the ith power frequency cycle ends, d i is the i-th second-order difference signal, is the mean value of the second-order difference signal in the first half of the period, is the mean value of the second-order difference signal in the second half of the period, is the average value of the first half of the period, is the mean value of the second half of the period;

[0091] Step 8: Comprehensively analyze whether a tree line failure has occurred;

[0092] The step 8 is specifically as follows:

[0093] In step 3, the second-order derivative function is is always positive; in step 5, the fluctuation amplitude of the discrete zero-sequence voltage effective value second-order difference signal in the first half of the time is smaller than the fluctuation amplitude in the second half of the time; in step 6, the absolute value of the skewness of the discrete zero-sequence voltage effective value second-order difference signal is less than 0.5; in step 7, the Pearson correlation coefficient of the discrete zero-sequence voltage effective value second-order difference signal in the first half is greater than 0, and the Pearson correlation coefficient in the second half is greater than that in the first half and greater than 0.3; among the above conditions, if three or more are met, the probability that the line fault is a tree-line fault is greater, and if less than three are met, the probability that the line fault is a tree-line fault is smaller.

[0094] In this embodiment, the original discrete signal from the beginning of the increase of the zero-sequence voltage to the first stable decrease is first collected, and the library function smooth in MATLAB is used to smooth the collected original signal to reduce the noise of the original signal, and then the effective value of the discrete signal in each power frequency cycle is calculated. The library functions polyfit and polyval in MATLAB are used to perform n-order fitting on the discrete zero-sequence voltage effective value signal to obtain its continuous function with respect to time. In order to ensure the accuracy of the fitting, n is a positive integer greater than 4. The second-order derivative function of the continuous function of the zero-sequence voltage effective value with respect to time is calculated, and the positive and negative conditions of the second-order derivative function in the middle period are determined. The obtained discrete zero-sequence voltage effective value signal is subjected to second-order difference with respect to time to obtain a quadratic approximation of the rate of change of the signal between adjacent points. Using time as the scale, the obtained discrete zero-sequence voltage effective value second-order differential signal is evenly divided into two sections, and the fluctuation amplitudes of the two sections of the differential signal are compared. The skewness of the obtained discrete zero-sequence voltage effective value second-order differential signal is calculated. Taking time as the scale, the obtained discrete zero-sequence voltage effective value second-order differential signal is evenly divided into two sections, and the Pearson correlation coefficients of the two differential signals are calculated and compared. The second-order derivative of the continuous function of the zero-sequence voltage effective value with respect to time is always positive in the middle period; the fluctuation amplitude of the discrete zero-sequence voltage effective value second-order differential signal in the first half of the time is smaller than the fluctuation amplitude in the second half of the time; the absolute value of the skewness of the discrete zero-sequence voltage effective value second-order differential signal is less than 0.5; the Pearson correlation coefficient of the discrete zero-sequence voltage effective value second-order differential signal in the first half is greater than 0, and the Pearson correlation coefficient in the second half is greater than that in the first half and greater than 0.3; among the above conditions, if three or more are met, the probability that the line fault is a tree-line fault is greater, and if less than three are met, the probability that the line fault is a tree-line fault is smaller.

[0095] In this embodiment, the original zero-sequence voltage discrete data is smoothed to reduce the noise interference of the original signal, and the calculation of the effective value in each power frequency cycle can highlight the characteristic quantity of the original zero-sequence voltage discrete signal in the power frequency. When a tree line fault occurs, the zero-sequence voltage is highly sensitive and changes violently, and the fluctuation degree of its second-order derivative is greater than that of the first-order derivative and the original data. Therefore, the second-order derivative of the zero-sequence voltage is used for analysis, which can better highlight the characteristic signal. When metal grounding, flame burning wire, broken line non-tree grounding and other faults occur, the fluctuation amplitude, Pearson correlation coefficient and skewness of the second-order derivative function of the zero-sequence voltage and the second-order differential signal of the zero-sequence voltage are different, so they can be used as the characteristic quantity of the zero-sequence voltage during line faults. Because when a tree line fault occurs in the distribution system under different ground capacitance currents or residual currents, the change trend of the zero-sequence voltage is the same, and the characteristic quantities are all compared with the first half and the second half of the discrete zero-sequence voltage effective value or with a certain constant, so there is no need to consider the size of the ground capacitance current and residual current of the distribution line.

[0096] The above-mentioned embodiments only express the specific implementation of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A method for detecting characteristic signals of tree-line faults based on zero-sequence voltage, characterized in that: The steps include: Step 1, collecting the original discrete signal from the time when the zero-sequence voltage starts to increase to the time before the first stable decrease, smoothing the signal and calculating the effective value of the discrete zero-sequence voltage signal in each cycle; Step 2: Fit the discrete zero-sequence voltage effective value signal to obtain its continuous function with respect to time; Step 3, calculate the second-order derivative of the continuous function, and determine whether the value of the second-order derivative in the middle period is positive or negative; Step 4, calculating the second-order differential signal of the discrete zero-sequence voltage effective value; Step 5, comparing the fluctuation amplitude of the second-order difference signal in the first half of the time period and the second half of the time period; Step 6: Calculate the skewness of the second-order difference signal; Step 7, comparing the Pearson correlation coefficient of the second-order difference signal in the first half of the time period and the second half of the time period; Step 8: Comprehensively analyze whether a tree line failure has occurred.

2. The method for detecting characteristic signals of tree-line faults based on zero-sequence voltage according to claim 1, characterized in that: The step 1 is specifically as follows: The library function smooth in MATLAB is used to smooth the collected original signal to reduce the noise of the original signal, and then the effective value of the discrete signal in each power frequency cycle is calculated; The smoothing process is as follows: In the formula, A y [n] is the discrete signal after the nth smoothing process, N is the window size of the smoothing process, k is the window radius of the smoothing process, and y[i] is the discrete signal before the ith smoothing process; The calculation of the effective value of the discrete signal in each power frequency cycle is as follows: Where U n is the effective value of the zero-sequence voltage in the nth power frequency cycle, M is the number of discrete signals in one power frequency cycle, u x It is the xth zero-sequence voltage discrete signal in one power frequency cycle.

3. The method for detecting tree-line fault characteristic signals based on zero-sequence voltage according to claim 2 is characterized in that: The step 2 is specifically as follows: The library functions polyfit and polyval in MATLAB are used to perform n-th order fitting on the discrete zero-sequence voltage effective value signal in step 1 to obtain its continuous function with respect to time, where n is a positive integer greater than 4.

4. The method for detecting characteristic signals of tree-line faults based on zero-sequence voltage according to claim 3 is characterized in that: The step 3 is as follows: Find the second-order derivative of the continuous function of the zero-sequence voltage effective value with respect to time obtained in step 2, and determine whether the second-order derivative is The positive and negative situation at t max is the maximum value of time.

5. The method for detecting characteristic signals of tree-line faults based on zero-sequence voltage according to claim 4 is characterized in that: The step 4 is specifically as follows: The discrete zero-sequence voltage effective value signal obtained in step 1 is subjected to second-order difference with respect to time to obtain a quadratic approximation of the rate of change of the signal between adjacent points; wherein, the second-order difference is specifically performed as follows: Wherein, F(n) is the nth value after the second-order difference, f(n+2) is the n+2th value of the discrete zero-sequence voltage effective value signal, f(n+1) is the n+1th value of the discrete zero-sequence voltage effective value signal, f(n) is the nth value of the discrete zero-sequence voltage effective value signal, and T is the power frequency period.

6. The method for detecting characteristic signals of tree-line faults based on zero-sequence voltage according to claim 5, characterized in that: The step 5 is specifically as follows: Taking time as the scale, the discrete zero-sequence voltage effective value second-order differential signal obtained in step 4 is evenly divided into two sections, and the fluctuation amplitudes of the two sections of differential signals are compared; The fluctuation range is as follows: In the formula, S1 is the fluctuation amplitude of the second-order difference signal in the first half of the time, S2 is the fluctuation amplitude of the second-order difference signal in the second half of the time, and x i is the i-th second-order difference signal, μ1 is the average amplitude of the second-order difference signal in the first half of the time, μ2 is the average amplitude of the second-order difference signal in the second half of the time, and A is the number of discrete second-order difference signals.

7. The method for detecting tree-line fault characteristic signals based on zero-sequence voltage according to claim 6, characterized in that: The step 6 is specifically as follows: Calculate the skewness of the discrete zero-sequence voltage effective value second-order difference signal obtained in step 4; the skewness is as follows: Where N is the number of discrete second-order difference signals, μ is the mean of the discrete second-order difference signals, σ is the standard deviation of the discrete second-order difference signals, and x i represents the i-th discrete second-order difference signal, and γ is the skewness.

8. The method for detecting tree-line fault characteristic signals based on zero-sequence voltage according to claim 7, characterized in that: The step 7 is specifically as follows: Taking time as the scale, the discrete zero-sequence voltage effective value second-order differential signal obtained in step 4 is evenly divided into two sections, and the Pearson correlation coefficients of the two sections of differential signals are respectively calculated and compared; The Pearson correlation coefficient is as follows: Where r1 is the Pearson correlation coefficient of the second-order difference signal in the first half of the time, r2 is the Pearson correlation coefficient of the second-order difference signal in the second half of the time, and t i is the time when the ith power frequency cycle ends, d i is the i-th second-order difference signal, is the mean value of the second-order difference signal in the first half of the period, is the mean value of the second-order difference signal in the second half of the period, is the average value of the first half of the period, is the average value of the second half of the period.

9. The method for detecting tree-line fault characteristic signals based on zero-sequence voltage according to claim 8, characterized in that: The step 8 is specifically as follows: In step 3, the second-order derivative function is is always positive; in step 5, the fluctuation amplitude of the discrete zero-sequence voltage effective value second-order difference signal in the first half of the time is smaller than the fluctuation amplitude in the second half of the time; in step 6, the absolute value of the skewness of the discrete zero-sequence voltage effective value second-order difference signal is less than 0.5; in step 7, the Pearson correlation coefficient of the discrete zero-sequence voltage effective value second-order difference signal in the first half is greater than 0, and the Pearson correlation coefficient in the second half is greater than that in the first half and greater than 0.3; among the above conditions, if three or more are met, the probability that the line fault is a tree-line fault is greater, and if less than three are met, the probability that the line fault is a tree-line fault is smaller.

Images