An arc fault identification method, computer device and readable storage medium

Abstract

This invention discloses an arc fault identification method, computer device, and readable storage medium, relating to the field of power detection. The method includes: acquiring power supply voltage, load voltage, and load current; calculating the arc voltage based on the power supply voltage and load voltage, and calculating the arc voltage energy value in conjunction with the sampling frequency; if the arc voltage energy value is greater than a set energy threshold, an arc is determined to exist, and an alarm message indicating the presence of an arc is generated; otherwise, an arc is determined not to exist; when an arc exists, the intensity of the arc is determined using a dual-channel one-dimensional convolutional neural network based on the arc voltage and load voltage; if the arc intensity is no arc, the alarm message is canceled; if the arc intensity is a slight arc or a severe arc, an alarm message indicating the intensity of the arc is generated, and the fault load type is determined using a two-dimensional convolutional neural network based on the load current. This invention improves the efficiency and accuracy of arc identification, and can further identify the intensity of the arc and the fault load type.

Description

Technical Field

[0001] This invention relates to the field of power detection, and in particular to an arc fault identification method, a computer device, and a readable storage medium. Background Technology

[0002] With the continuous development of power systems, the types of household appliances are increasing, and their power ratings vary. As these appliances age, the risk of electric arcing increases dramatically, easily leading to fires, property damage, and personal injury. Currently, existing arc detection and protection devices suffer from false alarms and missed alarms, resulting in a very low overall accuracy. These problems stem from the highly random nature of series arc faults and the fact that arc detection algorithms do not fully consider the physical characteristics of electric arcs.

[0003] Currently, arc detection primarily relies on single current waveform feature analysis, depending on the high-frequency current components and irregular distortions caused by the arc. However, using a single current signal cannot guarantee the highest accuracy of the detection results. Furthermore, the current waveforms of different types of loads vary significantly, making it difficult to quickly determine the presence of an arc by extracting a uniform feature. Using arc voltage information, on the other hand, easily provides a uniform feature for rapid arc detection. However, due to the randomness of arc fault location, arc voltage is generally not directly measurable, requiring indirect methods to calculate it by measuring the power supply voltage and load voltage. In addition to determining whether an arc fault has occurred, identifying the load type is also crucial. Existing methods typically integrate arc detection and load classification into a single algorithm, slowing down arc identification. Determining the presence of an arc is critical for safety protection and should have a significantly higher priority than load classification; moreover, different protective measures should be implemented based on the severity of different arc faults. In conclusion, current arc recognition algorithms are not perfect and need to incorporate new measurement data and employ simpler and more effective algorithms to ensure the speed of arc recognition.

[0004] Therefore, it is necessary to design a new arc detection method to improve the overall recognition accuracy and speed, and to determine the intensity of the arc, so as to maximize electrical safety. Summary of the Invention

[0005] The purpose of this invention is to provide an arc fault identification method, computer device, and readable storage medium, which can improve the efficiency and accuracy of arc identification, and further identify the intensity of the arc and the type of fault load.

[0006] To achieve the above objectives, the present invention provides the following solution:

[0007] An arc fault identification method, comprising:

[0008] Collect the power supply voltage, load voltage, and load current of the electrical appliance during operation according to the set sampling frequency and set sampling time.

[0009] Calculate the arc voltage based on the power supply voltage and the load voltage;

[0010] The arc voltage energy value is calculated based on the arc voltage and the sampling frequency.

[0011] Determine whether the arc voltage energy value is greater than the set energy threshold; if yes, determine that an arc exists and generate an alarm message indicating the presence of an arc; if no, determine that no arc exists and continue to collect the power supply voltage, load voltage and load current when the appliance is working.

[0012] When an electric arc is present, the intensity of the electric arc is determined by a pre-trained dual-channel one-dimensional convolutional neural network based on the electric arc voltage and the load voltage; the intensity of the electric arc is defined as no electric arc, slight electric arc, or severe electric arc.

[0013] If the arc intensity is no arc, the alarm information is canceled, and the power supply voltage, load voltage, and load current of the electrical appliance during operation are collected. If the arc intensity is a slight arc or a severe arc, an alarm information for the arc intensity is generated, and the fault load type is determined using a pre-trained two-dimensional convolutional neural network based on the load current.

[0014] To achieve the above objectives, the present invention provides the following solution:

[0015] A computer device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described arc fault identification method.

[0016] To achieve the above objectives, the present invention provides the following solution:

[0017] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described arc fault identification method.

[0018] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects: The present invention integrates power supply voltage, load voltage, and load current, separates the process of arc detection from the process of fault load classification, and performs them independently. The presence of an arc can be quickly and accurately determined by the arc voltage energy value alone. At the same time, the load voltage and a one-dimensional convolutional neural network are used for secondary detection, which can not only identify the intensity of the arc, but also further ensure the accuracy of arc identification. The most time-consuming and relatively unimportant fault load classification process is performed independently, and a two-dimensional convolutional neural network is used to identify the fault load type, which can not only identify the fault load type, but also further improve the efficiency of arc identification. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart of the arc fault identification method provided by the present invention;

[0021] Figure 2 This is a schematic diagram of the data acquisition component;

[0022] Figure 3 This is a waveform diagram of arc voltage in the absence of an electric arc;

[0023] Figure 4 The waveform of the arc voltage when an electric arc exists;

[0024] Figure 5 This is a histogram showing the distribution of arc voltage energy values.

[0025] Figure 6 The waveform of the load voltage without electric arc;

[0026] Figure 7 The load voltage waveform is for a slight electric arc.

[0027] Figure 8 The load voltage waveform is for a violent electric arc.

[0028] Figure 9 This is a schematic diagram of a dual-channel one-dimensional convolutional neural network.

[0029] Figure 10 This is a waveform diagram of the load current when the resistor is in normal short-term condition.

[0030] Figure 11 This is a waveform diagram of the load current during a short-time resistor fault.

[0031] Figure 12 This is a waveform diagram of the load current when the regulator is operating normally for a short period of time.

[0032] Figure 13 This is a waveform diagram of the load current during a short-term fault of the regulator.

[0033] Figure 14 The waveform of the load current when the electric drill is operating normally;

[0034] Figure 15 The waveform of the load current when the electric drill malfunctions;

[0035] Figure 16 This is a waveform diagram of the load current when the switching power supply is in normal operation for a short period of time.

[0036] Figure 17 The waveform of the load current during a short-time fault in the switching power supply;

[0037] Figure 18 This is a waveform diagram of the load current when the fluorescent lamp is operating normally for a short period of time.

[0038] Figure 19 The waveform of the load current during a short-term fault of a fluorescent lamp;

[0039] Figure 20 This is a schematic diagram of the MarKov transform method. Detailed Implementation

[0040] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] The purpose of this invention is to provide an arc fault identification method, computer device, and readable storage medium. The method prioritizes the tasks involved in arc fault identification into three priorities: arc presence identification, arc intensity identification, and fault load type identification. This greatly improves the speed of arc identification, detects dangers immediately, and issues alarm information. Then, lower priority identification tasks are performed to facilitate subsequent fault troubleshooting.

[0042] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0043] Example 1

[0044] like Figure 1As shown, this embodiment provides a method for identifying electric arc faults, including:

[0045] Step 1: Collect the power supply voltage, load voltage, and load current of the electrical appliance during operation according to the set sampling frequency and the set sampling time.

[0046] This invention simultaneously collects the power supply voltage, load voltage, and load current of household 220V AC appliances during operation. The data collection method can be as follows: Figure 2 As shown, the data acquisition components include: a Hall current sensor, a voltage sensor, an arc simulation device, and an oscilloscope. The sensor acquires signals every 0.5 seconds at a sampling rate of 10 kHz, with each acquisition lasting 120 ms, or 6 cycles of data. The acquired power supply voltage, load voltage, and load current are then subjected to sliding window filtering to remove noise, resulting in filtered power supply voltage, load voltage, and load current with more distinct characteristics.

[0047] Step 2: Calculate the arc voltage based on the power supply voltage and the load voltage.

[0048] Specifically, the arc voltage is obtained by subtracting the load voltage from the power supply voltage. The waveform of the arc voltage is as follows: Figure 3 and Figure 4 As shown, the specific calculation formula is as follows:

[0049] V arc (i)=V pow (i)-V load (i), i = 1, 2, ..., n;

[0050] Among them, V arc (i) represents the arc voltage at the i-th sampling point, V pow (i) represents the power supply voltage at the i-th sampling point, V load (i) represents the load voltage of the i-th sampling point, and n represents the number of sampling points. In this invention, n is 1200.

[0051] Step 3: Calculate the arc voltage energy value based on the arc voltage and the sampling frequency.

[0052] Specifically, the arc voltage energy value is calculated using the following formula:

[0053]

[0054] Where E is the arc voltage energy value and f is the sampling frequency, i.e., f = 10KHz.

[0055] Step 4: Determine whether the arc voltage energy value is greater than the set energy threshold; if yes, determine that an arc exists and generate an alarm message indicating the presence of an arc; if no, determine that no arc exists and continue to collect the power supply voltage, load voltage and load current when the appliance is working.

[0056] Specifically, firstly, all arc voltage energy values ​​during the experiment were determined experimentally. Then, statistical analysis was performed on all arc voltage energy values ​​during the experiment to construct a histogram of arc voltage energy value distribution, such as... Figure 5 As shown, the energy threshold is thus set to 7e. -5 .

[0057] By detecting the energy value of the electric arc voltage, it is possible to quickly determine whether there is an electric arc. This operation requires little computation and can identify the electric arc at the fastest speed.

[0058] Step 5: When an electric arc is present, the intensity of the electric arc is determined using a pre-trained dual-channel one-dimensional convolutional neural network based on the arc voltage and the load voltage. The intensity of the electric arc is defined as no arc, a slight arc, or a severe arc.

[0059] Specifically, the waveform of the load voltage is as follows: Figures 6 to 8 As shown, the arc voltage and load voltage are fed into a trained dual-channel one-dimensional convolutional neural network for inference and classification. The structure of the dual-channel one-dimensional convolutional neural network is as follows: Figure 9 As shown, the circuit consists of sequentially connected one-dimensional convolutions, one-dimensional max pooling, one-dimensional convolutions, one-dimensional max pooling, one-dimensional convolutions, one-dimensional max pooling, and two fully connected layers. ReLU activation functions are used between the one-dimensional convolutions and one-dimensional max pooling, and between the two fully connected layers. The SoftMax activation function is used after the last fully connected layer to output the intensity of the electric arc. Figure 9 In this context, k represents the kernel size, p represents the receptive field size, and s represents the stride.

[0060] First, the arc voltage and the load voltage are subjected to sliding window filtering respectively to obtain the filtered arc voltage and the filtered load voltage. Then, the filtered arc voltage and the filtered load voltage are input into a pre-trained dual-channel one-dimensional convolutional neural network to determine the intensity of the arc. The load voltage and arc voltage collected by the sensor are both discrete signals with 1200 sampling points. After sliding window filtering, the two one-dimensional signals are fused at the input layer and then fed into the one-dimensional convolutional neural network for classification. The intensity of the arc is output at the output layer.

[0061] Because the load voltage is distorted to varying degrees depending on the intensity of the electric arc, the more intense the arc, the more severe the load voltage distortion, resulting in a waveform similar to a square wave. Therefore, the intensity of the electric arc can be determined by detecting the square wave of the load voltage.

[0062] Step 6: If the arc intensity is no arc, cancel the alarm information and continue to collect the power supply voltage, load voltage and load current when the appliance is working; if the arc intensity is a slight arc or a severe arc, generate an alarm information for the arc intensity, and determine the fault load type using a pre-trained two-dimensional convolutional neural network based on the load current.

[0063] Different types of loads have different current waveforms, so this difference can be used to classify faulty loads using convolutional neural networks.

[0064] Specifically, based on the load current, a pre-trained two-dimensional convolutional neural network is used to determine the fault load type, including:

[0065] (1) Perform sliding window filtering on the load current to obtain the filtered load current.

[0066] (2) The filtered load current is randomly and continuously cropped to obtain the cropped load current. Specifically, the filtered load current containing six consecutive cycles is randomly and continuously cropped into a single data set containing only two consecutive cycles. The load current waveform before cropping is as follows: Figures 10 to 19 As shown.

[0067] (3) Convert the cropped load current into a color two-dimensional image. In this invention, the MarKov transform method is used to convert the cropped load current into a color two-dimensional image.

[0068] like Figure 20 As shown, the specific process of the MarKov transform method is as follows:

[0069] For a one-dimensional time-domain signal sequence of length N (the cropped load current) X = {x1, x2, ..., x...} N First, divide the distribution of X values ​​into Q discrete quantile units, and then use the quantile q... j (q j Label x ∈[1,Q]) and then x i Mapped to the corresponding quantile q j Above. Among them, x i Let be the load current at the i-th sampling point.

[0070] Transform a continuous time series into a quantile sequence {q} expressed in quantiles. j1 ,q j2 ,...,q jN The specific transition probabilities are obtained by statistically analyzing the changes in the quantile sequence at each time point.

[0071] Arrange all transition probabilities of the Markov chain along the transition rules to construct a Q×Q adjacent weighted matrix W (Markov transition probability matrix):

[0072]

[0073] Based on the above Markov transition probability matrix, the following Markov Transition Field (MTF) is derived:

[0074]

[0075] Where M is the Markov transformation matrix, w ij Let be the multi-step transition probability, indicating that at this moment, q is the transition probability. j Elements in the quantile region undergo multiple transitions to q in the next time step. i The probability of the quantile region.

[0076] The MTF values ​​are directly mapped to a pseudo-color image. The magnitude of each element in the Markov transformation matrix represents the intensity of a pixel's color, and the intensity of a pixel's color represents the probability of transitioning from one quantile to another.

[0077] (4) Input the color two-dimensional image into a pre-trained two-dimensional convolutional neural network to determine the fault load type. In this invention, the two-dimensional convolutional neural network uses ResNet18.

[0078] As one specific implementation, the fault load types include: resistor normal, resistor fault, regulator normal, regulator fault, electric drill normal, electric drill fault, switching power supply normal, switching power supply fault, fluorescent lamp normal, and fluorescent lamp fault.

[0079] Furthermore, if the fault load type determination result is that all loads are normal, i.e., the resistor is normal, the regulator is normal, the electric drill is normal, the switching power supply is normal, and the fluorescent lamp is normal, then the alarm information is canceled, and the power supply voltage, load voltage, and load current when the electrical appliance is working are collected.

[0080] If the fault load type determination result is any load fault, namely resistor fault, regulator fault, electric drill fault, switching power supply fault, or fluorescent lamp fault, then an alarm message for the fault load type is generated, and the power supply voltage, load voltage, and load current during the operation of the electrical appliance continue to be collected. The training samples for the two-dimensional convolutional neural network during the training process are shown in Table 1.

[0081] Table 1 Training Samples and Labels

[0082] Training samples Label Color image with normal resistance 0 Color image of resistor fault 1 Regulator normal color image 2 Color image of regulator malfunction 3 Normal color image of an electric drill 4 Color images of electric drill malfunctions 5 Normal color image of switching power supply 6 Color images of switching power supply failures 7 Normal color image under fluorescent light 8 Color images of fluorescent lamp malfunctions 9

[0083] In summary, this invention can achieve three functions: 1) quickly determining the presence of an electric arc; 2) determining the intensity of the electric arc; and 3) classifying the load type of the fault. The arc detection process is performed in segments, using minimal computation to detect the arc as quickly as possible and trigger an alarm; then, the intensity of the arc is determined, and the alarm level is output; finally, the load type is classified.

[0084] This invention introduces new measurement information, namely power supply voltage and load voltage, and achieves rapid arc determination by analyzing arc voltage. The arc identification task is completed step-by-step through priority classification, enabling the determination of arc presence, arc intensity, and load type. The advantages are that it can quickly identify the presence and intensity of an arc after it occurs, and ultimately classify the load type.

[0085] The arc fault identification method algorithm provided by this invention is written in Python. Based on the steps described above, by inputting the acquired raw waveform CSV file, it can complete three steps: arc presence / absence identification, arc intensity identification, and fault load type identification. Experimental analysis, as shown in Table 2, yielded the following results: the arc presence / absence identification speed is significantly improved compared to algorithms that integrate arc identification and load classification; the accuracy rates for the three stages reach 100%, 98.43%, and 99.64%, respectively.

[0086] Table 2 Experimental Analysis Results

[0087] Priority Job title Average time accuracy 1 Arc presence / absence identification 240ms 100％ 2 Arc intensity identification 5s149ms 98.43％ 3 Fault load type identification 5s376ms 99.64％

[0088] Compared to determining the type of electric arc and faulty load solely based on current, this invention integrates information from multiple sensors: power supply voltage, load voltage, and load current. The determination of the faulty arc and the classification of the faulty load are separated and performed independently. The presence of an arc can be quickly and accurately determined solely by the energy value of the arc voltage, while secondary detection using the load voltage further ensures the accuracy of identification. Isolating the most time-consuming and relatively unimportant faulty load classification further improves the efficiency of arc identification.

[0089] Example 2

[0090] A computer device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the arc fault identification method of Embodiment 1.

[0091] Example 3

[0092] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the arc fault identification method of Embodiment 1.

[0093] Example 4

[0094] A computer program product includes a computer program that, when executed by a processor, implements the arc fault identification method of Embodiment 1.

[0095] Example 5

[0096] A computer device, which may be a database, includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage medium. The database stores pending transactions. The I / O interfaces facilitate information exchange between the processor and external devices. The communication interface allows communication with external terminals via a network connection. When executed by the processor, the computer program implements the arc fault identification method described in Embodiment 1.

[0097] It should be noted that the object information (including but not limited to object device information, object personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this invention are all information and data authorized by the object or fully authorized by all parties, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0098] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided by this invention can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided by this invention may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided by this invention may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0099] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0100] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for identifying electric arc faults, characterized in that, The arc fault identification method includes: Collect the power supply voltage, load voltage, and load current of the electrical appliance during operation according to the set sampling frequency and set sampling time. Calculate the arc voltage based on the power supply voltage and the load voltage; The arc voltage energy value is calculated based on the arc voltage and the sampling frequency. Determine whether the arc voltage energy value is greater than the set energy threshold; if yes, determine that an arc exists and generate an alarm message indicating the presence of an arc; if no, determine that no arc exists and continue to collect the power supply voltage, load voltage and load current when the appliance is working. When an electric arc is present, the intensity of the electric arc is determined by a pre-trained dual-channel one-dimensional convolutional neural network based on the electric arc voltage and the load voltage; the intensity of the electric arc is defined as no electric arc, slight electric arc, or severe electric arc. If the arc intensity is no arc, the alarm information is canceled, and the power supply voltage, load voltage, and load current of the electrical appliance during operation are collected. If the arc intensity is a slight arc or a severe arc, an alarm information for the arc intensity is generated, and the fault load type is determined using a pre-trained two-dimensional convolutional neural network based on the load current.

2. The arc fault identification method according to claim 1, characterized in that, The arc voltage is calculated using the following formula: V arc (i)=V pow (i)-V load (i),i=1,2,...,n; Among them, V arc (i) represents the arc voltage at the i-th sampling point, V pow (i) represents the power supply voltage at the i-th sampling point, V load (i) represents the load voltage at the i-th sampling point, and n represents the number of sampling points.

3. The arc fault identification method according to claim 1, characterized in that, The arc voltage energy value is calculated using the following formula: Where E is the arc voltage energy value, V arc (i) represents the arc voltage at the i-th sampling point, f represents the sampling frequency, and n represents the number of sampling points.

4. The arc fault identification method according to claim 1, characterized in that, Based on the arc voltage and the load voltage, a pre-trained dual-channel one-dimensional convolutional neural network is used to determine the intensity of the arc, specifically including: The arc voltage and the load voltage are respectively subjected to sliding window filtering to obtain the filtered arc voltage and the filtered load voltage; The filtered arc voltage and the filtered load voltage are input into a pre-trained dual-channel one-dimensional convolutional neural network to determine the intensity of the arc.

5. The arc fault identification method according to claim 1, characterized in that, Based on the load current, a pre-trained two-dimensional convolutional neural network is used to determine the fault load type, specifically including: The load current is subjected to sliding window filtering to obtain the filtered load current. The filtered load current is randomly and continuously trimmed to obtain the trimmed load current; The cropped load current is converted into a color two-dimensional image; The color two-dimensional image is input into a pre-trained two-dimensional convolutional neural network to determine the fault load type.

6. The arc fault identification method according to claim 5, characterized in that, The MarKov transform method is used to convert the cropped load current into a color two-dimensional image.

7. The arc fault identification method according to claim 1, characterized in that, The fault load types include: resistor normal, resistor fault, regulator normal, regulator fault, electric drill normal, electric drill fault, switching power supply normal, switching power supply fault, fluorescent lamp normal, and fluorescent lamp fault.

8. The arc fault identification method according to claim 1, characterized in that, The arc fault identification method further includes: If the fault load type determination result is that all loads are normal, then the alarm information is canceled, and the power supply voltage, load voltage and load current when the electrical appliance is working are collected again. If the fault load type determination result is any type of load fault, an alarm message for the fault load type will be generated, and the power supply voltage, load voltage and load current during the operation of the electrical appliance will continue to be collected.

9. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the arc fault identification method according to any one of claims 1-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the arc fault identification method according to any one of claims 1-8.

Images