A new high-precision zero-sequence current adaptive sampling method

Abstract

The invention discloses a novel high-precision zero-sequence current adaptive sampling method, which relates to the technical field of small current fault line selection and solves the technical problems of poor zero-sequence current sampling accuracy, inaccurate positioning of small current fault line selection, and influence on overall maintenance. The invention improves the accuracy of overall positioning by filtering the collected zero-sequence current signal, and facilitates subsequent abnormal analysis based on the zero-sequence current. Secondly, when analyzing based on the zero-sequence current, data of multiple groups of sampling points are collected, data features are extracted using an artificial intelligence algorithm, and the distribution of the sampling points is analyzed according to the data features. According to different distribution conditions, data analysis corresponding to the sampling points is combined. When analyzing, the voltage, current, temperature and other factors are combined and analyzed to achieve the determination of specific sampling point abnormalities according to the accuracy of current positioning, and further improve the accuracy of overall zero-sequence current positioning.

Description

Technical Field

[0001] The invention relates to the technical field of small current fault line selection, and in particular to a novel high-precision zero-sequence current adaptive sampling method. Background Art

[0002] A low grounding current power grid refers to a power system whose neutral point is not grounded or is grounded through an arc suppression coil. When a single-phase grounding fault occurs in a power grid with this grounding method, no short circuit is formed and the fault current is very small. The regulations stipulate that the system can continue to operate for 1 to 2 hours after a single-phase grounding fault occurs. Although a single-phase grounding fault does not destroy the symmetrical operation of the system, it will cause the non-fault phase voltage to rise to the line voltage. Long-term operation is likely to cause insulation damage to the non-fault phase, forming a phase-to-phase short circuit. Arc grounding may also cause higher overvoltage in the power grid. Therefore, when a single-phase grounding fault occurs, the fault line should be identified and the fault should be eliminated as soon as possible. The traditional line selection method is to manually switch off each line one by one. When the faulty line is switched off, the fault signal disappears.

[0003] With the development of primary and secondary integration technology of distribution network, current transformers based on electronic principles have become key components of distribution automation systems. The corresponding feeder terminals (FTUs) are also required to realize accurate sampling of electronic zero-sequence current and phase current. According to Chinese patent application number CN202310167055.0, a zero-sequence current sampling system and sampling method for electronic transformers are disclosed. The system includes: a micro transformer, a signal conditioning module, an analog-to-digital conversion module, a signal processing module and a human-computer interaction module; the method includes the following steps: according to the system hardware parameters and characteristics, a current demarcation point is determined, and a current greater than this current is specified as a larger current, and a current less than this current is specified as a smaller current; the CPU is specified to use the data of channel 1 to determine whether the current type input by the system is a larger current or a smaller current, and the CPU sends the data of channel 1 when the system inputs a larger current, and sends the data of channel 2 when the system inputs a smaller current; the two channels of the ADC are calibrated separately; within the range specified by the hardware, any current of any size is input to the system for verification.

[0004] When a single-phase grounding fault occurs in a small current grounding system, the fault current is very small, usually less than 10A, and the sampling accuracy of the zero-sequence current transformer is not high. If a zero-sequence current transformer is not installed on the line, a power outage is required to install it. Or if a zero-sequence current transformer is not installed and the secondary current of the ABC three-phase current transformer is used to synthesize and calculate the zero-sequence current, the error will be greater. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides a novel high-precision zero-sequence current adaptive sampling method, which solves the problems of poor zero-sequence current sampling accuracy, inaccurate line selection and positioning of small current faults, and impact on overall maintenance.

[0006] To achieve the above objectives, the present invention is implemented by the following technical solutions: a new high-precision zero-sequence current adaptive sampling method, specifically comprising the following steps:

[0007] Step 1: Acquire the original signal in the three-phase system, perform wavelet transformation on the original signal to extract the zero-sequence current signal, and filter the zero-sequence current signal to obtain a filtered signal;

[0008] Step 2: Obtain the filtering signals corresponding to the sampling points in the three-phase system, and classify the sampling points into normal sampling points and abnormal sampling points according to the filtering signals;

[0009] Step 3: Analyze the abnormal sampling points obtained, extract the data features of the abnormal sampling points, and determine the distribution of the data features to generate stable distribution information and unstable distribution information;

[0010] Step 4: Analyze the obtained stable distribution information, analyze the sampling point parameters corresponding to the abnormal sampling point, analyze the voltage and current of the sampling point comprehensively to determine the specific abnormal point, and generate specific abnormal information;

[0011] Step 5: Analyze the obtained unstable distribution information, collect and analyze the temperature parameters corresponding to the abnormal sampling points, and analyze the changes in the zero-sequence current value to generate corresponding specific abnormal information.

[0012] As a further solution of the present invention: the specific method of obtaining the filtered signal in step 1 is:

[0013] The original signal data containing zero-sequence current is obtained, and the appropriate wavelet basis function is selected to perform multi-layer wavelet decomposition on the collected signal to obtain wavelet coefficients of different frequency bands. The decomposed wavelet coefficients are processed to retain the wavelet coefficients of zero-sequence current characteristics, and the signal is reconstructed using the processed wavelet coefficients to obtain a denoised filtered signal.

[0014] As a further solution of the present invention: the specific method of classifying according to the filtered signal in step 2 is:

[0015] All sampling points are obtained and labeled as a, where a=1, 2, ..., b, where b represents the number of sampling points, and the sampling point a is normalized. At the same time, the zero-sequence current value corresponding to the sampling point a is obtained and recorded as Ia. Then, the zero-sequence current mean of all sampling points a is calculated and recorded as Ip. The sampling points whose zero-sequence current Ia is greater than the zero-sequence current mean Ip are classified as abnormal sampling points, otherwise they are classified as normal sampling points.

[0016] As a further solution of the present invention: the specific method of analyzing the abnormal sampling points in step 3 is:

[0017] All abnormal sampling points are obtained and recorded as o, where o=1, 2, ..., p, and o represents the number of abnormal sampling points. The data features of the abnormal sampling points o are extracted through the artificial intelligence algorithm. Then, the distribution of the abnormal sampling points is determined based on the data features. The specific distribution is determined as follows:

[0018] The total number of abnormal sampling points is obtained, and the number of abnormal sampling points of concentrated clustering is obtained, and then the proportion of the concentrated clustered abnormal sampling points is calculated, and the calculated proportion is compared with the preset value. When the proportion is greater than the preset value, stable distribution information is generated. Conversely, when the proportion is less than the preset value, unstable distribution information is generated.

[0019] As a further solution of the present invention: the specific method of analyzing the stable distribution information in step 4 is:

[0020] The sampling point parameters corresponding to the abnormal sampling point o are obtained, where the sampling point parameters include the sampling point voltage Uo and the sampling point current Yo. At the same time, the sampling point voltage Uo is analyzed. The specific analysis method is as follows:

[0021] The voltage Uo of the abnormal sampling point is sorted from small to large to obtain the voltage sorting information, and the maximum value Umax and the minimum value Umin of the sampling point voltage are determined, and it is judged whether the voltage sorting information is the same as the label information. If they are the same, the maximum value Umax of the abnormal sampling point voltage is taken as the abnormal point to be analyzed. If they are not the same, the maximum value Umax of the abnormal sampling point voltage is selected as the abnormal point to be analyzed;

[0022] The specific method of analyzing the sampling point current Yo is as follows: obtaining the zero-sequence current direction corresponding to the abnormal sampling point o, and obtaining the current direction of the corresponding line as the standard direction, and comparing the zero-sequence current direction of the abnormal sampling point o with the standard direction, and screening out the abnormal sampling points whose zero-sequence current direction is different from the standard direction as pre-selected abnormal points.

[0023] As a further solution of the present invention: the specific method of analyzing and determining the specific abnormal point by comprehensively analyzing the voltage and current of the sampling point in step 4 is:

[0024] Then, the abnormal point to be analyzed is matched with the pre-selected abnormal point, and the abnormal sampling points with overlap are screened out and recorded as secondary pre-selected points. At the same time, the number of secondary pre-selected points is obtained. If the number of secondary pre-selected points is equal to 1, the secondary pre-selected point is used as the specific abnormal point, and specific abnormal information is generated. If the number of secondary pre-selected points is greater than 1, the secondary pre-selected point with the largest corresponding zero-sequence current value is selected as the specific abnormal point, and specific abnormal information is generated.

[0025] As a further solution of the present invention: the specific method of analyzing the unstable distribution information in step 5 is:

[0026] The real-time temperature corresponding to all abnormal sampling points o is obtained and recorded as To, and the temperature change within the time period t1 is obtained. At the same time, the change of the zero-sequence current Io of the abnormal sampling point o is obtained. If the temperature change is the same as the zero-sequence current change, an abnormal analysis signal is generated. Otherwise, if the temperature change is different from the zero-sequence current change, a normal analysis signal is generated.

[0027] As a further solution of the present invention: the specific method of analyzing the normal analysis signal and the abnormal analysis signal in step 5 is:

[0028] If a normal analysis signal is generated, the corresponding abnormal sampling point o is taken as a specific abnormal point and specific abnormal information is generated. If an abnormal analysis signal is generated, the load of the abnormal sampling point o is analyzed. If there is no corresponding load increase at the abnormal sampling point o, the corresponding abnormal sampling point o is taken as a specific abnormal point and specific abnormal information is generated. If there is a corresponding load increase at the abnormal sampling point o, a normal signal is generated and no processing is performed.

[0029] The present invention provides a novel high-precision zero-sequence current adaptive sampling method. Compared with the prior art, it has the following beneficial effects:

[0030] The present invention improves the accuracy of overall positioning by filtering the collected zero-sequence current signal, and facilitates subsequent abnormal analysis based on the zero-sequence current. Secondly, when analyzing based on the zero-sequence current, by collecting data from multiple groups of sampling points, data features are extracted using an artificial intelligence algorithm, and the distribution of the sampling points is analyzed based on the data features. According to different distribution conditions, the data corresponding to the sampling points are analyzed. During the analysis, the abnormalities of the sampling points are located by combining and analyzing factors such as voltage, current and temperature, thereby further improving the accuracy of overall zero-sequence current positioning. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 Figure 2 is a diagram of the method of the present invention. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] See also Figure 1 The present application provides a novel high-precision zero-sequence current adaptive sampling method, which specifically includes the following steps:

[0034] Step 1: Acquire the original signal in the three-phase system, perform wavelet transformation on the original signal to extract the zero-sequence current signal, and filter the zero-sequence current signal to obtain the filtered signal. The specific method of obtaining the filtered signal is:

[0035] First, the current signal of the power system is collected to obtain the original signal data containing the zero-sequence current, and the appropriate wavelet basis function is selected. The collected signal is subjected to multi-layer wavelet decomposition to obtain wavelet coefficients of different frequency bands. The decomposed wavelet coefficients are processed, such as threshold filtering, to remove noise and interference components, and the wavelet coefficients of the zero-sequence current characteristics are retained. The signal is reconstructed using the processed wavelet coefficients to obtain the denoised zero-sequence current signal, thereby achieving effective extraction from the noise and obtaining the filtered signal. Wavelet transform is a multi-scale time-frequency analysis method. Through variable-size window technology, the signal is locally analyzed to extract specific components in the signal, which will not be elaborated on here.

[0036] Assume that there is an original signal in a three-phase system, which contains multiple frequency components. The second main goal is to extract the zero-sequence current signal and filter it. First, the original signal is processed using wavelet transform, and the zero-sequence current signal is successfully extracted. Then a low-pass filter is designed to filter out the high-frequency noise and unnecessary frequency components in the zero-sequence current signal, and a clear filtered signal is obtained.

[0037] Step 2: Obtain the filtering signal corresponding to the sampling point in the three-phase system, and classify the sampling points into normal sampling points and abnormal sampling points according to the filtering signal, and generate sampling point classification information at the same time. The specific classification method is as follows:

[0038] All sampling points are obtained and labeled as a, and a=1, 2, ..., b, where b represents the number of sampling points, and the sampling point a is normalized, and the normalization usually adopts the Z-score standardization method. At the same time, the zero-sequence current value corresponding to the sampling point a is obtained and recorded as Ia. Then, the zero-sequence current mean of all sampling points a is calculated and recorded as Ip, and the sampling points whose zero-sequence current Ia is greater than the zero-sequence current mean Ip are classified as abnormal sampling points, otherwise they are classified as normal sampling points, and this includes sampling points that are less than or equal to the zero-sequence current mean Ip.

[0039] Step 3: Analyze the abnormal sampling points obtained, extract the data features of the abnormal sampling points, and determine the distribution of the data features to generate stable distribution information and unstable distribution information. The specific determination method is as follows:

[0040] All abnormal sampling points are obtained and recorded as o, where o=1, 2, ..., p, and o represents the number of abnormal sampling points, and the data features of the abnormal sampling point o are extracted through the artificial intelligence algorithm, and then the distribution of the abnormal sampling point is determined based on the data features. Specifically, the data features of the abnormal sampling point o are extracted using artificial intelligence algorithms, such as machine learning or deep learning technology. These features may include the amplitude, phase, and frequency of the zero-sequence current. Through algorithm processing, the key features in the data can be automatically identified. For example, the most critical features can be extracted through dimensionality reduction technologies such as principal component analysis (PCA) or autoencoders to reduce the impact of noise and redundant information. The potential connections between abnormal sampling points can be discovered using methods such as cluster analysis and association rule mining, thereby revealing possible fault modes or system defects, and the specific distribution is determined as follows:

[0041] The total number of abnormal sampling points is obtained, and the number of concentrated clustered abnormal sampling points is obtained, and then the proportion of concentrated clustered abnormal sampling points is calculated, and the calculated proportion is compared with the preset value, and the specific value of the preset value is set by the operator according to experience. When the proportion is greater than the preset value, it indicates that the abnormal sampling points are clustered and stable distribution information is generated. On the contrary, when the proportion is less than the preset value, it indicates that the abnormal sampling points are non-clustered and unstable distribution information is generated. For example, if it is found that the abnormal sampling points are mainly concentrated in a specific time period or a specific area.

[0042] Step 4: Analyze the obtained stable distribution information, analyze the sampling point parameters corresponding to the abnormal sampling point, analyze the voltage and current of the sampling point to determine the specific abnormal point, and generate specific abnormal information. The specific abnormal information is generated in the following way:

[0043] Get the sampling point parameters corresponding to the abnormal sampling point o, where the sampling point parameters include the sampling point voltage Uo and the sampling point current Yo, and the sampling point current here is the current in the three-phase line, and analyze the sampling point voltage Uo at the same time. The specific analysis method is:

[0044] The voltage Uo of the abnormal sampling point is sorted from small to large to obtain the voltage sorting information, and the maximum value Umax and the minimum value Umin of the sampling point voltage are determined, and it is judged whether the voltage sorting information is the same as the label information. If they are the same, the maximum value Umax of the abnormal sampling point voltage is used as the abnormal point to be analyzed. If they are not the same, the maximum value Umax of the abnormal sampling point voltage is selected as the abnormal point to be analyzed. Specifically, at the grounding fault point, the zero-sequence voltage is the largest and gradually decreases toward both ends of the line, so the point where the voltage maximum value appears is selected; specifically, in the actual application process, the sampling point voltage is the same value, and in the three-term system, it corresponds to only one voltage value.

[0045] The specific method of analyzing the sampling point current Yo is as follows: obtaining the zero-sequence current direction corresponding to the abnormal sampling point o, and obtaining the current direction of the corresponding line as the standard direction, and comparing the zero-sequence current direction of the abnormal sampling point o with the standard direction, and screening out the abnormal sampling points whose zero-sequence current direction is different from the standard direction as pre-selected abnormal points, and here all abnormal sampling points are analyzed, including the abnormal points to be analyzed obtained by the above processing;

[0046] Then, the abnormal point to be analyzed is matched with the pre-selected abnormal point, and the abnormal sampling points with overlap are screened out and recorded as secondary pre-selected points. The secondary pre-selected points here represent points where both current analysis and voltage analysis are abnormal. At the same time, the number of secondary pre-selected points is obtained. If the number of secondary pre-selected points is equal to 1, the secondary pre-selected point is used as a specific abnormal point, and specific abnormal information is generated. If the number of secondary pre-selected points is greater than 1, the secondary pre-selected point with the largest corresponding zero-sequence current value is selected as a specific abnormal point, and specific abnormal information is generated.

[0047] For example, current analysis may identify sampling points A, B, and C as abnormal, while voltage analysis pre-selects abnormal points B, D, and E. After matching, it can be found that point B appears in both analysis results. For example, if only sampling point B exists in the abnormal points of both analyses after matching, then point B will be determined as a specific abnormal point, and its abnormal information such as abnormal current value and voltage value will be recorded. For example, if points B and E are found to exist in the abnormal points of both analyses after matching, and the zero-sequence current value of point B is 1.50A and the zero-sequence current value of point E is 1.80A, then point E is selected as the specific abnormal point and its abnormal information is recorded.

[0048] Step 5: Analyze the obtained unstable distribution information, collect and analyze the temperature parameters corresponding to the abnormal sampling points, and analyze the changes in the zero-sequence current value to generate the corresponding specific abnormal information, and the specific analysis method is:

[0049] The real-time temperature corresponding to all abnormal sampling points o is obtained and recorded as To, and the temperature change within the time period t1 is obtained, and the specific value of the time period t1 here is set by the operator. At the same time, the change of the zero-sequence current Io of the abnormal sampling point o is obtained. If the temperature change is the same as the zero-sequence current change, an abnormal analysis signal is generated. Conversely, if the temperature change is different from the zero-sequence current change, a normal analysis signal is generated. Specifically, if the changes of the two are the same, it means that the temperature increases, and the corresponding zero-sequence current also increases. In this case, it means that the overall situation is abnormal. In real life, if the temperature increases, the corresponding zero-sequence current will decrease. Therefore, if they are the same, it means abnormality, and if they are different, it means normal.

[0050] If a normal analysis signal is generated, the corresponding abnormal sampling point o is taken as a specific abnormal point and specific abnormal information is generated. If an abnormal analysis signal is generated, the load of the abnormal sampling point o is analyzed. If there is no corresponding load increase at the abnormal sampling point o, the corresponding abnormal sampling point o is taken as a specific abnormal point and specific abnormal information is generated. If there is a corresponding load increase at the abnormal sampling point o, a normal signal is generated and no processing is performed. For example, if the temperature and zero-sequence current of a certain abnormal sampling point both rise within the time period t1, and the load monitoring shows an increase, the system will consider this to be a normal phenomenon and will not perform abnormal processing. For example, if the temperature and zero-sequence current of a certain abnormal sampling point both rise within the time period t1, but the load monitoring shows no increase, the system will record the specific abnormal information of the point, such as "the temperature and zero-sequence current of the sampling point both rise within the time period t1, but the load does not increase, and there may be current-related abnormalities.

[0051] In addition to directly measuring the corresponding temperature through the temperature sensor, the cable temperature can also be calculated through the movement of the load current. Combined with actual analysis, in the actual operation of the three-phase system, the current will generate heat due to the movement of the load current during transmission, which further causes the cable surface temperature to change, such as:

[0052] The resistance R of the cable can be calculated by the formula: R = ρ (L / A), where ρ is the resistivity, L is the cable length, and A is the cross-sectional area. For example, a 10-meter-long cable with a cross-sectional area of ​​4mm 2 The resistance of the copper core cable is 0.017×(10 / 4)=0.0425Ω. The heat generated by the cable Q can be calculated by Joule's law: Q=i 2 Rt, where i is the current through the cable, R is the cable resistance, and t is the time. For example, if the current is 50A and the resistance is 0.0425Ω, the heat generated in one second is 50 2 ×0.0425=1062.5 joules. Ideally, the cable surface temperature can be estimated using a simplified formula: Ts=T0+Q / (mc), where Ts is the surface temperature, T0 is the initial temperature, Q is the calorific value, m is the cable mass, and c is the specific heat capacity of the cable material. However, in actual use, the calculation formula for the cable surface temperature must be substituted into the environmental impact, which can be specifically expressed as: Ts=(T0+Q / (mc))×ai, where ai is the environmental temperature impact factor;

[0053] The surface temperature of the cable can be calculated based on the above formula.

[0054] Some of the data in the above formulas are dimensionless and numerically calculated. Meanwhile, the contents not described in detail in this specification belong to the prior art known to those skilled in the art.

[0055] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A novel high-precision zero-sequence current adaptive sampling method, characterized in that: The specific steps include: Step 1: Acquire the original signal in the three-phase system, perform wavelet transformation on the original signal to extract the zero-sequence current signal, and filter the zero-sequence current signal to obtain a filtered signal; Step 2: Obtain the filtering signals corresponding to the sampling points in the three-phase system, and classify the sampling points into normal sampling points and abnormal sampling points according to the filtering signals; Step 3: Analyze the abnormal sampling points obtained, extract the data features of the abnormal sampling points, and determine the distribution of the data features to generate stable distribution information and unstable distribution information. The specific analysis method is as follows: All abnormal sampling points are obtained and recorded as o, where o=1, 2, ..., p, and o represents the number of abnormal sampling points. The data features of the abnormal sampling points o are extracted through the artificial intelligence algorithm. Then, the distribution of the abnormal sampling points is determined based on the data features. The specific distribution is determined as follows: Obtain the total number of abnormal sampling points and the number of concentrated clustered abnormal sampling points, then calculate the proportion of concentrated clustered abnormal sampling points, and compare the calculated proportion with the preset value. When the proportion is greater than the preset value, stable distribution information is generated. Conversely, when the proportion is less than the preset value, unstable distribution information is generated. Step 4: Analyze the obtained stable distribution information, analyze the sampling point parameters corresponding to the abnormal sampling point, analyze the voltage and current of the sampling point to determine the specific abnormal point, and generate specific abnormal information. The specific analysis method is as follows: The sampling point parameters corresponding to the abnormal sampling point o are obtained, where the sampling point parameters include the sampling point voltage Uo and the sampling point current Yo. At the same time, the sampling point voltage Uo is analyzed. The specific analysis method is as follows: The voltage Uo of the abnormal sampling point is sorted from small to large to obtain the voltage sorting information, and the maximum value Umax and the minimum value Umin of the sampling point voltage are determined, and it is judged whether the voltage sorting information is the same as the label information. If they are the same, the maximum value Umax of the abnormal sampling point voltage is taken as the abnormal point to be analyzed. If they are not the same, the maximum value Umax of the abnormal sampling point voltage is selected as the abnormal point to be analyzed; The specific method of analyzing the sampling point current Yo is as follows: obtaining the zero-sequence current direction corresponding to the abnormal sampling point o, and obtaining the current direction of the corresponding line as the standard direction, and comparing the zero-sequence current direction of the abnormal sampling point o with the standard direction, and screening out the abnormal sampling points whose zero-sequence current direction is different from the standard direction as pre-selected abnormal points; Step 5: Analyze the obtained unstable distribution information, collect and analyze the temperature parameters corresponding to the abnormal sampling points, and analyze the changes in the zero-sequence current value to generate the corresponding specific abnormal information. The specific processing method is as follows: Then, the abnormal point to be analyzed is matched with the pre-selected abnormal point, and the abnormal sampling points with overlap are screened out and recorded as secondary pre-selected points. At the same time, the number of secondary pre-selected points is obtained. If the number of secondary pre-selected points is equal to 1, the secondary pre-selected point is taken as a specific abnormal point, and specific abnormal information is generated. If the number of secondary pre-selected points is greater than 1, the secondary pre-selected point with the largest corresponding zero-sequence current value is selected as a specific abnormal point, and specific abnormal information is generated; The real-time temperature corresponding to all abnormal sampling points o is obtained and recorded as To, and the temperature change within the time period t1 is obtained. At the same time, the change of the zero-sequence current Io of the abnormal sampling point o is obtained. If the temperature change is the same as the zero-sequence current change, an abnormal analysis signal is generated. Otherwise, if the temperature change is different from the zero-sequence current change, a normal analysis signal is generated.

2. A novel high-precision zero-sequence current adaptive sampling method according to claim 1, characterized in that: The specific method of obtaining the filtered signal in step 1 is: The original signal data containing zero-sequence current is obtained, and the appropriate wavelet basis function is selected to perform multi-layer wavelet decomposition on the collected signal to obtain wavelet coefficients of different frequency bands. The decomposed wavelet coefficients are processed to retain the wavelet coefficients of zero-sequence current characteristics, and the signal is reconstructed using the processed wavelet coefficients to obtain a denoised filtered signal.

3. A novel high-precision zero-sequence current adaptive sampling method according to claim 1, characterized in that: The specific method of classifying according to the filtered signal in step 2 is: All sampling points are obtained and labeled as a, where a=1, 2, ..., b, where b represents the number of sampling points, and the sampling point a is normalized. At the same time, the zero-sequence current value corresponding to the sampling point a is obtained and recorded as Ia. Then, the zero-sequence current mean of all sampling points a is calculated and recorded as Ip. The sampling points whose zero-sequence current Ia is greater than the zero-sequence current mean Ip are classified as abnormal sampling points, otherwise they are classified as normal sampling points.

4. A novel high-precision zero-sequence current adaptive sampling method according to claim 1, characterized in that: The specific method of analyzing the normal analysis signal and the abnormal analysis signal in step 5 is: If a normal analysis signal is generated, the corresponding abnormal sampling point o is taken as a specific abnormal point and specific abnormal information is generated. If an abnormal analysis signal is generated, the load of the abnormal sampling point o is analyzed. If there is no corresponding load increase at the abnormal sampling point o, the corresponding abnormal sampling point o is taken as a specific abnormal point and specific abnormal information is generated. If there is a corresponding load increase at the abnormal sampling point o, a normal signal is generated and no processing is performed.

Images