A Method and System for Grounding Current Analysis of Iron Core Clamping Based on Wavelet Transform Algorithm

Abstract

This invention discloses a method and system for analyzing grounding current in iron core clamps based on wavelet transform algorithm. The method includes: acquiring grounding current signals; performing wavelet packet decomposition of the signals using a complete binary tree structure, employing Shannon entropy as the cost function, and adaptively selecting the optimal wavelet packet basis using a bottom-up greedy strategy; extracting high-frequency detail coefficients of each decomposition layer and calculating the frequency band energy density gradient; establishing an adaptive threshold based on the 95th quantile of the background interval, identifying effective abrupt events by combining duration and energy constraints, and merging adjacent events with a time interval of less than 5ms. The method achieves accurate matching of the signal's time-frequency structure through adaptive decomposition of the optimal wavelet packet basis, improves the sensitivity of weak abrupt event detection through the frequency band energy density gradient, and effectively suppresses false alarms due to noise through the adaptive threshold and energy screening mechanism. It can accurately detect grounding current abrupt changes caused by faults such as multi-point grounding and partial discharge.

Description

Technical Field

[0001] This invention belongs to the field of power automation technology, and in particular relates to a method and system for analyzing grounding current of iron core clamps based on wavelet transform algorithm. Background Technology

[0002] Converter transformers are core equipment in high-voltage direct current (HVDC) transmission projects, and their operating conditions differ fundamentally from those of conventional AC transformers. Because the converter valve windings are connected to the converter valve bridge arm, the voltage waveform experienced by the converter transformer contains not only AC components but also a large number of harmonic voltages, with harmonic orders reaching up to 60th order and frequencies exceeding 3kHz. This unique operating condition results in the converter transformer core clamp grounding current exhibiting complex harmonic components, significant DC interference, and the potential inclusion of transient signals.

[0003] Ultra-high voltage direct current (UHVDC) converter stations require regular monitoring of the grounding current of the converter transformer core clamps. When the core or clamps are grounded at multiple points, a loop forms between these grounding points. Due to the low internal resistance of the loop, a large circulating current is generated, which can easily lead to localized overheating and, in severe cases, insulation breakdown. Therefore, achieving high-precision monitoring of the grounding current of the core clamps and accurate identification of sudden events is crucial for ensuring the safe and stable operation of the UHVDC transmission system.

[0004] Currently, existing technologies related to grounding current detection of converter transformer core clamps have the following main shortcomings:

[0005] First, it has poor adaptability to non-stationary signals. Existing technologies mostly employ windowed interpolation FFT algorithms for harmonic analysis, such as using a four-term Blackman-Harris window or Nuttall window to suppress spectral leakage. However, these methods are primarily designed for steady-state harmonics. When non-stationary abrupt changes exist in the signal (such as partial discharge or intermittent grounding), FFT cannot pinpoint the exact time of the abrupt change, and the spectral analysis results are affected by these changes, leading to errors. Actual measurement data shows that the fundamental frequency content in the converter transformer grounding current is only about 61.86%, and a large number of harmonic and abrupt change components cannot be accurately analyzed using traditional methods.

[0006] Secondly, the accuracy of the test is severely affected by harmonic interference. Commercially available clamp meters are designed for AC transformers with relatively few harmonic components. When measuring the grounding current of a converter transformer, the secondary side causes distortion, resulting in significant deviations from the true values ​​for both total current and power frequency current tests. Experiments show that under complex backgrounds with harmonics up to the 60th order and frequencies exceeding 3kHz, the measurement error of traditional instruments can reach 10%-20%.

[0007] Third, there is a lack of targeted indicators for extracting mutation features. Existing wavelet transform applications mostly use the modulus maxima method to detect mutations, which is not sensitive enough to weak mutation signals and is easily affected by noise interference, resulting in false detections. Furthermore, no energy density indicator with clear physical meaning has been established, making it impossible to effectively distinguish mutation features in different frequency bands.

[0008] Fourth, the diagnostic dimensions are limited. Existing technologies mostly use FFT or wavelet transform alone, lacking a multi-algorithm collaborative diagnostic mechanism, and cannot comprehensively utilize steady-state harmonic features and transient change features to identify fault types.

[0009] Therefore, there is an urgent need for a wavelet transform-based method and system for analyzing the grounding current of iron core clamps that can adapt to the complex operating conditions of converter transformers, accurately identify non-stationary abrupt signals, and have anti-interference capabilities, in order to overcome the shortcomings of existing technologies. Summary of the Invention

[0010] This invention provides a method and system for analyzing grounding current of iron core clamps based on wavelet transform algorithm, which solves the technical problems of poor adaptability to non-stationary abrupt change signals, low measurement accuracy under complex harmonic backgrounds, and lack of targeted abrupt change feature extraction indicators in existing technologies.

[0011] In a first aspect, the present invention provides a method for analyzing the grounding current of iron core clamps based on a wavelet transform algorithm, comprising:

[0012] The grounding current signal of the core clamp of the UHV converter transformer was collected to obtain the original current sequence;

[0013] The original current sequence is subjected to wavelet packet decomposition with a complete binary tree structure. During the decomposition process, each layer simultaneously performs recursive decomposition on both the low-frequency approximation component and the high-frequency detail component to obtain the wavelet packet coefficients of each decomposition node.

[0014] Using Shannon entropy as a sparsity measure, the cost value corresponding to the wavelet packet coefficients of each decomposition node is calculated. A bottom-up greedy search strategy is adopted to compare the sum of the cost values ​​of the parent node and the child node. The optimal basis that makes the signal representation most compact is adaptively selected from all orthogonal wavelet packet bases, resulting in an adaptive decomposition structure that matches the time-frequency characteristics of the grounding current.

[0015] Based on the adaptive decomposition structure, high-frequency detail coefficients of each decomposition layer are extracted, and the frequency band energy density gradient of each layer is calculated.

[0016] The fault-free period in the ground current signal is selected as the background interval. The percentile of the frequency band energy density gradient in each background interval is calculated. The weighting coefficient is set according to the number of decomposition layers to obtain the adaptive detection threshold of each layer.

[0017] The energy density gradient sequence of each frequency band is scanned to detect intervals that continuously exceed the adaptive detection threshold, and valid mutation events are identified based on the interval duration and energy magnitude.

[0018] Merge adjacent valid mutation events within the same layer whose time interval is less than a preset value, and output the occurrence time, duration, dominant decomposition layer, and event energy of each mutation event as the detection result of the mutation characteristics of the grounding current.

[0019] Secondly, the present invention provides a system for analyzing the grounding current of iron core clamps based on a wavelet transform algorithm, comprising:

[0020] The acquisition module is configured to collect the grounding current signal of the core clamp of the UHV converter transformer to obtain the original current sequence;

[0021] The decomposition module is configured to perform wavelet packet decomposition of the original current sequence in a complete binary tree structure. During the decomposition process, each layer simultaneously performs recursive decomposition of the low-frequency approximation component and the high-frequency detail component to obtain the wavelet packet coefficients of each decomposition node.

[0022] The first calculation module is configured to use Shannon entropy as a sparsity measure to calculate the cost value corresponding to the wavelet packet coefficients of each decomposition node. It adopts a bottom-up greedy search strategy, compares the sum of the cost value of the parent node and the cost value of the child node, and adaptively selects the optimal basis that makes the signal representation most compact among all orthogonal wavelet packet bases to obtain an adaptive decomposition structure that matches the time-frequency characteristics of the grounding current.

[0023] The second calculation module is configured to extract the high-frequency detail coefficients of each decomposition layer according to the adaptive decomposition structure and calculate the frequency band energy density gradient of each layer.

[0024] The third calculation module is configured to select the fault-free period in the ground current signal as the background interval, calculate the percentile of the frequency band energy density gradient in each background interval, and set the weight coefficient according to the number of decomposition layers to obtain the adaptive detection threshold of each layer.

[0025] The identification module is configured to scan the energy density gradient sequence of each frequency band, detect intervals that continuously exceed the adaptive detection threshold, and identify valid mutation events based on the interval duration and energy magnitude.

[0026] The output module is configured to merge adjacent valid mutation events with a time interval of less than a preset value within the same layer, and output the occurrence time, duration, dominant decomposition layer and event energy of each mutation event as the mutation characteristic detection result of the ground current.

[0027] Thirdly, an electronic device is provided, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the steps of the wavelet transform algorithm-based grounding current analysis method for iron core clamps according to any embodiment of the present invention.

[0028] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor performs the steps of the method for analyzing the grounding current of iron core clamps based on the wavelet transform algorithm according to any embodiment of the present invention.

[0029] This application presents a method and system for analyzing grounding current in iron core clamps based on wavelet transform algorithm. The method involves acquiring grounding current signals; performing wavelet packet decomposition on the signals using a complete binary tree structure; adaptively selecting the optimal wavelet packet basis using Shannon entropy as the cost function and a bottom-up greedy strategy; extracting high-frequency detail coefficients from each decomposition layer and calculating the frequency band energy density gradient; establishing an adaptive threshold based on the 95th quantile of the background interval; identifying effective abrupt events by combining duration and energy constraints; and merging adjacent events with a time interval of less than 5ms. The method achieves accurate matching of the signal's time-frequency structure through adaptive decomposition of the optimal wavelet packet basis, improves the sensitivity of weak abrupt event detection through the frequency band energy density gradient, and effectively suppresses false alarms due to noise through the adaptive threshold and energy screening mechanism. This method can accurately detect grounding current abrupt changes caused by faults such as multi-point grounding and partial discharge. Attached Figure Description

[0030] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 A flowchart illustrating a method for analyzing the grounding current of a core clamp based on a wavelet transform algorithm, provided in an embodiment of the present invention;

[0032] Figure 2 This is a structural block diagram of a core clamp grounding current analysis system based on wavelet transform algorithm provided in an embodiment of the present invention;

[0033] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0034] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, 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.

[0035] Please see Figure 1 The diagram shows a flowchart of a method for analyzing the grounding current of iron core clamps based on wavelet transform algorithm according to this application.

[0036] like Figure 1 As shown, the method for analyzing the grounding current of iron core clamps based on the wavelet transform algorithm specifically includes the following steps:

[0037] Step S101: Collect the grounding current signal of the core clamp of the UHV converter transformer to obtain the original current sequence.

[0038] In this step, a high-precision leakage current sensor detects the leakage current in the transformer core / clamping and converts it into a corresponding analog small signal. Subsequently, the signal is amplified and conditioned, and then sampled and converted from analog to digital by a high-precision ADC. In a preferred embodiment of the invention, the sampling frequency is set to fs = 10kHz, the number of sampling points is N = 10000, corresponding to a 1-second signal duration. The ADC uses a 16-bit high-precision chip from Analog Devices (ADI), with a throughput rate of up to 250KSPS and a dynamic range of 93.8dB, ensuring the accuracy of wideband signal acquisition.

[0039] Step S102: Perform wavelet packet decomposition of the original current sequence using a complete binary tree structure. During the decomposition process, each layer simultaneously performs recursive decomposition of the low-frequency approximation component and the high-frequency detail component to obtain the wavelet packet coefficients of each decomposition node.

[0040] In this step, the wavelet packet function in the wavelet packet decomposition of the complete binary tree structure satisfies the following recurrence relation:

[0041] ,

[0042] ,

[0043] In the formula, For wavelet packet basis functions, These are the coefficients of the low-pass filter. These are the coefficients of the high-pass filter.

[0044] Wavelet packet coefficients are calculated using the following formula:

[0045] ,

[0046] ,

[0047] In the formula, To decompose the scale, The translation coefficient is... This is the grounding current signal.

[0048] Step S103: Using Shannon entropy as a sparsity measure, calculate the cost value corresponding to the wavelet packet coefficients of each decomposition node. Adopt a bottom-up greedy search strategy, compare the sum of the cost values ​​of the parent node and the child node, and adaptively select the optimal basis that makes the signal representation most compact among all orthogonal wavelet packet bases to obtain an adaptive decomposition structure that matches the time-frequency characteristics of the grounding current.

[0049] In this step, for each decomposition node, the wavelet packet coefficient vector Calculate the energy normalization probability:

[0050] ,

[0051] Calculate the Shannon entropy value:

[0052] ,

[0053] In the formula, For the coefficient of total energy, It is a small positive number. For the sake of value.

[0054] In the bottom-up greedy search strategy, starting from the deepest level of the decomposition tree, for each non-leaf node... ,set up For nodes Its own value and Low-frequency sub-nodes and high-frequency child nodes The optimal cost of the subtree is then the node The optimal cost of the subtree is:

[0055] ,

[0056] like Then retain the node. Then prune its subtrees; otherwise, retain the child nodes and continue decomposing downwards.

[0057] Step S104: Based on the adaptive decomposition structure, extract the high-frequency detail coefficients of each decomposition layer and calculate the frequency band energy density gradient of each layer.

[0058] In this step, the band energy density gradient is calculated using the following formula:

[0059] ,

[0060] In the formula, For the first High-frequency detail coefficients and bandwidth of the layer Time resolution , The sampling frequency.

[0061] Step S105: Select the fault-free period in the ground current signal as the background interval, calculate the percentile of the frequency band energy density gradient in each background interval, and set the weight coefficient according to the number of decomposition layers to obtain the adaptive detection threshold of each layer.

[0062] In this step, the adaptive detection threshold is calculated using the following formula:

[0063] ,

[0064] In the formula, This represents the 95th quantile of the energy density gradient value of the j-th frequency band within the background interval. These are the weighting coefficients for the third and fourth layers. For the first, second and fifth floors .

[0065] Step S106: Scan the energy density gradient sequence of each frequency band, detect intervals that continuously exceed the adaptive detection threshold, and identify valid mutation events based on the interval duration and energy magnitude.

[0066] The criteria for identifying valid mutation events are:

[0067] Let the starting index of the candidate interval be... The ending index is Then the event duration and total event energy in this interval are respectively:

[0068] ,

[0069] ,

[0070] when and When this interval is determined to be a valid mutation event, then... This represents the mean value of the energy density gradient of the corresponding layer frequency band within the background interval.

[0071] Step S107: Merge adjacent valid abrupt events within the same layer whose time interval is less than a preset value, and output the occurrence time, duration, dominant decomposition layer, and event energy of each abrupt event as the detection result of the abrupt change characteristics of the grounding current.

[0072] In summary, the method of this application acquires ground current signals; performs wavelet packet decomposition of the signals using a complete binary tree structure, employs Shannon entropy as the cost function, and adopts a bottom-up greedy strategy to adaptively select the optimal wavelet packet basis; extracts high-frequency detail coefficients of each decomposition layer and calculates the frequency band energy density gradient; establishes an adaptive threshold based on the 95th quantile of the background interval, identifies effective abrupt events by combining duration and energy constraints, and merges adjacent events with a time interval of less than 5ms; achieves accurate matching of the signal's time-frequency structure through adaptive decomposition of the optimal wavelet packet basis, improves the sensitivity of weak abrupt event detection through the frequency band energy density gradient, and effectively suppresses noise false alarms through the adaptive threshold and energy screening mechanism, thus accurately detecting ground current abrupt features caused by faults such as multi-point grounding and partial discharge.

[0073] Please see Figure 2 The diagram shows a structural block diagram of a core clamp grounding current analysis system based on wavelet transform algorithm according to this application.

[0074] like Figure 2 As shown, the iron core clamp grounding current analysis system 200 based on wavelet transform algorithm includes an acquisition module 210, a decomposition module 220, a first calculation module 230, a second calculation module 240, a third calculation module 250, an identification module 260, and an output module 270.

[0075] The acquisition module 210 is configured to acquire the grounding current signal of the core clamp of the UHV converter transformer to obtain the original current sequence; the decomposition module 220 is configured to perform wavelet packet decomposition of the original current sequence using a complete binary tree structure, wherein each layer of the decomposition process simultaneously performs recursive decomposition on the low-frequency approximation component and the high-frequency detail component to obtain the wavelet packet coefficients of each decomposition node; the first calculation module 230 is configured to use Shannon entropy as a sparsity measure to calculate the cost value corresponding to the wavelet packet coefficients of each decomposition node, adopt a bottom-up greedy search strategy, compare the sum of the cost values ​​of the parent node and the child node, and adaptively select the optimal basis that makes the signal representation most compact among all orthogonal wavelet packet bases to obtain an adaptive decomposition structure that matches the time-frequency characteristics of the grounding current; the second calculation module 240 is configured to perform decomposition based on the... An adaptive decomposition structure is used to extract high-frequency detail coefficients from each decomposition layer and calculate the frequency band energy density gradient of each layer. A third calculation module 250 is configured to select a fault-free period in the ground current signal as a background interval, calculate the percentile of the frequency band energy density gradient within the background interval of each layer, and set weight coefficients according to the number of decomposition layers to obtain the adaptive detection threshold for each layer. An identification module 260 is configured to scan the frequency band energy density gradient sequence of each layer, detect intervals that continuously exceed the adaptive detection threshold, and identify valid mutation events based on the interval duration and energy magnitude. An output module 270 is configured to merge adjacent valid mutation events with a time interval less than a preset value within the same layer, and output the occurrence time, duration, dominant decomposition layer, and event energy of each mutation event as the mutation feature detection result of the ground current.

[0076] It should be understood that Figure 2 The modules and references described in the document Figure 1 The steps described in the text correspond to those in the method described above. Therefore, the operations, features, and corresponding technical effects described above also apply to the method described in the text. Figure 2 The various modules in the document will not be described in detail here.

[0077] In other embodiments, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor performs the wavelet transform algorithm-based grounding current analysis method for iron core clamps in any of the above method embodiments.

[0078] In one embodiment, the computer-readable storage medium of the present invention stores computer-executable instructions, which are configured as follows:

[0079] The grounding current signal of the core clamp of the UHV converter transformer was collected to obtain the original current sequence;

[0080] The original current sequence is subjected to wavelet packet decomposition with a complete binary tree structure. During the decomposition process, each layer simultaneously performs recursive decomposition on both the low-frequency approximation component and the high-frequency detail component to obtain the wavelet packet coefficients of each decomposition node.

[0081] Using Shannon entropy as a sparsity measure, the cost value corresponding to the wavelet packet coefficients of each decomposition node is calculated. A bottom-up greedy search strategy is adopted to compare the sum of the cost values ​​of the parent node and the child node. The optimal basis that makes the signal representation most compact is adaptively selected from all orthogonal wavelet packet bases, resulting in an adaptive decomposition structure that matches the time-frequency characteristics of the grounding current.

[0082] Based on the adaptive decomposition structure, high-frequency detail coefficients of each decomposition layer are extracted, and the frequency band energy density gradient of each layer is calculated.

[0083] The fault-free period in the ground current signal is selected as the background interval. The percentile of the frequency band energy density gradient in each background interval is calculated. The weighting coefficient is set according to the number of decomposition layers to obtain the adaptive detection threshold of each layer.

[0084] The energy density gradient sequence of each frequency band is scanned to detect intervals that continuously exceed the adaptive detection threshold, and valid mutation events are identified based on the interval duration and energy magnitude.

[0085] Merge adjacent valid mutation events within the same layer whose time interval is less than a preset value, and output the occurrence time, duration, dominant decomposition layer, and event energy of each mutation event as the detection result of the mutation characteristics of the grounding current.

[0086] Computer-readable storage media may include a stored program area and a stored data area, wherein the stored program area may store an operating system and an application program required for at least one function; the stored data area may store data created based on the use of the wavelet transform-based iron core clamp grounding current analysis system, etc. Furthermore, the computer-readable storage medium may include high-speed random access memory, and may also include memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, the computer-readable storage medium may optionally include memory remotely located relative to a processor, which can be connected to the wavelet transform-based iron core clamp grounding current analysis system via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0087] Figure 3 This is a schematic diagram of the structure of the electronic device provided in the embodiment of the present invention, such as... Figure 3As shown, the device includes a processor 310 and a memory 320. The electronic device may also include an input device 330 and an output device 340. The processor 310, memory 320, input device 330, and output device 340 can be connected via a bus or other means. Figure 3 Taking a bus connection as an example, the memory 320 is the computer-readable storage medium described above. The processor 310 executes various server functions and data processing by running non-volatile software programs, instructions, and modules stored in the memory 320, thereby implementing the wavelet transform-based grounding current analysis method for iron core clamps described in the above embodiment. The input device 330 can receive input digital or character information and generate key signal inputs related to user settings and function control of the wavelet transform-based iron core clamp grounding current analysis system. The output device 340 may include a display screen or other display device.

[0088] The aforementioned electronic device can execute the method provided in the embodiments of the present invention, and has the corresponding functional modules and beneficial effects for executing the method. Technical details not described in detail in this embodiment can be found in the method provided in the embodiments of the present invention.

[0089] In one implementation, the above-described electronic device is applied to a core clamp grounding current analysis system based on wavelet transform algorithm, serving as a client, and includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to:

[0090] The grounding current signal of the core clamp of the UHV converter transformer was collected to obtain the original current sequence;

[0091] The original current sequence is subjected to wavelet packet decomposition with a complete binary tree structure. During the decomposition process, each layer simultaneously performs recursive decomposition on both the low-frequency approximation component and the high-frequency detail component to obtain the wavelet packet coefficients of each decomposition node.

[0092] Using Shannon entropy as a sparsity measure, the cost value corresponding to the wavelet packet coefficients of each decomposition node is calculated. A bottom-up greedy search strategy is adopted to compare the sum of the cost values ​​of the parent node and the child node. The optimal basis that makes the signal representation most compact is adaptively selected from all orthogonal wavelet packet bases, resulting in an adaptive decomposition structure that matches the time-frequency characteristics of the grounding current.

[0093] Based on the adaptive decomposition structure, high-frequency detail coefficients of each decomposition layer are extracted, and the frequency band energy density gradient of each layer is calculated.

[0094] The fault-free period in the ground current signal is selected as the background interval. The percentile of the frequency band energy density gradient in each background interval is calculated. The weighting coefficient is set according to the number of decomposition layers to obtain the adaptive detection threshold of each layer.

[0095] The energy density gradient sequence of each frequency band is scanned to detect intervals that continuously exceed the adaptive detection threshold, and valid mutation events are identified based on the interval duration and energy magnitude.

[0096] Merge adjacent valid mutation events within the same layer whose time interval is less than a preset value, and output the occurrence time, duration, dominant decomposition layer, and event energy of each mutation event as the detection result of the mutation characteristics of the grounding current.

[0097] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0098] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for analyzing grounding current of iron core clamps based on wavelet transform algorithm, characterized in that, include: The grounding current signal of the core clamp of the UHV converter transformer was collected to obtain the original current sequence; The original current sequence is subjected to wavelet packet decomposition with a complete binary tree structure. During the decomposition process, each layer simultaneously performs recursive decomposition on both the low-frequency approximation component and the high-frequency detail component to obtain the wavelet packet coefficients of each decomposition node. Using Shannon entropy as a sparsity measure, the cost value corresponding to the wavelet packet coefficients of each decomposition node is calculated. A bottom-up greedy search strategy is adopted to compare the sum of the cost values ​​of the parent node and the child node. The optimal basis that makes the signal representation most compact is adaptively selected from all orthogonal wavelet packet bases, resulting in an adaptive decomposition structure that matches the time-frequency characteristics of the grounding current. Based on the adaptive decomposition structure, high-frequency detail coefficients of each decomposition layer are extracted, and the frequency band energy density gradient of each layer is calculated. The fault-free period in the ground current signal is selected as the background interval. The percentile of the frequency band energy density gradient in each background interval is calculated. The weighting coefficient is set according to the number of decomposition layers to obtain the adaptive detection threshold of each layer. The energy density gradient sequence of each frequency band is scanned to detect intervals that continuously exceed the adaptive detection threshold, and valid mutation events are identified based on the interval duration and energy magnitude. Merge adjacent valid mutation events within the same layer whose time interval is less than a preset value, and output the occurrence time, duration, dominant decomposition layer, and event energy of each mutation event as the detection result of the mutation characteristics of the grounding current.

2. The method for analyzing grounding current of iron core clamps based on wavelet transform algorithm according to claim 1, characterized in that, In the wavelet packet decomposition of the complete binary tree structure, the wavelet packet function satisfies the following recurrence relation: ， ， In the formula, For wavelet packet basis functions, These are the coefficients of the low-pass filter. These are the coefficients of the high-pass filter.

3. The method for analyzing grounding current of iron core clamps based on wavelet transform algorithm according to claim 2, characterized in that, The wavelet packet coefficients are calculated using the following formula: ， ， In the formula, To decompose the scale, The translation coefficient is... This is the grounding current signal.

4. The method for analyzing grounding current of iron core clamps based on wavelet transform algorithm according to claim 1, characterized in that, The specific method for calculating the cost value using Shannon entropy as a measure of sparsity is as follows: For the wavelet packet coefficient vector of each decomposition node Calculate the energy normalization probability: ， Calculate the Shannon entropy value: ， In the formula, For the coefficient of total energy, It is a small positive number. For the sake of value.

5. The method for analyzing grounding current of iron core clamps based on wavelet transform algorithm according to claim 1, characterized in that, In the bottom-up greedy search strategy, starting from the deepest level of the decomposition tree, for each non-leaf node... ,set up For nodes Its own value and Low-frequency sub-nodes and high-frequency child nodes The optimal cost of the subtree is then the node The optimal cost of the subtree is: ， like Then retain the node. Then prune its subtrees; otherwise, retain the child nodes and continue decomposing downwards.

6. The method for analyzing grounding current of iron core clamps based on wavelet transform algorithm according to claim 1, characterized in that, The frequency band energy density gradient is calculated using the following formula: ， In the formula, For the first High-frequency detail coefficients and bandwidth of the layer Time resolution , The sampling frequency.

7. The method for analyzing grounding current of iron core clamps based on wavelet transform algorithm according to claim 1, characterized in that, The adaptive detection threshold is calculated using the following formula: ， In the formula, This represents the 95th quantile of the energy density gradient value of the j-th frequency band within the background interval. These are the weighting coefficients for the third and fourth layers. For the first, second and fifth floors .

8. The method for analyzing grounding current of iron core clamps based on wavelet transform algorithm according to claim 1, characterized in that, The criteria for identifying valid mutation events are as follows: Let the starting index of the candidate interval be... The ending index is Then the event duration and total event energy in this interval are respectively: ， ， when and When this interval is determined to be a valid mutation event, then... This represents the mean value of the energy density gradient of the corresponding layer frequency band within the background interval.

9. A system for analyzing grounding current of iron core clamps based on wavelet transform algorithm, characterized in that, include: The acquisition module is configured to collect the grounding current signal of the core clamp of the UHV converter transformer to obtain the original current sequence; The decomposition module is configured to perform wavelet packet decomposition of the original current sequence in a complete binary tree structure. During the decomposition process, each layer simultaneously performs recursive decomposition of the low-frequency approximation component and the high-frequency detail component to obtain the wavelet packet coefficients of each decomposition node. The first calculation module is configured to use Shannon entropy as a sparsity measure to calculate the cost value corresponding to the wavelet packet coefficients of each decomposition node. It adopts a bottom-up greedy search strategy, compares the sum of the cost value of the parent node and the cost value of the child node, and adaptively selects the optimal basis that makes the signal representation most compact among all orthogonal wavelet packet bases to obtain an adaptive decomposition structure that matches the time-frequency characteristics of the grounding current. The second calculation module is configured to extract the high-frequency detail coefficients of each decomposition layer according to the adaptive decomposition structure and calculate the frequency band energy density gradient of each layer. The third calculation module is configured to select the fault-free period in the ground current signal as the background interval, calculate the percentile of the frequency band energy density gradient in each background interval, and set the weight coefficient according to the number of decomposition layers to obtain the adaptive detection threshold of each layer. The identification module is configured to scan the energy density gradient sequence of each frequency band, detect intervals that continuously exceed the adaptive detection threshold, and identify valid mutation events based on the interval duration and energy magnitude. The output module is configured to merge adjacent valid mutation events with a time interval of less than a preset value within the same layer, and output the occurrence time, duration, dominant decomposition layer and event energy of each mutation event as the mutation characteristic detection result of the ground current.

Images