Novel arc light high-resistance fault detection method for power distribution network

Abstract

The invention relates to a novel arc light high-resistance fault detection method for a power distribution network. The arc light high-resistance fault detection method comprises the following steps: firstly, acquiring zero-sequence current of each feeder line, calculating a kurtosis value of the zero-sequence current of each feeder line within a quarter period after a fault occurs, and starting the arc light high-resistance fault detection method when the kurtosis value of any feeder line is greater than a set threshold value; secondly, acquiring a bus zero-sequence voltage, performing trapezoidal sliding window integration on the bus zero-sequence voltage, and performing differential calculation on the integrated bus zero-sequence voltage to obtain a characteristic zero-sequence voltage; then, performing low-pass filtering on the zero-sequence current of each feeder line to obtain characteristic zero-sequence current of each feeder line; and finally, calculating the transient characteristic power ratio of each feeder by using the obtained characteristic zero-sequence voltage and the characteristic zero-sequence current of each feeder, and realizing arc light high-resistance fault detection according to the positive and negative differences of the transient characteristic power ratios.

Description

Technical Field

[0001] The invention relates to a new method for detecting arc high-resistance faults in a distribution network, and belongs to the technical field of relay protection of distribution networks in power systems. Background Art

[0002] High-resistance faults account for 5% to 20% of medium-voltage distribution network faults. Due to the high transition resistance, these faults often produce weak fault signals and are often accompanied by arcing, which can easily lead to major accidents such as fires. Detecting arc-type high-resistance faults is particularly challenging. The neutral point of my country's distribution networks is often ungrounded or grounded via arc suppression coils. Arc suppression coils further weaken the fault signature, hindering line selection for arc-type high-resistance faults. Existing transient power direction methods are limited by the weak signature of arc-type high-resistance faults, making them difficult to effectively detect. To overcome this technical bottleneck, there is an urgent need to improve the transient characteristic power direction method and develop a new arc-type high-resistance fault detection method to accurately extract fault information, improve detection reliability, and ensure the safe and stable operation of the distribution network. Summary of the Invention

[0003] The purpose of the present invention is to provide a new method for detecting arc high-resistance faults in distribution networks, so as to solve the problem that arc high-resistance faults are difficult to detect effectively using the existing transient power direction method due to their weak characteristics.

[0004] The technical solution adopted by the present invention is a new method for detecting arc high-resistance faults in distribution networks based on an improved transient characteristic power direction method, which specifically includes the following steps:

[0005] Step 1: Obtain the zero-sequence current i of each feeder 0h (t e ), h is the feeder number, h=1,2,3…,n, e is the sampling point number, calculate the kurtosis value K of the zero-sequence current of each feeder within a quarter cycle after the fault occurs h , when the kurtosis value of any feeder is greater than the set threshold β, the protection device is activated and then the process goes to step 2;

[0006] Step 2: Get the bus zero-sequence voltage u0(t e ), perform trapezoidal sliding window integration on the busbar zero-sequence voltage to obtain u 0J (t e ), and then the integrated bus zero sequence voltage u 0J (t e ) to perform differential calculation and obtain the characteristic zero-sequence voltage u 0W (t e );

[0007] Step 3: Calculate the zero-sequence current i of each feeder 0h (t e ) is low-pass filtered to obtain the characteristic zero-sequence current i of each feeder lh (t e), and the characteristic zero-sequence voltage u in step 2 0W (t e ) Calculate the transient characteristic power ratio T of each feeder h , and according to the transient characteristic power ratio T h The positive and negative difference of the sensor can realize arc high resistance fault detection.

[0008] Preferably, the specific process of using the zero-sequence current of each feeder to construct the starting criterion in step 1 is:

[0009] Step 1.1: Obtain the zero-sequence current i of each feeder 0h (t e ), h=1,2,3…,n, calculate the kurtosis value K of the zero-sequence current of each feeder within a quarter cycle after the fault occurs h , kurtosis value K h The specific calculation formula is shown as follows:

[0010]

[0011] Where i j is the zero-sequence current sampling value corresponding to the j-th sampling point, n is the number of zero-sequence current sampling points within a quarter cycle after the fault occurs, μ is the mean value of the zero-sequence current within a quarter cycle after the fault occurs, σ is the standard deviation of the zero-sequence current within a quarter cycle after the fault occurs,

[0012] Step 1.2: When the kurtosis value of any feeder is greater than the set threshold β, the protection device is activated and then the process goes to step 2;

[0013] The principle for selecting the threshold β is as follows: when the system is operating normally, the current waveform is a standard sine wave with a kurtosis value of 1.5; after a fault occurs, the waveform will experience a transient process, resulting in a sudden increase in the kurtosis value. Due to the existence of three-phase asymmetric operation or noise in the actual system, a margin of 2 to 3 times is considered here, and the value of β is set to 4.5, that is, there is a K h When >4.5, the protection device is activated.

[0014] Preferably, step 2 includes the following specific steps:

[0015] Step 2.1: Get the busbar zero-sequence voltage u0(t e ), a sliding trapezoid with a window size of 7 sampling points is used to integrate the bus zero-sequence voltage, as shown in the following formula:

[0016]

[0017] Where, f sis the sampling rate, k is the first sampling point in the sampling window, l=1,2,3,4,5 are the corresponding different sampling points in the sampling window, u0(t k ) is the busbar zero-sequence voltage sampling value corresponding to the kth sampling point in the sampling window, u0(t k+l ) is the busbar zero-sequence voltage sampling value corresponding to the k+1th sampling point to the k+5th sampling point in the sampling window, u0(t k+6 ) is the bus zero-sequence voltage sampling value corresponding to the k+6th sampling point in the sampling window;

[0018] Step 2.2: For the integral result u 0J (t e ) performs central difference processing to obtain the characteristic zero-sequence voltage u 0W (t e ), as shown in the following formula:

[0019]

[0020] Where Δt is the time interval between adjacent sampling points.

[0021] Preferably, the specific process of step 3 is:

[0022] Step 3.1: Calculate the zero-sequence current i of each feeder 0h (t e ) is low-pass filtered to obtain the characteristic zero-sequence current i of each feeder lh (t e );

[0023] Step 3.2: For the characteristic zero-sequence voltage u in step 2 0W (t e ) and the characteristic zero-sequence current i of each feeder in step 3.1 lh (t e ) Extract the transient characteristic quantity of one quarter of the power frequency cycle and obtain the transient characteristic zero sequence voltage and transient characteristic zero-sequence current of each feeder Calculate the transient characteristic power Q of each feeder h , as shown in the following formula:

[0024]

[0025] Where T is a power frequency cycle, h=1,2,3…,n;

[0026] Step 3.3: Define the transient characteristic power ratio T of the hth feeder h is the transient characteristic power Q of the feeder h The ratio of the absolute value of the product of the transient characteristic powers of all feeders, h=1,2,3…,n, T h The specific calculation formula is shown as follows:

[0027]

[0028] When there is a transient characteristic power ratio T of a feeder h If the positive and negative signs are opposite to those of other feeders, it is determined that an arc high-resistance fault has occurred on the feeder.

[0029] Working principle of the invention

[0030] 1. Arc high-resistance ground fault equivalent circuit

[0031] When an arc high-resistance fault occurs, the feeder inductance, feeder resistance and arc suppression coil resistance can be ignored. Therefore, the zero-sequence equivalent network of the resonant grounding system can be simplified to Figure 2 As shown in the figure, L p is the arc suppression coil inductance. The system has n feeders in total, where the yth feeder is assumed to be a faulty feeder and m is a healthy feeder, m=1,2,…y-1,y+1,…,n, C 0h is the zero-sequence capacitance of the h-th feeder to the ground, h=1,2,…,n, R is the arc high-resistance grounding resistance; i 0h is the zero-sequence current of the h-th feeder, i 0f with i L They are respectively expressed as fault point current and neutral point current, u 0f and u0(t) represent the phase voltage at the fault point and the bus zero-sequence voltage respectively.

[0032] according to Figure 2 For the network shown, write the differential equation:

[0033]

[0034] Where, is the sum of the capacitance of each feeder to ground.

[0035] Combining the above formulas, we can construct the differential equation:

[0036]

[0037] The characteristic roots of the above equation are:

[0038]

[0039] When the system is in an underdamped state, R satisfies:

[0040]

[0041] At this time, the attenuation factor δ and the damped oscillation angular frequency ω f The following equation is satisfied.

[0042]

[0043] When a high-resistance ground fault occurs in a resonant grounding system, the system is generally in an underdamped state. 0f =U ms sin(ω0t1+θ),U ms and θ are the phase voltage amplitude and initial phase angle before the fault, respectively, and ω0 is the power frequency angular frequency. L for:

[0044]

[0045] Among them, B, A1, and A2 are the coefficients for solving the neutral point current differential equation. And the transient component of the neutral point current i L_z for:

[0046] i L_z e -δt (A1cos(ω f t1)+A2sin(ω f t1))

[0047] Derived i L The first and second derivatives of , the specific results are:

[0048]

[0049] i L The initial conditions are:

[0050] i L (0 - )=i L (0 + )=0

[0051] From the above derivation formula, and assuming t = 0, we can get:

[0052]

[0053] Calculation can be obtained:

[0054]

[0055] Furthermore, the busbar zero-sequence voltage u0(t) is obtained as:

[0056]

[0057] The busbar zero-sequence voltage transient component u 0_z (t) is:

[0058] u 0_z (t) = L p e -δt [(A2ωf -A1δ)cos(ω f t)-(A1ω f +A2δ)sin(ω f t)]

[0059] From this, the healthy feeder zero-sequence current i can be obtained 0m for:

[0060]

[0061] Where m = 1, 2,…y-1, y+1,…, n.

[0062] The transient component i of the zero-sequence current of the healthy feeder 0m_z for:

[0063]

[0064] Furthermore, based on Figure 2 The current flow direction can be calculated to obtain the zero-sequence current i of the fault feeder y 0y for:

[0065]

[0066] The transient component i of zero-sequence current of fault feeder y is 0y_z for:

[0067]

[0068] 2. Analysis of the applicability of the transient power direction method to high-resistance ground faults in resonant grounding systems

[0069] The basic principle of the transient current power direction method is: the zero-sequence current at the outlet of a healthy line has the same polarity as the derivative of the busbar zero-sequence voltage, and the polarity is opposite to that of the transient zero-sequence current at the outlet of the fault line. The line with a negative direction coefficient Q is selected as the fault line. Define the transient zero-sequence current i at the outlet of a certain line x as x (t) and the direction coefficient of the zero sequence voltage u0(t) is Q x , whose expression is:

[0070]

[0071] Assume that the transient power direction coefficient of the mth healthy line is Q m , the transient power direction coefficient of fault line y is Q y . We can get:

[0072]

[0073] Where m = 1, 2,…y-1, y+1,…, n.

[0074] Divide the above formula into two parts, where:

[0075]

[0076] Calculate Q separately y1 and Q y2 :

[0077]

[0078] From the above formula, we can know that when a single-phase grounding fault occurs in the distribution network, the positive and negative values ​​of the transient power of the fault feeder are related to A1, A2, attenuation factor δ and damped oscillation angular frequency ω. f The positive and negative polarity cannot be determined by this formula alone, so the transient power direction method may fail under certain working conditions.

[0079] 3. Kurtosis

[0080] Kurtosis is a statistic used to describe the characteristics of data distribution, reflecting the degree of deviation of the data distribution from the normal distribution. The calculation formula of the zero-sequence current kurtosis value of each feeder in the present invention is as follows:

[0081]

[0082] Where i j is the zero-sequence current sampling value corresponding to the j-th sampling point, n is the number of zero-sequence current sampling points within a quarter cycle after the fault occurs, μ is the mean value of the zero-sequence current within a quarter cycle after the fault occurs, σ is the standard deviation of the zero-sequence current within a quarter cycle after the fault occurs,

[0083] When the system is operating normally, the current waveform is a standard sine wave with a kurtosis value of 1.5. After a fault occurs, the waveform will experience a transient process, resulting in a sudden increase in the kurtosis value of the waveform. Due to the existence of three-phase asymmetric operation or noise in the actual system, the present invention considers a margin of 2 to 3 times and sets the value of β to 4.5. That is, when the kurtosis value of any feeder is greater than the set threshold β, the protection device is activated.

[0084] Compared with the prior art, the present invention has the following advantages:

[0085] 1) In terms of high-resistance fault detection and initiation: Traditional zero-sequence voltage initiation methods are insensitive to weak faults, require more than one cycle of sampling data for calculation, have slow response times, and are susceptible to false tripping due to transient overvoltage interference. In low-current grounding systems, they often fail due to insufficient zero-sequence voltage rise. Existing methods calculate the zero-sequence current kurtosis of each feeder within a quarter cycle after the fault, initiating the initiation if any kurtosis exceeds 4.5. This method accurately captures transient faults by exploiting the non-normal distortion of the current waveform, shortening response time and enhancing anti-interference capabilities. It is particularly suitable for low-current grounding systems, improving the protection's speed, reliability, and adaptability.

[0086] 2) Signal processing: The integration-then-differentiation approach for busbar zero-sequence voltage offers significant advantages. Integration accumulates the zero-sequence voltage energy of weak faults, converting small voltage amplitudes, such as those from high-resistance ground faults, into recognizable energy increments while suppressing high-frequency noise and reducing false trips. Differentiation captures the instantaneous rate of voltage change of a fault, amplifying transient mutations and enabling rapid response. This combination transcends traditional amplitude dependence, enhancing detection sensitivity and interference resistance, ensuring optimal protection sensitivity, speed, and reliability in scenarios such as high-resistance faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] Figure 1 This is a flow chart of the arc high resistance fault detection method of the present invention;

[0088] Figure 2 It is the zero-sequence equivalent network of the resonant grounding system in the present invention;

[0089] Figure 3 Schematic diagram of a simulation model of a 10kV resonant grounding system according to an embodiment of the present invention;

[0090] Figure 4 is the Emanuel arc operating condition model in an embodiment of the present invention;

[0091] Figure 5 (a) is the bus zero-sequence voltage u0(t e );

[0092] Figure 5 (b) is the trapezoidal sliding window integral u of the bus zero-sequence voltage in case 1 according to the embodiment of the present invention. 0J (t e );

[0093] Figure 5 (c) is the characteristic zero-sequence voltage u under Case 1 in the embodiment of the present invention 0W (t e );

[0094] Figure 6 (a) is the zero-sequence current waveform of each feeder in case 1 according to the embodiment of the present invention. 0h (t e );

[0095] Figure 6 (b) is the transient characteristic zero-sequence current of each feeder in case 1 according to the embodiment of the present invention. DETAILED DESCRIPTION

[0096] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0097] A new method for detecting arc high-resistance faults in distribution networks:

[0098] Step 1: Obtain the zero-sequence current i of each feeder 0h (t e ), h is the feeder number, h=1,2,3…,n, e is the sampling point number, calculate the kurtosis value K of the zero-sequence current of each feeder within a quarter cycle after the fault occurs h , when the kurtosis value of any feeder is greater than the set threshold β, the protection device is activated and then the process goes to step 2;

[0099] Step 2: Get the bus zero-sequence voltage u0(t e ), perform trapezoidal sliding window integration on the busbar zero-sequence voltage to obtain u 0J (t e ), and then the integrated bus zero sequence voltage u 0J (t e ) to perform differential calculation and obtain the characteristic zero-sequence voltage u 0W (t e );

[0100] Step 3: Calculate the zero-sequence current i of each feeder 0h (t e ) is low-pass filtered to obtain the characteristic zero-sequence current i of each feeder lh (t e ), and the characteristic zero-sequence voltage u in step 2 0W (t e ) Calculate the transient characteristic power ratio T of each feeder h , and according to the transient characteristic power ratio T h The positive and negative difference of the sensor can realize arc high resistance fault detection.

[0101] Specifically, the specific process of using the zero-sequence current of each feeder to construct the starting criterion in step 1 is as follows:

[0102] Step 1.1: Obtain the zero-sequence current i of each feeder 0h (t e ), h=1,2,3…,n, calculate the kurtosis value K of the zero-sequence current of each feeder within a quarter cycle after the fault occurs h , kurtosis value K h The specific calculation formula is shown as follows:

[0103]

[0104] Where i j is the zero-sequence current sampling value corresponding to the j-th sampling point, n is the number of zero-sequence current sampling points within a quarter cycle after the fault occurs, μ is the mean value of the zero-sequence current within a quarter cycle after the fault occurs, σ is the standard deviation of the zero-sequence current within a quarter cycle after the fault occurs,

[0105] Step 1.2: When the kurtosis value of any feeder is greater than the set threshold β, the protection device is activated and then the process goes to step 2.

[0106] The principle for selecting the threshold β is as follows: when the system is operating normally, the current waveform is a standard sine wave with a kurtosis value of 1.5; after a fault occurs, the waveform will experience a transient process, resulting in a sudden increase in the kurtosis value. Due to the existence of three-phase asymmetric operation or noise in the actual system, a margin of 2 to 3 times is considered here, and the value of β is set to 4.5, that is, there is a K h When >4.5, the protection device is activated.

[0107] Specifically, step 2 includes the following specific steps:

[0108] Step 2.1: Get the busbar zero-sequence voltage u0(t e ), a sliding trapezoid with a window size of 7 sampling points is used to integrate the bus zero-sequence voltage, as shown in the following formula:

[0109]

[0110] Where, f s is the sampling rate, k is the first sampling point in the sampling window, l=1,2,3,4,5 are the corresponding different sampling points in the sampling window, u0(t k ) is the busbar zero-sequence voltage sampling value corresponding to the kth sampling point in the sampling window, u0(t k+l ) is the busbar zero-sequence voltage sampling value corresponding to the k+1th sampling point to the k+5th sampling point in the sampling window, u0(t k+6 ) is the bus zero-sequence voltage sampling value corresponding to the k+6th sampling point in the sampling window.

[0111] Step 2.2: For the integral result u 0J (t e ) performs central difference processing to obtain the characteristic zero-sequence voltage u 0W (t e ), as shown in the following formula:

[0112]

[0113] Where Δt is the time interval between adjacent sampling points.

[0114] Specifically, the specific process of step 3 is:

[0115] Step 3.1: Calculate the zero-sequence current i of each feeder 0h (t e ) is low-pass filtered to obtain the characteristic zero-sequence current i of each feeder lh (te );

[0116] Step 3.2: For the characteristic zero-sequence voltage u in step 2 0W (t e ) and the characteristic zero-sequence current i of each feeder in step 3.1 lh (t e ) Extract the transient characteristic quantity of one quarter of the power frequency cycle and obtain the transient characteristic zero sequence voltage and transient characteristic zero-sequence current of each feeder Calculate the transient characteristic power Q of each feeder h , as shown in the following formula:

[0117]

[0118] Where T is a power frequency cycle, h = 1, 2, 3…, n.

[0119] Step 3.3: Define the transient characteristic power ratio T of the hth feeder h is the transient characteristic power Q of the feeder h The ratio of the absolute value of the product of the transient characteristic powers of all feeders, h=1,2,3…,n, T h The specific calculation formula is shown as follows:

[0120]

[0121] When there is a transient characteristic power ratio T of a feeder h If the positive and negative signs are opposite to those of other feeders, it is determined that an arc high-resistance fault has occurred on the feeder.

[0122] Example

[0123] like Figure 3 As shown, a neutral point arc suppression coil distribution network is built, and four feeders l1~l4 are set. The line parameters are shown in Table 1. The system sampling rate is 10kHz. The system adopts overcompensation with an overcompensation degree of 5%. The arc suppression coil inductance L p =0.547H, damping resistor r L =5.1557Ω, i1, i2, i3, i4 are the currents flowing through each feeder. Assume that an arc high-resistance fault occurs at 5 km from l1 at 0.2 s.

[0124] Table 1 Line specific parameters

[0125]

[0126] The Emanuel arc model is used to simulate high-resistance faults. The Emanuel arc model consists of two diodes connected to two anti-parallel DC sources and two asymmetric resistors, such as Figure 4 Shown: Two DC sources Vp 、V n and diode D p 、D n Forming positive and negative half-cycle current paths; two DC sources V p 、V n The simulated arc voltage, whose value depends on the voltage level and asymmetry modeling of the system, varies randomly and independently according to time. s >V p When V s <V n When the current reverses, V n <V s <V p During this time, no current flows. p and V n The size will increase the randomness and arc extinction time of asymmetric faults. At the same time, through the two nonlinear resistors R p and R n Different values ​​are taken to simulate asymmetric current. The specific parameters are shown in Table 2.

[0127] Table 2 Parameters of three types of arc models

[0128] arc <![CDATA[V p / kV]]> <![CDATA[V n / kV]]> <![CDATA[R p / Oh]]> <![CDATA[R n / Oh]]> Step size / ms Case 1 2.2±20.45％ 2.0±20％ 800 750 0.1 Case 2 3.8±10％ 3.6±10％ 400 350 0.1 Case 3 1.2±10％ 1.0±10％ 1200 1100 0.1

[0129] Taking the arc model in Case 1 as an example, first calculate the bus zero-sequence voltage u0(t e ) Perform trapezoidal sliding window integration to obtain u 0J (t e ),like Figure 5 As shown in (b), it can be seen from the figure that the zero-sequence voltage energy of a weak fault accumulates, and the small voltage is converted into a recognizable energy increment. Then the integrated bus zero-sequence voltage u 0J (t e ) to perform differential calculation and obtain the characteristic zero-sequence voltage u 0W (t e ),like Figure 5 As shown in (c), it can be seen that the differential captures the voltage change rate at the moment of the fault and amplifies the transient mutation characteristics. 0h (t e ) to perform low-pass filtering to obtain the transient characteristic zero-sequence current of each feeder like Figure 6 As shown in (b), it can be seen from the figure that the characteristic frequency band filtering minimizes signal distortion, so that the high-frequency noise and low-frequency interference in the zero-sequence current of each feeder are filtered out, providing pure and effective characteristic quantity input for subsequent fault detection.

[0130] Using the transient characteristic zero sequence voltage and transient characteristic zero-sequence current of each feeder Calculate the transient characteristic power ratio T of each feeder h , calculate the transient characteristic power ratio T of each feeder of the above three types of arc models respectively h , the results are shown in Table 3.

[0131] Table 3 Detection of three types of arc high resistance faults

[0132] arc <![CDATA[T1 / var]]> [T2 / var] <![CDATA[T3 / var]]> <![CDATA[T4 / var]]> Testing results Case 1 -2.3196 0.0316 0.8306 0.4329 efficient Case 2 -35.1973 0.2325 7.1184 3.4210 efficient Case 3 -1.1796 0.0190 0.4898 0.2568 efficient

[0133] As shown in Table 3, in the three types of arc high-resistance fault detection cases, the transient characteristic power ratio T1 of feeder 1 is negative, while the transient characteristic power ratios of other feeders are positive. It can be determined that feeder 1 is the faulty feeder, and accurate detection of arc high-resistance faults can be achieved.

Claims

1. A new method for detecting arc high-resistance faults in distribution networks, characterized by: Step 1: Obtain the zero-sequence current i of each feeder 0h (t e ), h is the feeder number, h=1,2,3…,n, e is the sampling point number, calculate the kurtosis value K of the zero-sequence current of each feeder within a quarter cycle after the fault occurs h , when the kurtosis value of any feeder is greater than the set threshold β, the protection device is activated and then the process goes to step 2; Step 2: Get the bus zero-sequence voltage u0(t e ), perform trapezoidal sliding window integration on the busbar zero-sequence voltage to obtain u 0J (t e ), and then the integrated bus zero sequence voltage u 0J (t e ) to perform differential calculation and obtain the characteristic zero-sequence voltage u 0W (t e ); Step 3: Calculate the zero-sequence current i of each feeder 0h (t e ) is low-pass filtered to obtain the characteristic zero-sequence current i of each feeder lh (t e ), and the characteristic zero-sequence voltage u in step 2 0W (t e ) Calculate the transient characteristic power ratio T of each feeder h , and according to the transient characteristic power ratio T h The positive and negative difference of the sensor can realize arc high resistance fault detection.

2. A new method for detecting arc high-resistance faults in distribution networks according to claim 1, characterized in that: The specific process of using the zero-sequence current of each feeder to construct the starting criterion in step 1 is as follows: Step 1.1: Obtain the zero-sequence current i of each feeder 0h (t e ), h=1,2,3…,n, calculate the kurtosis value K of the zero-sequence current of each feeder within a quarter cycle after the fault occurs h , kurtosis value K h The specific calculation formula is shown as follows: Where i j is the zero-sequence current sampling value corresponding to the j-th sampling point, n is the number of zero-sequence current sampling points within a quarter cycle after the fault occurs, μ is the mean value of the zero-sequence current within a quarter cycle after the fault occurs, σ is the standard deviation of the zero-sequence current within a quarter cycle after the fault occurs, Step 1.2: When the kurtosis value of any feeder is greater than the set threshold β, the protection device is activated and then the process goes to step 2; The principle for selecting the threshold β is as follows: when the system is operating normally, the current waveform is a standard sine wave with a kurtosis value of 1.5; after a fault occurs, the waveform will experience a transient process, resulting in a sudden increase in the kurtosis value. Due to the existence of three-phase asymmetric operation or noise in the actual system, a margin of 2 to 3 times is considered here, and the value of β is set to 4.5, that is, there is a K h When the temperature is >4.5, the protection device is activated.

3. A new method for detecting arc high-resistance faults in distribution networks according to claim 1, characterized in that: The step 2 includes the following specific steps: Step 2.1: Get the busbar zero-sequence voltage u0(t e ), a sliding trapezoid with a window size of 7 sampling points is used to integrate the bus zero-sequence voltage, as shown in the following formula: Where, f s is the sampling rate, k is the first sampling point in the sampling window, l=1,2,3,4,5 are the corresponding different sampling points in the sampling window, u0(t k ) is the busbar zero-sequence voltage sampling value corresponding to the kth sampling point in the sampling window, u0(t k+l ) is the busbar zero-sequence voltage sampling value corresponding to the k+1th sampling point to the k+5th sampling point in the sampling window, u0(t k+6 ) is the bus zero-sequence voltage sampling value corresponding to the k+6th sampling point in the sampling window; Step 2.2: For the integral result u 0J (t e ) performs central difference processing to obtain the characteristic zero-sequence voltage u 0W (t e ), as shown in the following formula: Where Δt is the time interval between adjacent sampling points.

4. A new method for detecting arc high-resistance faults in distribution networks according to claim 1, characterized in that: The specific process of step 3 is as follows: Step 3.1: Calculate the zero-sequence current i of each feeder 0h (t e ) is low-pass filtered to obtain the characteristic zero-sequence current i of each feeder lh (t e ); Step 3.2: For the characteristic zero-sequence voltage u in step 2 0W (t e ) and the characteristic zero-sequence current i of each feeder in step 3.1 lh (t e ) Extract the transient characteristic quantity of one quarter of the power frequency cycle and obtain the transient characteristic zero sequence voltage and transient characteristic zero-sequence current of each feeder Calculate the transient characteristic power Q of each feeder h , as shown in the following formula: Where T is a power frequency cycle, h=1,2,3…,n; Step 3.3: Define the transient characteristic power ratio T of the hth feeder h is the transient characteristic power Q of the feeder h The ratio of the absolute value of the product of the transient characteristic powers of all feeders, h=1,2,3…,n, T h The specific calculation formula is shown as follows: When there is a transient characteristic power ratio T of a feeder h If the positive and negative signs are opposite to those of other feeders, it is determined that an arc high-resistance fault has occurred on the feeder.

Images