Switchgear live display online monitoring method based on signal processing

Abstract

The application relates to the technical field of digital data processing, in particular to a switch cabinet live display online monitoring method based on signal processing, which comprises the following steps: calculating a pulse current trend index according to the change trend of a pulse current signal of a switch cabinet live display and the frequency of rising changes in the signal; combining the characteristics of partial discharge, calculating a partial discharge influence coefficient of the pulse current signal according to the change rate, area and duration of each wave peak region in the pulse current signal; obtaining an adaptive threshold value according to the partial discharge influence coefficient, carrying out wavelet denoising according to the adaptive threshold value, carrying out wavelet analysis on the denoised pulse current signal, and judging whether partial discharge occurs. Thus, the switch cabinet live display online monitoring is realized, the background noise in the pulse current signal is effectively removed, the details of the partial discharge signal are retained, and the accuracy and reliability of the switch cabinet partial discharge online monitoring are improved.

Description

Technical Field

[0001] This application relates to the field of digital data processing technology, specifically to an online monitoring method for a switchgear live display based on signal processing. Background Technology

[0002] Switchgear, as a core component of power systems, bears the important responsibility of power distribution, control, and protection. Its health directly affects the reliability and security of the power grid. In recent years, partial discharge, as an early warning signal of insulation degradation in power equipment, has received widespread attention. The occurrence of partial discharge is often a direct reflection of potential faults such as aging of insulation materials, moisture absorption, and mechanical damage. Timely detection and assessment of partial discharge activity are of great significance for preventing major electrical accidents. Against this backdrop, the development of efficient and accurate online monitoring technology is particularly urgent, aiming to monitor the operating status of switchgear in real time, promptly detect and locate partial discharge phenomena, and provide a scientific basis for preventive maintenance.

[0003] Existing partial discharge detection methods have limitations, are difficult to operate, require high-level equipment modification, are complex to install, and are costly, which limits their widespread application in large-scale power facilities. Furthermore, the internal environment of switchgear is complex, with electromagnetic interference and mechanical vibration being prevalent. These factors can easily overlap with partial discharge signals, resulting in a low signal-to-noise ratio in the monitoring data and increasing the difficulty of signal analysis. Summary of the Invention

[0004] To address the aforementioned technical problems, this application provides an online monitoring method for switchgear live indicators based on signal processing, thereby resolving the existing issues.

[0005] The online monitoring method for switchgear live indicator based on signal processing in this application adopts the following technical solution:

[0006] One embodiment of this application provides an online monitoring method for a switchgear live indicator based on signal processing, the method comprising the following steps:

[0007] (1) Collect the pulse current signal of the switch cabinet's live display to obtain the pulse current sequence;

[0008] (2) Based on the pulse current sequence, construct the moving average sequence and rising frequency sequence of the pulse current using the moving average method; obtain the trend index of the pulse current based on the data change trend in the moving average sequence and rising frequency sequence.

[0009] (3) Divide the time-domain waveform of the pulse current signal to obtain each peak region; obtain the change coefficient of each peak region based on the rise and fall speed of the data in each peak region; obtain the partial discharge influence coefficient of the pulse current based on the change coefficient and trend index.

[0010] (4) The pulse current signal is denoised by combining the partial discharge influence coefficient of the pulse current signal with a denoising algorithm;

[0011] (5) Perform wavelet analysis on the denoised pulse current signal to determine whether partial discharge has occurred.

[0012] In one embodiment, the process of constructing the moving average sequence and the rising frequency sequence of the pulse current is as follows:

[0013] The pulse current sequence is used as the input of the moving average method. The length of the subsequence when performing the moving average is set, and the output is each moving average number. The sequence composed of all moving average numbers is used as the moving average sequence of the pulse current.

[0014] Obtain the subsequence corresponding to each moving average calculation, and count the frequency of rising conditions in each subsequence. The rising condition is defined as the latter element being greater than the former element in any two adjacent elements within a subsequence. The sequence composed of the frequencies of the rising conditions in all subsequences is denoted as the rising frequency sequence of the pulse current.

[0015] In one embodiment, the process of obtaining the trend index of the pulse current is as follows:

[0016] The growth trend of pulse current is constructed based on the changes in data in the moving average sequence and rising frequency sequence of pulse current.

[0017] The non-uniform growth exponent of the moving average sequence of pulse current is calculated based on the changes of adjacent elements in the moving average sequence of pulse current.

[0018] Based on the aforementioned growth trend and the aforementioned non-uniform growth index, the trend index of the pulse current is determined;

[0019] The trend index of pulse current is positively correlated with the growth trend of pulse current and negatively correlated with the non-uniform growth index of the moving average sequence of pulse current.

[0020] In one embodiment, the process of obtaining the growth trend of the pulse current is as follows:

[0021] Obtain the first-order difference sequence of the rising frequency sequence of the pulse current, calculate the sum of all elements in the first-order difference sequence, and denote it as the first sum.

[0022] In the moving average sequence of pulse current, the growth rate between any two adjacent elements is calculated; among all the growth rates in the moving average sequence, the proportion of positive growth rates is statistically analyzed; the fluctuation rise coefficient of pulse current is positively correlated with the growth rate between any two adjacent elements and the proportion of positive growth rates, respectively.

[0023] Based on the first sum and the fluctuation rise coefficient of the pulse current, the growth trend of the pulse current is determined;

[0024] The growth trend of the pulse current is positively correlated with the first sum of the pulse current and the fluctuation rise coefficient.

[0025] In one embodiment, the non-uniform growth index of the moving average sequence of the pulse current is: the degree of dispersion of the growth rate between any two adjacent elements in the moving average sequence of the pulse current.

[0026] In one embodiment, the process of dividing the time-domain waveform of the pulse current signal to obtain each peak region specifically involves:

[0027] In the time-domain waveform diagram of the pulse current signal, the point where the pulse current signal waveform intersects with the baseline is taken as the dividing point, and the area enclosed by the pulse current signal waveform and the baseline between two adjacent dividing points is taken as a peak region.

[0028] In one embodiment, the process of obtaining the variation coefficients of each peak region is as follows:

[0029] The rate at which each peak region rises from the baseline to the peak value is taken as the rate of abrupt rise of each peak region.

[0030] The rate at which each peak region descends from its peak value to the baseline is taken as the abrupt descent rate of each peak region.

[0031] Calculate the difference between the abrupt increase rate and the abrupt decrease rate in each peak region, and denote it as the first difference; calculate the result of the exponential function with the natural constant as the base and the first difference as the exponent, and denote it as the average rate of change in each peak region.

[0032] The mean values ​​of the abrupt increase rate and the abrupt decrease rate of each peak region are calculated and denoted as the first mean value. The degree of abrupt change in each peak region is positively correlated with the area per unit time of each peak region and the first mean value.

[0033] Based on the average rate of change and the degree of abrupt change in each peak region, the change coefficient of each peak region is determined.

[0034] The coefficient of change of each peak region is positively correlated with the degree of abrupt change of each peak region, and negatively correlated with the average rate of change of each peak region.

[0035] In one embodiment, the process of obtaining the partial discharge influence coefficient of the pulse current is as follows:

[0036] Based on the pulse current sequence corresponding to the pulse current signal in each peak region, the trend index of the pulse current in each peak region is obtained using the same calculation method as the trend index.

[0037] The current mutation rate index of each peak region is positively correlated with the change coefficient of each peak region and the trend index of the corresponding pulse current.

[0038] The positive fusion result of the current mutation rate exponents of all peak regions is used as the partial discharge influence coefficient of the pulse current signal.

[0039] In one embodiment, the process of denoising the pulse current signal is as follows:

[0040] The original threshold of the wavelet threshold denoising algorithm is determined by using a threshold calculation method. The ratio of the original threshold to the local discharge influence coefficient after normalization of the pulse current signal is used as the adaptive threshold. The adjusted adaptive threshold is used as the segmentation threshold, and wavelet denoising is performed to obtain the denoised pulse current signal.

[0041] In one embodiment, performing wavelet analysis on the denoised pulse current signal to determine whether partial discharge has occurred specifically involves:

[0042] The denoised pulse current signal is used as input, and wavelet analysis is used to identify partial discharge. The output is the partial discharge judgment result of the pulse current signal.

[0043] This application has at least the following beneficial effects:

[0044] This application calculates the pulse current trend index based on the changing trend of the pulse current signal and the frequency of its rising changes. Then, combining the characteristics of partial discharge, it calculates the current mutation rate index based on the rate of change, area, and duration of each peak region in the pulse current signal, thereby obtaining the partial discharge influence coefficient of the pulse current signal. Finally, it performs adaptive threshold adjustment based on the partial discharge influence coefficient and uses the adjusted adaptive threshold as the segmentation threshold, employing a soft thresholding function for wavelet denoising. This approach eliminates the need for complex modifications to the switchgear and simultaneously considers the changing trend of the pulse current signal and signal changes similar to partial discharge characteristics within the signal. It can more accurately distinguish between partial discharge signals and non-partial discharge noise, and the adaptive threshold adjustment effectively removes background noise while preserving the details of the partial discharge signal to the maximum extent, ensuring the accuracy and completeness of signal analysis and improving the accuracy and reliability of online monitoring of partial discharge in switchgear. Attached Figure Description

[0045] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 A flowchart of the online monitoring method for a switchgear live indicator based on signal processing provided in this application;

[0047] Figure 2 This is a schematic diagram illustrating the steps of an online monitoring method for a switchgear live indicator based on signal processing.

[0048] Figure 3 A schematic diagram illustrating the steps to obtain the trend index of pulse current. Detailed Implementation

[0049] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the online monitoring method for a switchgear live display based on signal processing proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0050] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0051] The following description, in conjunction with the accompanying drawings, details the specific scheme of the online monitoring method for switchgear live indicators based on signal processing provided in this application.

[0052] One embodiment of this application provides an online monitoring method for switchgear live display based on signal processing.

[0053] Specifically, the following online monitoring method for switchgear live indicator based on signal processing is provided. Please refer to [link / reference]. Figure 1 The method includes the following steps:

[0054] Step S1: Collect the pulse current signal from the live indicator of the switchgear to obtain the pulse current sequence.

[0055] By using the internal sensor of the live display as a partial discharge sensor, pulse current signals are acquired at the phase hole of the live display in the switchgear. The pulse current signals are obtained and sent to an oscilloscope to obtain the time-domain waveform of the pulse current signals. The discrete numerical sequence corresponding to the time-domain waveform of the pulse current signals is output and the output result is recorded as the pulse current sequence.

[0056] Step S2: Based on the pulse current sequence, construct the moving average sequence and rising frequency sequence of the pulse current using the moving average method; obtain the trend index of the pulse current based on the data change trends in the moving average sequence and rising frequency sequence.

[0057] Since the normal operating frequency of current in a power system is 50–60 Hz, while the frequency of partial discharge pulse current signals is much higher than the normal operating frequency, typically ranging from several hundred kHz to GHz, the acquired pulse current signal is a current signal with a frequency much higher than the normal operating frequency. However, various noise sources in a power system can also cause high-frequency current fluctuations, making it possible that the acquired pulse current signal is not caused by partial discharge. In this case, it is necessary to analyze the acquired pulse current signal based on the characteristics of partial discharge to determine whether it is a partial discharge pulse current signal.

[0058] (1) Based on the pulse current sequence, a moving average sequence and a rising frequency sequence of the pulse current are constructed using the moving average method, specifically as follows:

[0059] During the operation of power equipment, partial discharge is caused by factors such as the deterioration of insulation materials. However, partial discharge may also be affected by the operating environment at the time, so it cannot be guaranteed that the severity of partial discharge increases monotonically. However, the deterioration of insulation materials is irreversible. Before maintenance or repair, the degree of deterioration becomes more and more serious. Therefore, by observing the frequency of partial discharge and the overall trend of changes over a long period of time, there should be a characteristic that the frequency of partial discharge is increasing and the overall abnormal trend of partial discharge pulse current is rising.

[0060] The pulse current sequence is used as the input of the moving average method. The subsequence length is set when performing the moving average. It should be noted that the implementer can set the subsequence length himself. In this embodiment, the subsequence length is set to 20. The output is each moving average number. The sequence composed of all moving average numbers is used as the moving average sequence of the pulse current.

[0061] Obtain the subsequence corresponding to each moving average calculation, and count the frequency of rising conditions in each subsequence. The rising condition is defined as the latter element being greater than the former element in any two adjacent elements within a subsequence. The sequence composed of the frequencies of the rising conditions in all subsequences is denoted as the rising frequency sequence of the pulse current.

[0062] Preferably, as an embodiment of this application, the sequence of all moving averages arranged in ascending order of time can be used as the moving average sequence of the pulse current; the sequence of the frequency of rise of all subsequences arranged in ascending order of time can be used as the rising frequency sequence of the pulse current.

[0063] (2) The growth trend of pulse current is constructed based on the data changes in the moving average sequence and rising frequency sequence of pulse current, specifically:

[0064] Obtain the first-order difference sequence of the rising frequency sequence of the pulse current, calculate the sum of all elements in the first-order difference sequence, and denote it as the first sum.

[0065] In the moving average sequence of pulse current, the growth rate between any two adjacent elements is calculated; among all the growth rates in the moving average sequence, the proportion of positive growth rates is statistically analyzed; the fluctuation rise coefficient of pulse current is positively correlated with the growth rate between any two adjacent elements and the proportion of positive growth rates, respectively.

[0066] The growth trend of the pulse current is positively correlated with the first sum of the pulse current and the fluctuation rise coefficient.

[0067] Preferably, as an embodiment of this application, the sum of the growth rates between any two adjacent elements of the moving average sequence is calculated and denoted as the growth rate sum; the fluctuation rise coefficient of the pulse current can be the product of the growth rate sum and the proportion of the positive growth rate; the growth trend of the pulse current can be the product of the first sum of the pulse current and the fluctuation rise coefficient.

[0068] If a pulse current signal may be caused by partial discharge, then the pulse current signal will generally show an upward trend, and the frequency of partial discharge will increase. As a result, the overall frequency sequence will show an increasing trend, and the pulse current signal will also fluctuate upward. In this case, the pulse current growth trend is relatively large. Conversely, the smaller the pulse current growth trend, the lower the probability of partial discharge.

[0069] (3) Calculate the non-uniform growth index of the moving average sequence of pulse current based on the changes of adjacent elements in the moving average sequence of pulse current:

[0070] The variance of the growth rate between any two adjacent elements in the moving average sequence of the pulse current is used as the non-uniform growth index of the moving average sequence of the pulse current.

[0071] To avoid a situation where the pulse current signal decreases for most of the time, only occasionally experiencing a sharp rise that ultimately leads to an upward trend, the elements in the moving average sequence of the pulse current need to grow uniformly. The more uniform the growth of elements in the moving average sequence, the more it conforms to the characteristics of partial discharge, and the lower the non-uniform growth exponent. Conversely, the worse the uniformity of element growth in the moving average sequence of the pulse current, the higher the non-uniform growth exponent, and the lower the probability of partial discharge.

[0072] (4) Calculate the trend index of pulse current based on the growth trend of pulse current and the non-uniform growth index of the moving average sequence:

[0073] The trend index of pulse current is positively correlated with the growth trend of pulse current and negatively correlated with the non-uniform growth index of the moving average sequence of pulse current.

[0074] Preferably, as an embodiment of this application, the trend index of the pulse current can be the ratio of the growth trend of the pulse current to the non-uniform growth index of the moving average sequence.

[0075] Step S3: Divide the time-domain waveform of the pulse current signal to obtain each peak region; obtain the change coefficient of each peak region based on the rise and fall speed of the data in each peak region; obtain the partial discharge influence coefficient of the pulse current according to the change coefficient and trend index.

[0076] Judging solely by the increase in pulse current can lead to misdiagnosis. For example, with the long-term operation of power equipment, various problems and noises gradually increase, causing the pulse current to show an increasing trend to some extent. Therefore, it is necessary to further screen and distinguish based on other characteristics of partial discharge. Pulse current fluctuations caused by partial discharge are often transient, sudden, and short-lived; that is, the larger the current surge occurs in a short period of time, and the shorter the recovery time, the better. In contrast, pulses caused by other types of signals or noise interference are usually not as instantaneous and drastic as those caused by partial discharge.

[0077] (1) In the time-domain waveform diagram of the pulse current signal, the point where the pulse current signal waveform intersects with the baseline is taken as the dividing point, and the area enclosed by the pulse current signal waveform and the baseline between two adjacent dividing points is taken as a peak area, and the area of ​​each peak area is obtained.

[0078] Preferably, as an embodiment of this application, the area of ​​the crest region can be obtained by using the minimum bounding matrix of each crest region, and the area of ​​the crest region can be approximated in the minimum bounding matrix using the Monte Carlo method. The Monte Carlo method is a well-known technique, and the specific process will not be described in detail.

[0079] (2) The variation coefficient of each peak region is obtained based on the rate of increase and decrease of data in each peak region, specifically:

[0080] (2.1) Calculate the rate of change of rise and the rate of change of fall in each peak region:

[0081] The rate at which each peak region rises from the baseline to the peak value is taken as the rate of abrupt rise of each peak region.

[0082] The rate at which each peak region descends from its peak value to the baseline is taken as the abrupt descent rate of each peak region.

[0083] Preferably, as an embodiment of this application, the time from the left critical point to the peak point of each peak region is recorded as the rise time of each peak region; the time from the peak point to the right critical point of each peak region is recorded as the rise time of each peak region; the speed at which each peak region rises from the baseline to the peak point can be the ratio of the height of the peak point to the rise time of each peak region; the speed at which each peak region falls from the peak point to the baseline can be the ratio of the height of the peak point to the fall time of each peak region.

[0084] (2.2) Calculate the variation coefficients of each wave crest region:

[0085] Calculate the difference between the abrupt increase rate and the abrupt decrease rate in each peak region, and denote it as the first difference; calculate the result of the exponential function with the natural constant as the base and the first difference as the exponent, and denote it as the average rate of change in each peak region.

[0086] The mean values ​​of the abrupt increase rate and the abrupt decrease rate of each peak region are calculated and denoted as the first mean value. The degree of abrupt change in each peak region is positively correlated with the area per unit time of each peak region and the first mean value.

[0087] The coefficient of change of each peak region is positively correlated with the degree of abrupt change of each peak region, and negatively correlated with the average rate of change of each peak region.

[0088] Preferably, as an embodiment of this application, the time from the left critical point to the right critical point of each peak region is taken as the time length of each peak region; the area per unit time of each peak region can be the ratio of the area of ​​each peak region to the time length; the degree of change of each peak region can be the product of the area per unit time of each peak region and the first mean; the coefficient of change of each peak region can be the ratio of the degree of change of each peak region to the average rate of change.

[0089] (3) Calculate the current mutation rate index for each peak region based on the change coefficient and trend index:

[0090] Based on the pulse current sequence corresponding to the pulse current signal in each peak region, the trend index of the pulse current in each peak region is obtained by using the same calculation method as the trend index mentioned above; the current change rate index of each peak region is positively correlated with the change coefficient of each peak region and the corresponding trend index of the pulse current.

[0091] Preferably, as an embodiment of this application, the current change rate index of each peak region can be the product of the change coefficient of each peak region and the trend index of the corresponding pulse current.

[0092] When a pulse current signal in the peak region contains a peak caused by partial discharge, the pulse current trend exponent of the pulse current signal is relatively large. Since the peak region is caused by a large current surge occurring and recovering within a very short time, both the rise and fall rates of the surge should be large and similar, meaning the average rate of change in the peak region is relatively small. The large surge and recovery within a short time also results in a large degree of change in the peak region, meaning a large coefficient of change, and thus a large current surge rate exponent. The larger this value, the more likely the peak region is to correspond to partial discharge. Conversely, if the peak region is caused by other types of signals or noise interference, the current surge rate exponent is relatively small.

[0093] (4) The positive fusion result of the current mutation rate index of all peak regions is used as the partial discharge influence coefficient of the pulse current signal.

[0094] Preferably, as an embodiment of this application, the forward fusion result can be the sum of the current mutation rate indices of all peak regions; in other embodiments, the forward fusion result can also be the multiplication or averaging of the current mutation rate indices of all peak regions, etc., which shall be determined according to the actual situation during application, and this application does not impose any special restrictions.

[0095] If the pulse current signal is generated due to partial discharge, it will have more peak regions that conform to the characteristics of partial discharge, resulting in a larger partial discharge influence coefficient in the final pulse current signal. Conversely, if the pulse current signal is generated due to various noises or other reasons, the partial discharge influence coefficient will be smaller.

[0096] Step S4: Denoise the pulse current signal based on the partial discharge influence coefficient of the pulse current signal combined with a denoising algorithm.

[0097] If the partial discharge influence coefficient of a pulse current signal is large, it indicates that there may be more partial discharge features in the signal. Therefore, more signal details need to be preserved. When performing wavelet denoising, the threshold should be lowered to retain more wavelet coefficients, thus preserving more signal details. Conversely, if the partial discharge influence coefficient is small, it indicates that the possibility of partial discharge is low, and the signal may mainly consist of noise or other non-partial discharge signals. More forceful denoising can be performed. When performing wavelet denoising, the threshold should be increased to reduce the number of wavelet coefficients, thereby obtaining a smoother and cleaner signal.

[0098] (1) Obtain the normalized value of the partial discharge influence coefficient of the pulse current signal;

[0099] (2) Calculate the adaptive adjustment threshold of the pulse current signal and perform denoising using a denoising algorithm:

[0100] The original threshold of the wavelet threshold denoising algorithm is determined by using the VisuShrink threshold calculation method. The ratio of the original threshold to the local discharge influence coefficient after normalization of the pulse current signal is used as the adaptive threshold. The adaptive threshold is then used as the segmentation threshold, and wavelet denoising is performed using a soft threshold function to obtain the denoised pulse current signal.

[0101] It should be noted that implementers may also use other threshold calculation methods to determine the original threshold of the wavelet threshold denoising algorithm, and this embodiment does not impose specific restrictions.

[0102] Step S5: Perform wavelet analysis on the denoised pulse current signal to determine whether partial discharge has occurred.

[0103] The denoised pulse current signal is used as input, and wavelet analysis is employed to identify partial discharges. The output is the partial discharge judgment result based on the pulse current signal. Partial discharge identification based on wavelet analysis is a well-known technique, and the specific process will not be elaborated further. When identifying partial discharges using wavelet analysis, the features of the partial discharge signal are first extracted in the wavelet domain. The denoised signal makes the features of the partial discharge more apparent, thereby improving the accuracy of online monitoring of partial discharges in switchgear.

[0104] The steps of the above method are illustrated in the diagram below. Figure 2 As shown; a schematic diagram illustrating the steps for obtaining the trend index of pulse current is shown below. Figure 3 As shown.

[0105] In summary, this embodiment of the application calculates the pulse current trend index by analyzing the changing trend of the pulse current signal and the frequency of its rising changes. Then, combining the characteristics of partial discharge, it calculates the current mutation rate index based on the rate of change, area, and duration of each peak region in the pulse current signal, thereby obtaining the partial discharge influence coefficient of the pulse current signal. Finally, it performs adaptive threshold adjustment based on the partial discharge influence coefficient and uses the adjusted adaptive threshold as the segmentation threshold, employing a soft thresholding function for wavelet denoising. This approach eliminates the need for complex modifications to the switchgear and simultaneously considers the changing trend of the pulse current signal and signal changes similar to partial discharge characteristics within the signal. It can more accurately distinguish between partial discharge signals and non-partial discharge noise, and the adaptive threshold adjustment effectively removes background noise while preserving the details of the partial discharge signal to the maximum extent, ensuring the accuracy and completeness of signal analysis and improving the accuracy and reliability of online monitoring of partial discharge in switchgear.

[0106] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this application. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0107] The various embodiments in this application are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0108] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them; modifications to the technical solutions described in the foregoing embodiments, or equivalent substitutions of some of the technical features, do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for online monitoring of switchgear live indicators based on signal processing, characterized in that, The method includes the following steps: (1) Collect the pulse current signal of the switch cabinet's live display to obtain the pulse current sequence; (2) Based on the pulse current sequence, construct the moving average sequence and rising frequency sequence of the pulse current using the moving average method; obtain the trend index of the pulse current based on the data change trend in the moving average sequence and rising frequency sequence. (3) Divide the time-domain waveform of the pulse current signal to obtain each peak region; obtain the change coefficient of each peak region based on the rise and fall speed of the data in each peak region; obtain the partial discharge influence coefficient of the pulse current based on the change coefficient and trend index. (4) The pulse current signal is denoised by combining the partial discharge influence coefficient of the pulse current signal with a denoising algorithm; (5) Perform wavelet analysis on the denoised pulse current signal to determine whether partial discharge has occurred.

2. The online monitoring method for switchgear live indicator based on signal processing as described in claim 1, characterized in that, The process of constructing the moving average sequence and rising frequency sequence of the pulse current is as follows: The pulse current sequence is used as the input of the moving average method. The length of the subsequence when performing the moving average is set, and the output is each moving average number. The sequence composed of all moving average numbers is used as the moving average sequence of the pulse current. Obtain the subsequence corresponding to each moving average calculation, and count the frequency of rising conditions in each subsequence. The rising condition is defined as the latter element being greater than the former element in any two adjacent elements within a subsequence. The sequence composed of the frequencies of the rising conditions in all subsequences is denoted as the rising frequency sequence of the pulse current.

3. The online monitoring method for switchgear live indicator based on signal processing as described in claim 1, characterized in that, The process for obtaining the trend index of the pulse current is as follows: The growth trend of pulse current is constructed based on the changes in data in the moving average sequence and rising frequency sequence of pulse current. The non-uniform growth exponent of the moving average sequence of pulse current is calculated based on the changes of adjacent elements in the moving average sequence of pulse current. Based on the aforementioned growth trend and the aforementioned non-uniform growth index, the trend index of the pulse current is determined; The trend index of pulse current is positively correlated with the growth trend of pulse current and negatively correlated with the non-uniform growth index of the moving average sequence of pulse current.

4. The online monitoring method for switchgear live indicator based on signal processing as described in claim 3, characterized in that, The process for obtaining the growth trend of the pulse current is as follows: Obtain the first-order difference sequence of the rising frequency sequence of the pulse current, calculate the sum of all elements in the first-order difference sequence, and denote it as the first sum. In the moving average sequence of pulse current, the growth rate between any two adjacent elements is calculated; among all the growth rates in the moving average sequence, the proportion of positive growth rates is statistically analyzed; the fluctuation rise coefficient of pulse current is positively correlated with the growth rate between any two adjacent elements and the proportion of positive growth rates, respectively. Based on the first sum and the fluctuation rise coefficient of the pulse current, the growth trend of the pulse current is determined; The growth trend of the pulse current is positively correlated with the first sum of the pulse current and the fluctuation rise coefficient.

5. The online monitoring method for switchgear live indicator based on signal processing as described in claim 4, characterized in that, The non-uniform growth index of the moving average sequence of the pulse current is: the degree of dispersion of the growth rate between any two adjacent elements in the moving average sequence of the pulse current.

6. The online monitoring method for switchgear live indicator based on signal processing as described in claim 1, characterized in that, The process of dividing the time-domain waveform of the pulse current signal to obtain each peak region is as follows: In the time-domain waveform diagram of the pulse current signal, the point where the pulse current signal waveform intersects with the baseline is taken as the dividing point, and the area enclosed by the pulse current signal waveform and the baseline between two adjacent dividing points is taken as a peak region.

7. The online monitoring method for switchgear live indicator based on signal processing as described in claim 1, characterized in that, The process for obtaining the variation coefficients of each peak region is as follows: The rate at which each peak region rises from the baseline to the peak value is taken as the rate of abrupt rise of each peak region. The rate at which each peak region descends from its peak value to the baseline is taken as the abrupt descent rate of each peak region. Calculate the difference between the rate of sudden increase and the rate of sudden decrease in each peak region, and denote it as the first difference; The result of the calculation of the exponential function with the natural constant as the base and the first difference as the exponent is denoted as the average rate of change of each peak region; The mean values ​​of the abrupt increase rate and the abrupt decrease rate of each peak region are calculated and denoted as the first mean value. The degree of abrupt change in each peak region is positively correlated with the area per unit time of each peak region and the first mean value. Based on the average rate of change and the degree of abrupt change in each peak region, the change coefficient of each peak region is determined. The coefficient of change of each peak region is positively correlated with the degree of abrupt change of each peak region, and negatively correlated with the average rate of change of each peak region.

8. The online monitoring method for switchgear live indicator based on signal processing as described in claim 1, characterized in that, The process for obtaining the partial discharge influence coefficient of the pulse current is as follows: Based on the pulse current sequence corresponding to the pulse current signal in each peak region, the trend index of the pulse current in each peak region is obtained using the same calculation method as the trend index. The current mutation rate index of each peak region is positively correlated with the change coefficient of each peak region and the trend index of the corresponding pulse current. The positive fusion result of the current mutation rate exponents of all peak regions is used as the partial discharge influence coefficient of the pulse current signal.

9. The online monitoring method for switchgear live indicator based on signal processing as described in claim 1, characterized in that, The process of denoising the pulse current signal is as follows: The original threshold of the wavelet threshold denoising algorithm is determined by using a threshold calculation method. The ratio of the original threshold to the local discharge influence coefficient after normalization of the pulse current signal is used as the adaptive threshold. The adjusted adaptive threshold is used as the segmentation threshold, and wavelet denoising is performed to obtain the denoised pulse current signal.

10. The online monitoring method for switchgear live indicator based on signal processing as described in claim 1, characterized in that, The step of performing wavelet analysis on the denoised pulse current signal to determine whether partial discharge has occurred is as follows: The denoised pulse current signal is used as input, and wavelet analysis is used to identify partial discharge. The output is the partial discharge judgment result of the pulse current signal.

Images