Coiflet wavelet relative spectral entropy line selection method and system for single-phase earth fault

Abstract

The invention provides a Coiflet wavelet relative spectral entropy line selection method for a single-phase earth fault. The method specifically comprises the following steps: S1, collecting zero-sequence current of each line connected with a bus; s2, performing Coiflet wavelet decomposition on the zero-sequence current of each line connected with the bus to obtain wavelet coefficients of each line in different frequency bands, and determining a fault occurrence moment according to change characteristics of the wavelet coefficients in different frequency bands; s3, taking the fault occurrence moment as a demarcation point, respectively calculating energy spectrums of different characteristic frequency bands of each line before and after the fault, and calculating a comprehensive relative spectrum entropy of the different characteristic frequency bands of each line before and after the fault; and S4, taking different characteristic frequency bands as different analysis scales, selecting a fault line in each analysis scale according to the comprehensive relative spectral entropy of each line, and realizing single-phase earth fault line selection. The method comprehensively considers the transient and steady state characteristics before and after the fault time, and improves the adaptability and accuracy of line selection.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network fault line selection, and in particular to a Coiflet wavelet relative spectral entropy line selection method and system for single-phase grounding fault. Background Art

[0002] With the continued growth of comprehensive power load demand and the rise of distributed power generation technology, distribution networks are becoming increasingly complex and randomized. This makes distribution network lines and equipment susceptible to various faults during operation, which not only impacts the stability of power system operation but also easily damages power-consuming equipment. my country's distribution networks generally use low-current grounding, and single-phase ground faults account for over 80% of all faults. While a fault can continue for 1-2 hours, improving power supply reliability, the voltage in the non-faulty phase will increase by √3 times its original value, impacting insulation and potentially causing the fault to expand, impacting the safe and stable operation of distribution network equipment and the reliability of the power supply. Therefore, timely identification of faulty lines is crucial.

[0003] At present, the line selection methods for single-phase grounding faults in distribution networks are mainly divided into two categories: steady-state line selection method and transient line selection method.

[0004] (1) Steady-state line selection method

[0005] The steady-state line selection method is a method for selecting the fault line based on the steady-state response of the system after the fault. It generally uses the magnitude and direction of the zero-sequence current of each line after the fault to compare and judge, including the zero-sequence current fundamental amplitude and phase ratio method, the zero-sequence active component direction method, the zero-sequence current harmonic amplitude and phase ratio method, the negative-sequence current method, etc.

[0006] Zero-sequence current fundamental amplitude and phase comparison method: This method compares the magnitude and phase of the fundamental component of the zero-sequence current on each line to select the fault line. The line with the largest amplitude or the opposite phase to the zero-sequence voltage is the fault line. This method is suitable for systems with ungrounded neutral points. While its principle is simple, it cannot eliminate the influence of current transformer imbalance and is susceptible to system operating mode, line length, transition resistance, and other factors, leading to misselection, multiple selection, or missed selection. It cannot meet the requirements of highly variable system operation.

[0007] Zero-sequence active component direction method: This method uses the direction of each line's zero-sequence active component to select a line. The direction of the active current is related to the location of the fault point. Therefore, after a fault occurs, the active current direction of the faulty line will differ from that of the normal line. This method is applicable to systems grounded via arc suppression coils and is not affected by arc suppression coils. However, it is also susceptible to factors such as system operation mode, transition resistance, and current transformer characteristics, which may lead to misjudgment or missed detection.

[0008] Zero-sequence current harmonic amplitude and phase comparison method: This method uses the harmonic components of each line's zero-sequence current to select the line. The amplitude and direction of the harmonic current in the faulty line differ significantly from those in the non-faulty line. This method is applicable to both ungrounded and arc-suppression coil-grounded systems, and offers good anti-interference capabilities and accuracy. However, it is susceptible to factors such as system operating mode, transition resistance, and current transformer characteristics.

[0009] Negative-sequence current method: This method uses the phase and magnitude of negative-sequence current to select a line. Negative-sequence current only exists in faulted lines, and its phase is 180 degrees out of phase with the zero-sequence current of healthy lines. Therefore, by comparing the magnitude and direction of negative-sequence currents on each line, the faulty line can be identified. However, this method is only applicable when current transformers are connected in a full star configuration and is less effective for selecting lines with low zero-sequence currents.

[0010] (2) Transient line selection method

[0011] The transient line selection method is a method of selecting the fault line based on the transient characteristics of the system after the fault. Usually, the transient component of the fault line is significantly greater than that of the non-fault line. Common transient line selection methods include active power method, wavelet method and energy method.

[0012] Active power method: The faulty line is identified by comparing the changes in active power of each line at the moment of the fault. When a single-phase ground fault occurs, the current in the faulty line suddenly increases, accompanied by a decrease in voltage, which causes a significant change in the active power of that line. Non-fault lines, on the other hand, experience relatively small changes in current and voltage, resulting in smaller changes in active power. Therefore, by comparing the changes in active power of each line, the faulty line can be identified. This method is simple in principle, easy to implement, and unaffected by arc suppression coils, making it highly applicable. However, when the system is large or the network structure is complex, the active power changes of each line may interfere with each other, resulting in inaccurate line selection. Furthermore, this method places high demands on the sampling accuracy and synchronization of current and voltage, requiring high-precision measurement equipment and data processing technology.

[0013] Wavelet method: This method uses wavelet transforms to perform multi-scale analysis on transient zero-sequence currents, extracting characteristic parameters of the fault line, such as modulus maxima and energy distribution. By comparing these characteristic parameters, the fault line can be identified. This method is applicable to both ungrounded and arc-suppression coil-grounded systems. It effectively extracts sudden changes and non-stationary features in transient signals, improving the accuracy and reliability of line selection. However, it is important to be aware of its limitations and select appropriate equipment and sampling techniques to improve line selection accuracy and reliability.

[0014] The energy method uses the difference in energy distribution between faulty and non-faulty lines after a fault to select a line. Specifically, when a single-phase ground fault occurs, the current in the faulty line suddenly increases, and its energy change differs from that of the non-faulty line. By comparing the energy changes of each line, the faulty line can be identified. This method has the advantage of not requiring complex mathematical transformations and signal processing of the current signal, making it suitable for systems with arc suppression coils. However, the energy method requires high accuracy and synchronization of the current transformers, and is less applicable in large-scale systems or complex network structures.

[0015] In practical applications, the energy line selection method can be combined with other methods to improve the accuracy and reliability of line selection. Existing research often combines the energy method with wavelet transforms to form a comprehensive line selection strategy, which can reduce the influence of sampling asynchrony, transition resistance, fault location, and fault angle. However, most of these methods only analyze the transient zero-sequence current characteristics after the fault, and use the maximum energy modulus value of the characteristic frequency band and its polarity opposite to that of the non-fault line as the line selection criterion to select the fault line. In actual fault processes, the energy relationship and polarity characteristics of the fault line and non-fault line may not hold. In particular, in cases such as the reverse polarity of the zero-sequence CT, the incorrect zero-sequence CT ratio setting, or the lack of CT connection on some non-fault lines, this can lead to misjudgment or missed detection.

[0016] In summary, both steady-state and transient line selection methods are based on post-fault characteristic components, offering high accuracy and reliability and applicable to various grounding systems. However, these methods are susceptible to sampling accuracy, CT polarity, CT imbalance, system size, and unbalanced load current, often leading to misjudgments or missed faults.

[0017] In view of this, the present invention proposes a Coiflet wavelet relative spectral entropy line selection method and system for single-phase grounding fault. Summary of the Invention

[0018] The purpose of the present invention is to propose a Coiflet wavelet relative spectral entropy line selection method and system for single-phase grounding faults, which comprehensively considers the transient and steady-state characteristics before and after the fault moment to improve the adaptability and accuracy of line selection. It is not only applicable to traditional typical scenarios, but also to extreme scenarios such as zero-sequence CT imbalance, reverse CT polarity, and load current imbalance.

[0019] To achieve the above object, the technical solution of the present invention is as follows:

[0020] On the one hand, the present invention proposes a Coiflet wavelet relative spectral entropy line selection method for single-phase grounding fault, which specifically includes the following steps:

[0021] S1, collect the zero-sequence current of each line connected to the bus;

[0022] S2. Perform Coiflet wavelet decomposition on the zero-sequence current of each line connected to the busbar to obtain the wavelet coefficients of each line in different frequency bands, and determine the time of fault occurrence based on the changing characteristics of the wavelet coefficients in different frequency bands;

[0023] S3. Taking the fault occurrence time as the dividing point, calculate the energy spectrum of different characteristic frequency bands of each line before and after the fault, and calculate the comprehensive relative spectral entropy of different characteristic frequency bands of each line before and after the fault;

[0024] S4. Different characteristic frequency bands are used as different analysis scales. In each analysis scale, the fault line is selected according to the comprehensive relative spectral entropy of each line to realize single-phase grounding fault line selection.

[0025] Preferably, the Coiflet 2 wavelet is used to perform wavelet decomposition on the zero-sequence current of each line connected to the bus.

[0026] Preferably, the step S2 specifically includes the following steps:

[0027] S21, performing Coiflet wavelet decomposition on the zero-sequence current of each line to obtain the wavelet coefficients of different characteristic frequency bands of each line;

[0028] S22. Calculate the change in the zero-sequence current amplitude of each line, and select the line with the most obvious change as the analysis modulus;

[0029] S23. Determine the maximum point of the wavelet coefficient of each characteristic frequency band of the analysis modulus, which is the initial moment of the fault in each frequency band space.

[0030] Preferably, the energy spectrum of the jth characteristic frequency band of the zero-sequence current signal of the ith line is calculated as follows:

[0031]

[0032] Where: is the coefficient of the (j,i) frequency band obtained by Coiflet wavelet decomposition, n=1,2,3,……N; N is the total number of wavelet coefficients of the jth characteristic frequency band; when (j,i) is the frequency band before the fault, e ji is the energy spectrum before the fault, denoted as e ji-pre ; (j,i) is the frequency band after the fault, e ji is the energy spectrum after the fault, denoted as e ji-post .

[0033] Preferably, the relative spectral entropy is calculated as follows:

[0034] Before the fault, the proportion of the energy spectrum of the i-th signal in the j-band scale to the total energy spectrum is:

[0035]

[0036] Where: M is the total number of lines carried by the busbar.

[0037] After the fault, the proportion of the energy spectrum of the i-th signal in the j-band scale to the total energy spectrum is:

[0038]

[0039] The relative spectral entropy H of the i-th signal before the fault relative to the signal after the fault on the j-band scale ji-pre for:

[0040] H ji-pre =|p ji-pre ln(p ji-pre / e ji-post )|

[0041] The relative spectral entropy H of the i-th signal after the fault relative to that before the fault on the j-band scale ji-post for:

[0042] H ji-post =|p ji-post ln(p ji-post / e ji-pre )|

[0043] Then the comprehensive relative spectral entropy H of the i-th signal before and after the fault on the j-band scale is ji for:

[0044] H ji =H ji-pre +H ji-post .

[0045] Preferably, the S4 is specifically:

[0046] According to the comprehensive relative spectral entropy H of the i-th signal before and after the fault on the j-band scale ji , select the line with the largest relative spectral entropy in each frequency band scale, which is the line selection result under the characteristic frequency band scale, that is,

[0047] L j =argmax(H ji )

[0048] Where: argmax() represents the position of the maximum value; L j is the line selection result under the j-band scale. The line with the most selected lines in each frequency band scale is the actual fault line.

[0049] Preferably, when the comprehensive relative entropy of each line meets the following conditions, it is determined that a single-phase grounding fault occurs on the bus:

[0050]

[0051] Where: K is the reliability coefficient, which is generally 0.69; H1, H2, and H3 are the comprehensive relative entropy H of the zero-sequence current signal of each line. Li The first three values; the comprehensive relative entropy H of the zero-sequence current signal of the i-th line Li The expression is:

[0052]

[0053] On the other hand, the present invention also proposes a Coiflet wavelet relative spectral entropy line selection system for single-phase grounding faults, including a processor, a memory, and a computer program stored on the memory. When the processor executes the computer program, it specifically executes any step in the above-mentioned Coiflet wavelet relative spectral entropy line selection method for single-phase grounding faults.

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] The present invention proposes a new line selection method for single-phase grounding faults based on Cofile wavelet relative spectral entropy. This method comprehensively considers the transient and steady-state characteristics before and after the fault, improving the adaptability and accuracy of line selection. It is not only applicable to traditional typical scenarios, but also to extreme scenarios such as zero-sequence CT imbalance, reverse CT polarity, and load current imbalance. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 This is the flow chart of the Coiflet wavelet relative spectral entropy line selection process of the present invention. DETAILED DESCRIPTION

[0057] The following is combined with Figure 1 , the technical solution of the present invention is described in detail.

[0058] Starting from the zero-sequence current energy spectrum before and after the fault, the present invention proposes a line selection method and system for single-phase grounding fault using Coiflet wavelet relative spectral entropy based on the excellent ability of Coiflet wavelet in processing local signal characteristics and the characteristic that relative entropy can better characterize the difference in signal probability distribution before and after the fault.

[0059] like Figure 1 The present invention proposes a Coiflet wavelet relative spectral entropy line selection method for single-phase grounding fault, which specifically includes the following steps:

[0060] S1, collect the zero-sequence current of each line connected to the bus;

[0061] S2. Perform Coiflet wavelet decomposition on the zero-sequence current of each line connected to the busbar to obtain the wavelet coefficients of each line in different frequency bands, and determine the time of fault occurrence based on the changing characteristics of the wavelet coefficients in different frequency bands;

[0062] S3. Taking the fault occurrence time as the dividing point, calculate the energy spectrum of different characteristic frequency bands of each line before and after the fault, and calculate the comprehensive relative spectral entropy of different characteristic frequency bands of each line before and after the fault;

[0063] S4. Use different characteristic frequency bands as different analysis scales, select the fault line according to the comprehensive relative spectral entropy of each line in each analysis scale, and integrate the results of each analysis scale to achieve single-phase grounding fault line selection.

[0064] In this embodiment, the Coiflet 2 wavelet is used to perform wavelet decomposition on the zero-sequence current of each line connected to the bus.

[0065] Wavelet transform is an effective method for extracting fault information features, and the selection of its wavelet basis is key to the algorithm's implementation. When processing fault voltage and current signals, it is necessary not only to detect the moment of signal disturbance but also to reconstruct and analyze the characteristic parts before and after the disturbance. Therefore, a wavelet basis with better symmetry is selected to reduce signal distortion during the decomposition and reconstruction process.

[0066] The Coiflet wavelet is a compactly supported wavelet function in which both the scaling function and the wavelet function have high vanishing momentum and its waveform is nearly symmetrical. It is often used for signal feature processing. The orthogonal Coiflet wavelet of order N satisfies the following conditions:

[0067]

[0068] Where: is the scaling function; ψ(t) is the wavelet function. ψ(t) is called the dual-scale equation of wavelet, and its expressions are:

[0069]

[0070] Where: p k ,q k Indicates the low-pass and high-pass filter coefficients. This application uses Coiflet 2 wavelet to extract fault features.

[0071] In this embodiment, S2 specifically includes the following steps:

[0072] S21, performing Coiflet wavelet decomposition on the zero-sequence current of each line to obtain the wavelet coefficients of different characteristic frequency bands of each line;

[0073] Taking the sampling frequency of 12500Hz as an example, a 6-layer wavelet decomposition is performed, and the multi-layer characteristic frequency bands are 0~98Hz, 98~195Hz, 195~390Hz, 390~781Hz, 781~1562Hz, 1562~3125Hz, and 3125~6250Hz.

[0074] S22. Calculate the change in the zero-sequence current amplitude of each line, and select the line with the most obvious change as the analysis modulus;

[0075] S23. Determine the maximum point of the wavelet coefficient of each characteristic frequency band of the analysis modulus, which is the initial moment of the fault in each frequency band space.

[0076] In this embodiment, the energy spectrum of the jth characteristic frequency band of the zero-sequence current signal of the i-th line is calculated as follows:

[0077]

[0078] Where: is the coefficient of the (j,i) frequency band obtained by Coiflet wavelet decomposition, n=1,2,3,……N; N is the total number of wavelet coefficients of the jth characteristic frequency band; when (j,i) is the frequency band before the fault, e ji is the energy spectrum before the fault, denoted as e ji-pre , otherwise it is the energy spectrum after the fault, recorded as e ji-post .

[0079] Relative spectral entropy is often used to measure the difference between two signals. A larger relative spectral entropy indicates a more pronounced difference between the waveform signals. When a single-phase ground fault occurs, the current flowing through the faulted line equals the sum of the ground capacitance of the non-faulted line and the current in the neutral arc suppression coil. In other words, the difference in current before and after the fault on the faulted line is typically more pronounced than that on the non-faulted line. This difference is measured using relative spectral entropy.

[0080] In this embodiment, the relative spectral entropy is calculated as follows:

[0081] Before the fault, the proportion of the energy spectrum of the i-th signal in the j-band scale to the total energy spectrum is:

[0082]

[0083] Where: M is the total number of lines carried by the busbar.

[0084] After the fault, the proportion of the energy spectrum of the i-th signal in the j-band scale to the total energy spectrum is:

[0085]

[0086] The relative spectral entropy H of the i-th signal before the fault relative to the signal after the fault on the j-band scale ji-pre for:

[0087] H ji-pre =|p ji-pre ln(p ji-pre / e ji-post )|

[0088] The relative spectral entropy H of the i-th signal after the fault relative to that before the fault on the j-band scale ji-post for:

[0089] H ji-post =|p ji-post ln(p ji-post / e ji-pre )|

[0090] Then the comprehensive relative spectral entropy H of the i-th signal before and after the fault on the j-band scale is ji for:

[0091] H ji =H ji-pre +H ji-post .

[0092] In this embodiment, the S4 is specifically:

[0093] According to the comprehensive relative spectral entropy H of the i-th signal before and after the fault on the j-band scale ji , select the line with the largest relative spectral entropy in each frequency band scale, which is the line selection result under the characteristic frequency band scale, that is,

[0094] L j =argmax(H ji )

[0095] Where: argmax() represents the position of the maximum value; L j is the line selection result under the j-band scale. The line with the most selected lines in each frequency band scale is the actual fault line.

[0096] When a single-phase grounding fault occurs on the busbar, the comprehensive relative entropy of the zero-sequence current of the lines is relatively close. That is to say, when a single-phase grounding fault occurs on the busbar, the comprehensive relative entropy of each line meets the following conditions:

[0097]

[0098] Where: K is the reliability coefficient, which is generally 0.69; H1, H2, and H3 are the comprehensive relative entropy H of the zero-sequence current signal of each line. Li The first three values; the comprehensive relative entropy H of the zero-sequence current signal of the i-th line Li The expression is:

[0099]

[0100] On the other hand, the present invention also proposes a Coiflet wavelet relative spectral entropy line selection system for single-phase grounding faults, including a processor, a memory, and a computer program stored on the memory. When the processor executes the computer program, it specifically executes any step in the above-mentioned Coiflet wavelet relative spectral entropy line selection method for single-phase grounding faults.

[0101] The feasibility and effectiveness of the proposed method were verified using 37 sets of fault data collected from actual substations. These data included 12 sets of typical scenarios (i.e., correct zero-sequence CT polarity, CT ratio, and secondary circuit, and CT imbalance meeting requirements), and 25 sets of extreme scenarios (i.e., reversed CT polarity, unbalanced CT configuration, and ineffective CT connection to the non-fault line).

[0102] The line selection results of the present invention, the fundamental wave amplitude and phase ratio method, the harmonic amplitude and phase ratio method, and the first half-wave method are shown in Table 1.

[0103] Table 1 Statistics of line selection results of various small current grounding line selection algorithms in different scenarios

[0104] method Fundamental wave amplitude and phase ratio method Harmonic amplitude and phase ratio method First Half-Wave Method This patented method Typical scenarios 4 10 7 12 Extreme scenarios 9 10 13 19

[0105] As can be seen from Table 1, compared with the currently commonly used transient steady-state line selection algorithm, the method proposed in this patent is not only applicable to typical scenarios, but also to extreme scenarios such as reverse polarity of zero-sequence CTs and unbalanced CT configuration, and has a higher line selection accuracy.

[0106] The above are preferred embodiments of the present invention. Any changes made according to the technical solution of the present invention, as long as the resulting functions and effects do not exceed the scope of the technical solution of the present invention, shall fall within the scope of protection of the present invention.

Claims

1. A Coiflet wavelet relative spectral entropy line selection method for single-phase grounding fault, characterized by: The specific steps include: S1, collect the zero-sequence current of each line connected to the bus; S2. Perform Coiflet wavelet decomposition on the zero-sequence current of each line connected to the busbar to obtain the wavelet coefficients of each line in different frequency bands, and determine the time of fault occurrence based on the changing characteristics of the wavelet coefficients in different frequency bands; S3. Taking the fault occurrence time as the dividing point, calculate the energy spectrum of different characteristic frequency bands of each line before and after the fault, and calculate the comprehensive relative spectral entropy of different characteristic frequency bands of each line before and after the fault; S4. Different characteristic frequency bands are used as different analysis scales. In each analysis scale, the fault line is selected according to the comprehensive relative spectral entropy of each line to realize single-phase grounding fault line selection.

2. The Coiflet wavelet relative spectral entropy line selection method for single-phase grounding fault according to claim 1 is characterized in that: The Coiflet 2 wavelet is used to perform wavelet decomposition on the zero-sequence current of each line connected to the bus.

3. The Coiflet wavelet relative spectral entropy line selection method for single-phase grounding fault according to claim 2 is characterized in that: The S2 specifically includes the following steps: S21, performing Coiflet wavelet decomposition on the zero-sequence current of each line to obtain the wavelet coefficients of different characteristic frequency bands of each line; S22. Calculate the change in the zero-sequence current amplitude of each line, and select the line with the most obvious change as the analysis modulus; S23. Determine the maximum point of the wavelet coefficient of each characteristic frequency band of the analysis modulus, which is the initial moment of the fault in each frequency band space.

4. The Coiflet wavelet relative spectral entropy line selection method for single-phase grounding fault according to claim 1 is characterized in that: The energy spectrum of the jth characteristic frequency band of the zero-sequence current signal of the i-th line is calculated as follows: Where: is the coefficient of the (j,i) frequency band obtained by Coiflet wavelet decomposition, n=1,2,3,……N; N is the total number of wavelet coefficients of the jth characteristic frequency band; (j,i) is the frequency band before the fault, e ji is the energy spectrum before the fault, denoted as e ji-pre ; (j,i) is the frequency band after the fault, e ji is the energy spectrum after the fault, denoted as e ji-post .

5. The Coiflet wavelet relative spectral entropy line selection method for single-phase grounding fault according to claim 4 is characterized in that: The relative spectral entropy is calculated as follows: Before the fault, the proportion of the energy spectrum of the i-th signal in the j-band scale to the total energy spectrum is: Where: M is the total number of lines carried by the bus; After the fault, the proportion of the energy spectrum of the i-th signal in the j-band scale to the total energy spectrum is: The relative spectral entropy H of the i-th signal before the fault relative to the signal after the fault on the j-band scale ji-pre for: A ji-pre =|p ji-pre ln(p ji-pre / have been ji-post )| The relative spectral entropy H of the i-th signal after the fault relative to that before the fault on the j-band scale ji-post for: A ji-post =|p ji-post ln(p ji-post / have been ji-pre )| Then the comprehensive relative spectral entropy H of the i-th signal before and after the fault on the j-band scale is ji for: H ji =H ji-pre +H ji-post 。 6. The Coiflet wavelet relative spectral entropy line selection method for single-phase grounding fault according to claim 5 is characterized in that: The S4 is specifically: According to the comprehensive relative spectral entropy H of the i-th signal before and after the fault on the j-band scale ji , select the line with the largest relative spectral entropy in each frequency band scale, which is the line selection result under the characteristic frequency band scale, that is, L j =argmax(H ji ) Where: argmax() represents the position of the maximum value; L j is the line selection result under the j-band scale. The line with the most selected lines in each frequency band scale is the actual fault line.

7. The Coiflet wavelet relative spectral entropy line selection method for single-phase grounding fault according to claim 5 is characterized in that: When the comprehensive relative entropy of each line meets the following conditions, it is determined that a single-phase grounding fault has occurred on the bus: Where: K is the reliability coefficient; H1, H2, H3 are the comprehensive relative entropy H of the zero-sequence current signal of each line respectively. Li The first three values; the comprehensive relative entropy H of the zero-sequence current signal of the i-th line Li The expression is:

8. The Coiflet wavelet relative spectral entropy line selection method for single-phase grounding fault according to claim 7 is characterized in that: The reliability coefficient K is 0.

69.

9. A Coiflet wavelet relative spectral entropy line selection system for single-phase grounding fault, characterized by: The method comprises a processor, a memory and a computer program stored in the memory. When the processor executes the computer program, the method specifically performs the steps in the Coiflet wavelet relative spectral entropy line selection method for single-phase grounding fault according to any one of claims 1 to 8.

Images