Electric arc detection false alarm prevention method based on current spectrum sample kurtosis

Abstract

The invention discloses an arc detection false alarm prevention method based on current spectrum sample kurtosis. The method comprises the following steps: acquiring current spectrum data of a phase difference working condition by using an oscilloscope according to a sampling frequency; analyzing a probability density distribution curve of the current spectrum data, and selecting a wave crest frequency band KSpan and a peak point frequency band KArc according to spectrum characteristics when an electric arc occurs; setting a kurtosis judgment threshold value GArc according to the selected wave crest frequency band KSpan and the spectrum characteristics when the electric arc is generated; sampling PV current data in real time, selecting all sample points of a peak frequency band KSpan, obtaining a sampling point serial number K corresponding to a peak point, and calculating sample kurtosis G; when G > = GArc and K belongs to KArc, arc generation is determined. According to the invention, detection misinformation caused by noise of the photovoltaic inverter in different working environments can be effectively avoided. The false alarm rate is effectively reduced, and the traditional false alarm rate exceeding 30% is reduced to be within 5%.

Description

Technical Field

[0001] The present invention relates to an arc detection false alarm prevention method based on current spectrum sample kurtosis, belonging to the technical field related to arc detection false alarm prevention. Background Art

[0002] In the field of arc signal detection, the traditional method is to compare the performance of DC power under non-arcing conditions and select time-domain or frequency-domain amplitudes to set warning thresholds to determine whether an arc is present in the line in real time. However, PV inverters are affected by switching frequency, panel optimizer operating frequency, and the mains power environment. Even if an arc does not occur, the current signal may reach the set threshold, resulting in a serious false alarm and adversely affecting the power generation of the PV power generation system. Summary of the Invention

[0003] The purpose of the present invention is to solve the above-mentioned deficiencies in the prior art. Aiming at the problem that traditional arc detection has false alarms that affect the operation of photovoltaic power generation systems, an arc detection anti-false alarm method based on the kurtosis of current spectrum samples is proposed.

[0004] In order to achieve the above object, the technical solution adopted by the present invention is:

[0005] The arc detection and false alarm prevention method based on the kurtosis of current spectrum samples includes the following steps:

[0006] S1 uses an oscilloscope to obtain current spectrum data of phase difference working conditions according to the sampling frequency;

[0007] S2 analyzes the probability density distribution curve of the current spectrum data and selects the peak frequency segment K according to the spectrum characteristics when the arc occurs. Span and peak frequency band K Arc ;

[0008] S3 selects the peak frequency segment K Span The spectral characteristics when the arc occurs are used to set the kurtosis judgment threshold G Atc ;

[0009] S4 samples the PV current data in real time and selects the peak frequency segment K Span All sample points, obtain the sampling point sequence number K corresponding to the peak point, and calculate the sample kurtosis G;

[0010] S5 when G≥G Arc And K∈K Arc When , it is determined that an arc has occurred.

[0011] Preferably, in step S4, all sample points include x1 to xn, Where M4 represents the fourth-order sample central moment, M2 represents the sample variance, and M2 2 represents the square of the sample variance;

[0012] Where n is the number of samples, x i is the sample point amplitude, is the mean of all sample points;

[0013] Where n is the number of samples, x i is the sample point amplitude, is the mean of all sample points.

[0014] Preferably, in step S3, G Arc Greater than the kurtosis when no arc is generated.

[0015] Preferably, in step S3, G is performed according to the test results. Arc adjust, K is the peak frequency band when the arc is triggered Span The corresponding sample kurtosis G is the ensemble mean, δ is the adjustment coefficient, and δ ranges from 0.72 to 0.9.

[0016] Preferably, in step S2, K Span 2KHz~15KHz, K Arc =(3KHz, 6KHz).

[0017] Preferably, in step S1, an oscilloscope is used to obtain differentiated direct current arc tests under PV low-voltage conditions, rated voltage conditions, and high-voltage conditions to obtain current spectrum data.

[0018] The beneficial effects of the present invention are mainly reflected in:

[0019] 1. It can effectively avoid false detection caused by noise from photovoltaic inverters in different working environments.

[0020] 2. The false alarm rate is effectively reduced, from the traditional false alarm rate of over 30% to less than 5%. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Other features, objects and advantages of the present application will become more apparent upon reading the detailed description of non-limiting embodiments made with reference to the following drawings:

[0022] Figure 1 It is a flow chart of the arc detection and false alarm prevention method based on the kurtosis of current spectrum samples of the present invention.

[0023] Figure 2 This is a spectrum waveform diagram when an arc occurs under high voltage full load conditions in the present invention.

[0024] Figure 3This is a spectrum waveform diagram when an arc occurs under low voltage and full load conditions in the present invention.

[0025] Figure 4 It is a spectrum waveform diagram when arc occurs under rated voltage in the present invention. DETAILED DESCRIPTION

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings 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 making creative efforts shall fall within the scope of protection of the present invention.

[0027] The present application will be further described in detail below with reference to the accompanying drawings and examples. It will be understood that the specific embodiments described herein are intended only to explain the relevant inventions and are not intended to limit the inventions. It should also be noted that, for ease of description, only portions relevant to the relevant inventions are shown in the accompanying drawings. It should be noted that, unless there is a conflict, the embodiments and features in the embodiments of this application may be combined with each other.

[0028] The present invention provides an arc detection and anti-false alarm method based on the kurtosis of current spectrum samples, such as Figure 1 As shown, the following steps are included:

[0029] First, an oscilloscope is used to obtain the current spectrum data of the phase difference working condition according to the sampling frequency.

[0030] Analyze the probability density distribution curve of current spectrum data and select the peak frequency segment K according to the spectrum characteristics when the arc occurs. Span and peak frequency band K Arc .

[0031] According to the selected peak frequency segment K Span The spectral characteristics when the arc occurs are used to set the kurtosis judgment threshold G Arc .

[0032] Sampling PV current data in real time, selecting peak frequency segment K Span All sample points are obtained, the sampling point sequence number K corresponding to the peak point is obtained, and the sample kurtosis G is calculated.

[0033] When G ≥ G Arc And K∈K Arc When , it is determined that an arc has occurred.

[0034] Specific implementation instructions:

[0035] Based on the arc data collected from multiple arcing experiments and analysis of its spectrum characteristics, current transformers and AD sampling chips are generally used for implementation. Any solution that can achieve current spectrum data acquisition is within the scope of protection of this case.

[0036] More specifically, the PV current sampling frequency is 256kHz, and the FFT (Fast Fourier Transform) operation points are 1024. This means the signal spectrum range is 128kHz, with a resolution of 125Hz. The selected arc identification frequency range is 2kHz to 15kHz, and the judgment is based on the kurtosis of the 112 sample data points and the maximum frequency point range. Of course, the sample data can be flexibly adjusted based on frequency, operation points, and other factors.

[0037] Then analyze the probability density distribution curve of the current spectrum data, and select the peak frequency segment K according to the spectrum characteristics when the arc occurs. Span and peak frequency band K Arc .

[0038] pass Figures 2 to 4 Spectrum waveform diagram, generally choose K Span 2KHz~15KHz, K Arc =(3KHz, 6KHz), of course, this selection is only for parameter determination in this embodiment, and it can also be adjusted according to the actual data situation. Span and peak frequency band K Arc choice.

[0039] The main use is made of the spectrum characteristics when the arc occurs. It can be observed that a peak with a large kurtosis will be generated in the frequency range of 2K~15KHz, and the peak point is at 3KHz~6KHz. Different from the different background noise of the machine under different working conditions or the spectrum amplitude protrusion caused by the power tracking process, although there is a significant amplitude change in the arc characteristic frequency band, the kurtosis and peak point are obviously different from the protrusion caused by the arc. Based on this phenomenon, false alarms can be effectively reduced.

[0040] Next, according to the selected peak frequency segment K Span The spectral characteristics when the arc occurs are used to set the kurtosis judgment threshold G Arc .

[0041] Among them, G Arc The kurtosis is greater than that when no arc is generated. In addition, when setting it, it is also necessary to consider whether the set kurtosis threshold can be reached when the arc is generated. This threshold is generally set artificially in the early stage.

[0042] During the actual detection, the PV current data is sampled in real time, and the peak frequency segment K is selected. Span All sample points are obtained, the sampling point sequence number K corresponding to the peak point is obtained, and the sample kurtosis G is calculated.

[0043] When G ≥ G Arc And K∈K Arc When the condition is not met, it is determined that an arc has occurred. If this condition is not met, the real-time sampling of PV current data can be continued.

[0044] This allows arc signals to be identified and avoids false alarms when no arc has occurred.

[0045] In a specific embodiment, all sample points include x1 to xn, Where M4 represents the fourth-order sample central moment, M2 represents the sample variance, and M2 2 represents the square of the sample variance;

[0046] Where n is the number of samples, x i is the sample point amplitude, is the mean of all sample points;

[0047] Where n is the number of samples, x i is the sample point amplitude, is the mean of all sample points.

[0048] This achieves the calculation of sample kurtosis G.

[0049] In a specific embodiment, G is performed according to the test results. Arc adjust, K is the peak frequency band when the arc is triggered Span The corresponding sample kurtosis G is the ensemble mean, δ is the adjustment coefficient, and δ ranges from 0.72 to 0.9.

[0050] That is, in the actual application process, as the number of samples and triggers accumulate, correlation adjustment can further reduce the false alarm rate, and by adjusting the mean and coefficient of the trigger samples, G Arc Dynamic updates.

[0051] In a specific embodiment, an oscilloscope is used to obtain arcing tests of differentiated direct current under PV low-voltage conditions, rated voltage conditions, and high-voltage conditions to obtain current spectrum data.

[0052] Using an oscilloscope, we conducted arcing experiments with DC currents of varying magnitudes (2A to 16A) under PV low voltage (100V), rated voltage (360V), and high voltage (450V). The interval current was 2A. A total of 24 sets of arc samples were collected under different PV configurations. The data probability density distribution curve was analyzed, and the kurtosis of each of the 24 sample arrays and the frequency corresponding to the maximum peak point were calculated. The above data represents a specific application experiment.

[0053] Summarizing the 24 sets of sample data, we found that the maximum peak of the arc spectrum is between 3kHz and 6kHz. The DC arc's impact on the PV current frequency band is primarily concentrated between 2kHz and 15kHz. Based on the consistency of the arc behavior, the peak values of the 24 sets of arc data between 2kHz and 15kHz were calculated, with kurtosis ranging from 3.19 to 4.67.

[0054] Based on the arc's spectrum performance, we selected a suitable computing chip and sampling circuit to complete the development board design, and configured the system's sampling frequency to ensure that all frequency points of the arc's characteristics can be observed.

[0055] By real-time sampling of the inverter-mounted PV current spectrum data, monitoring (30s / time) is performed to summarize the frequency corresponding to the maximum peak point without arcing in different configurations and the peak value in the range of 2 to 15 kHz.

[0056] The frequency corresponding to the maximum peak point is relatively random, but the peak value of 2K~15KHz is between 0.26 and 0.96, which is much smaller than the kurtosis when the arc occurs. That is, it can be determined that there is no intersection between the kurtosis when the arc is not generated and the kurtosis when the arc exists. The kurtosis can be detected in real time by PV current kurtosis and the kurtosis can be calculated to avoid the problem of arc false alarm.

[0057] According to the arc characteristic frequency band, the 2K~15KHz kurtosis threshold GArc is set to 2.5, and the peak point frequency range KArc = (3K, 6K) is selected as the basis for determining whether an arc has occurred. When the arc detection data meets the 2K~15KHz peak value greater than the set threshold GArc, and the maximum peak point falls in the interval KArc, it is determined that an arc has occurred.

[0058] The data in this embodiment is only a set of application data, which is used to illustrate the arc determination in this case, and is not intended to limit the threshold value of this case.

[0059] From the above description, we can see that it can effectively avoid false alarms caused by noise from photovoltaic inverters in different working environments. The false alarm rate is effectively reduced from the traditional false alarm rate of over 30% to less than 5%.

[0060] The term "comprise" or any other similar term is intended to cover non-exclusive inclusion such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0061] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

Claims

1. Arc detection and false alarm prevention method based on current spectrum sample kurtosis, characterized by The steps include: S1 uses an oscilloscope to obtain current spectrum data of phase difference working conditions according to the sampling frequency; S2 analyzes the probability density distribution curve of the current spectrum data and selects the peak frequency segment K according to the spectrum characteristics when the arc occurs. Span and peak frequency band K Arc ; S3 selects the peak frequency segment K Span The spectral characteristics when the arc occurs are used to set the kurtosis judgment threshold G Arc ; S4 samples the PV current data in real time and selects the peak frequency segment K Span All sample points, obtain the sampling point sequence number K corresponding to the peak point, and calculate the sample kurtosis G; S5 when G≥G Arc And K∈K Arc When , it is determined that an arc has occurred.

2. The arc detection and false alarm prevention method based on current spectrum sample kurtosis according to claim 1, characterized in that: In step S4, all sample points include x1 to xn. Where M4 represents the fourth-order sample central moment, M2 represents the sample variance, and M2 2 represents the square of the sample variance; Where n is the number of samples, x i is the sample point amplitude, is the mean of all sample points; Where n is the number of samples, x i is the sample point amplitude, is the mean of all sample points.

3. The arc detection and false alarm prevention method based on current spectrum sample kurtosis according to claim 1, characterized in that: In step S3, G Arc Greater than the kurtosis when no arc is generated.

4. The arc detection and false alarm prevention method based on current spectrum sample kurtosis according to claim 3, characterized in that: In step S3, G is performed according to the test results. Arc adjust, K is the peak frequency band when the arc is triggered Span The corresponding sample kurtosis G is the ensemble mean, δ is the adjustment coefficient, and δ ranges from 0.72 to 0.

9.

5. The arc detection and false alarm prevention method based on current spectrum sample kurtosis according to claim 1, characterized in that: In step S2, K Span 2KHz~15KHz, K Arc =(3KHz, 6KHz).

6. The arc detection and false alarm prevention method based on current spectrum sample kurtosis according to claim 1, characterized in that: In step S1, an oscilloscope is used to obtain differentiated direct current arc tests under PV low-voltage conditions, rated voltage conditions, and high-voltage conditions to obtain current spectrum data.

Images