Photovoltaic DC system fault detection method and device based on line spectrum suppression

Abstract

The invention discloses a photovoltaic direct current system fault detection method and device based on line spectrum suppression, and the method comprises the steps: extracting a signal peak frequency through a time-frequency spectrum, analyzing the time-frequency stability of the signal peak frequency, and recognizing a periodic line spectrum signal. The method comprises the following steps of: dividing a frequency band by combining envelope on a frequency spectrum and minimum value point detection, decomposing a signal by using empirical wavelet transform, eliminating a modal component containing a line spectrum signal, only retaining a non-stationary modal component related to a fault feature, and reconstructing the non-stationary modal component to obtain a fault feature signal. And extracting statistical features such as root mean square value, variance and kurtosis of the fault feature signal, and comparing the statistical features with a feature threshold to perform fault arc detection. A photovoltaic system direct current arc detection device is designed based on the algorithm, the device effectively solves the problem that periodic signals and ripple noise in a photovoltaic direct current system interfere fault features under the condition that the low-voltage operation fault features are not obvious, and the universality of photovoltaic direct current system fault recognition is improved.

Description

Technical Field

[0001] The present application relates to the technical field of photovoltaic DC system fault diagnosis, and in particular to a photovoltaic DC system fault detection method and device based on line spectrum suppression. Background Art

[0002] A DC system fault arc is a non-stationary electrical discharge phenomenon caused by poor contact, line aging or insulation damage. It manifests as violent fluctuations in current and voltage, accompanied by high-frequency nonlinear signal characteristics, and its frequency and amplitude change randomly over time. This type of fault arc is usually superimposed on the normal operating signal, mixed with periodic signals and ripple noise, easily masked, and has suddenness and instability. If the fault arc is not detected in time, it may cause power loss, equipment overheating, component burning, and even cause serious fire accidents, directly threatening the operating safety and equipment life of the photovoltaic system.

[0003] Fault arc detection mainly relies on the time domain, frequency domain or time-frequency domain characteristics of the signal for analysis. In the time domain analysis, the voltage and current waveform changes on the line when the arc occurs are used for detection. The arc current change rate under a specific time window is used as a characteristic quantity to make a certain degree of judgment on the arc. However, since the photovoltaic system is greatly affected by environmental factors (such as light intensity), the current changes significantly, and the current changes caused by different arc burning conditions are quite different, which makes the method based on the time domain characteristic quantity susceptible to external interference, high misjudgment rate, insufficient anti-interference ability, and difficult to achieve accurate fault arc detection. In the frequency domain analysis, the change in the harmonic content of the high frequency part of the signal before and after the arc occurs is used as the detection basis to avoid the problem that the time domain method is susceptible to interference. The frequency domain detection based on Fourier transform has high accuracy and is not affected by the arc type, but it lacks time domain information, is complex in calculation, and has high requirements for waveform stability. When a busbar series arc occurs in the photovoltaic system, the low-frequency part of the current signal will be significantly reduced. This phenomenon is greatly affected by the arc position. At present, there is an urgent need for a detection method that can accurately retain the non-stationary characteristics of the fault arc, so as to highlight the fault signal characteristics under complex working conditions and improve the accuracy and reliability of fault diagnosis. Summary of the invention

[0004] The purpose of this application is to provide a photovoltaic DC system fault detection method and device based on line spectrum suppression, which suppresses periodic line spectrum signals and noise interference, accurately extracts the non-stationary characteristics of the fault arc, realizes reliable fault detection of the photovoltaic DC system under complex working conditions, and improves the accuracy and robustness of fault diagnosis.

[0005] To achieve the above object, the present invention adopts the following technical solution: a photovoltaic DC system fault detection method based on line spectrum suppression, characterized in that it includes the following steps:

[0006] S1: Initialize the arc fault detection device and use the acquisition module to collect the voltage signal of the photovoltaic DC system from the voltage sampling point in real time. The acquisition frequency is f 0 ;

[0007] S2: Perform fast Fourier transform (FFT) and short-time Fourier transform (STFT) on the signal to generate the single-sided spectrum and time-frequency spectrum Y of the signal 1 , find the local peak frequency in each time frame and form a time-frequency spectrum matrix Y containing only the peak frequency 4 ;

[0008] S3: Count the number of times each frequency point appears in different time frames as frequency stability, select stable frequency points as starting frequency points, and use the frequency offset between time frames and the offset between the path point and the starting frequency as the total deviation constraint to perform dynamic programming to generate a path. This path is the path where the main frequency changes with the time frame;

[0009] S4: Line spectrum recognition is performed using path offset length and node energy entropy as threshold indicators. The signal is subjected to Fourier transform (FFT) and the minimum value point of the envelope on the spectrum is extracted. The intervals are merged according to the extracted line spectrum frequency information to obtain the frequency band division boundary.

[0010] S5: Perform empirical wavelet transform (EWT) on the signal according to the divided frequency bands to obtain multiple modal components (IMFs), and reconstruct the modal components without line spectrum frequencies to obtain the fault characteristic signal;

[0011] S6: Extract the threshold values ​​of the reconstructed signal RMS value, variance, kurtosis, peak factor, pulse factor, waveform factor and other indicators to perform fault arc detection and determine whether an arc is generated.

[0012] Further, the step S1 is specifically as follows: initializing the arc fault detection device, using the ADC acquisition module to obtain the voltage signal from the voltage sampling point of the photovoltaic DC system in real time, and the sampling frequency f 0 At least twice the maximum frequency of the effective fault arc characteristic frequency band. If the sampling hardware allows, selecting a higher sampling frequency can more accurately capture the high-frequency characteristics and non-stationarity of the fault arc signal. 0 Set it to 100kHz~1MHz, and the sampling time is T. .

[0013] Further, the step S2 is specifically as follows: in an optional embodiment S2 of the present application, a fast Fourier transform (FFT) is performed on the collected signal x(t) to generate a single-sided spectrum X(f).

[0014] The short-time Fourier transform (STFT) of the acquired signal spectrum is used to weight the signal segment by segment using the Hanning window, and the window length is set to N win, the length of the window overlap is set to N overlap , the number of fast Fourier transform points is N FFT .

[0015] After applying the window function to each signal segment, a fast Fourier transform (FFT) is performed to obtain the spectrum of the time frame. The spectrum of all time frames is stacked to obtain the time-frequency spectrum matrix Y1.

[0016] In each time frame of Y1, the local peak frequency in the frequency dimension is detected point by point, and the judgment condition of the local maximum frequency point is:

[0017]

[0018] Where f is the current frequency point, is the spectrum resolution, t is the time frame index, N t is the total number of time frames.

[0019] The local peak frequency points are retained and the rest are set to zero to obtain the peak frequency matrix Y4.

[0020] Furthermore, the step S3 is specifically as follows: for the peak frequency matrix Y4 generated in S2, traverse all time frames, count the number of occurrences of non-zero values ​​of each frequency point in all time frames, and use it as the frequency stability N(f) of the frequency point.

[0021] In order to reduce the amount of calculation, a preliminary screening is performed to filter out non-stationary frequency points. The preliminary screening process is as follows:

[0022]

[0023] Wherein, Td is the time proportion threshold, and S'(f) is the initial frequency point set.

[0024] Taking the frequency point of S'(f) as the starting frequency point, dynamic programming is used to find the line spectrum stable path in the peak frequency matrix Y4. The total deviation constraint Cost(f t )as follows:

[0025] Cost(f t )=α·|f t -f t-1 |+β·|f t -f start |

[0026] In the formula, f start is the starting frequency point of the initial frequency point set, |f t -f t-1 | is the frequency offset cost between time frames, |f t -f start| is the global offset cost between the current path frequency point and the starting frequency point, and α and β are cost weight parameters.

[0027] The cumulative minimum cost value C(f,t) of each frequency point is calculated in each time frame. The recursive formula is as follows:

[0028] C(f,t)=min(C(f t-1 ,t-1)+Cost(f t ))

[0029] Among them, C(f,t) is the cumulative minimum cost value of the current time frame t, C(f t-1 ,t-1) is the cumulative minimum cost value of the previous time frame t-1, Cost(f t ) is the total deviation constraint for the current time frame.

[0030] Get the frequency point with the minimum cost value in the last time frame Get the main frequency path

[0031] Further, the step S4 is specifically as follows: the line spectrum signal has frequency stability and period stability in the process of changing with time, and the path offset length L is selected path =C(f,t) and node energy entropy H are used for line spectrum recognition. The node energy entropy calculation formula is as follows:

[0032]

[0033] Among them, E i is the energy value of the node on the path P(t,f).

[0034] The path offset length threshold L 0 and node energy entropy threshold H 0 Determine the stable line spectrum frequency points in the signal, and determine the distribution interval of the line spectrum frequency with the line spectrum frequency points as a reference.

[0035] The single-sided spectrum X(f) obtained in step S2 is subjected to sliding window analysis to extract the upper envelope to obtain X upper (f) Detect the envelope minimum point to obtain the initial frequency band division boundary.

[0036] The adjacent frequency bands containing line spectrum frequency points are merged into one frequency band, and the final frequency band division boundary B = {[f j ]|j=1,2,...}.

[0037] Further, the step S5 is specifically as follows: performing EWT decomposition according to the frequency band division boundary B, for each frequency band interval [f i ,f i+1 ], construct a bandpass filter Wi (f), perform modal decomposition on the acquired signal x(t):

[0038]

[0039] In the formula, and Represent the Fourier transform and inverse transform of the signal respectively. Each IMF i (t) corresponds to the characteristics of the signal in the i-th frequency band.

[0040] Reconstruct the frequency band interval that does not contain the line spectrum frequency to obtain the fault characteristic signal x r (t):

[0041]

[0042] Further, the step S6 is specifically as follows: extracting the reconstructed signal x r The statistical characteristics of (t) are analyzed and the occurrence of arc fault is determined by setting the characteristic value threshold. The selection of statistical characteristics should be based on the following perspectives:

[0043] Energy change: reflects the increase in energy level and fluctuation of fault signals;

[0044] Waveform characteristics: Capture the characteristics of instantaneous spikes or complex waveforms in fault signals;

[0045] Transient characteristics: short-term and drastic changes caused by fault arc.

[0046] The beneficial effects of the present invention are:

[0047] (1) This application uses frequency stability and energy stability analysis combined with a dynamic programming algorithm to accurately identify and eliminate the stable line spectrum frequency interference in the photovoltaic DC system, avoid the line spectrum frequency from masking the fault characteristic signal, and can display the fault characteristics more clearly.

[0048] (2) The present application can effectively separate normal signals and fault signals under complex working conditions such as low-voltage operation of the photovoltaic system and inverter startup, retain the characteristics of inverter startup and arc fault, accurately judge the occurrence of arc fault, reduce misjudgment and missed judgment, and has strong robustness and anti-interference capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0050] Figure 1 A photovoltaic DC system fault detection device based on line spectrum suppression is provided in an embodiment of the present application;

[0051] Figure 2 A flow chart of suppressing signal line spectrum frequency provided in an embodiment of the present application. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0053] Figure 1 The present application provides a method and device for photovoltaic DC system fault detection based on line spectrum suppression, which includes: a signal acquisition module S101, a signal processing unit S102, a fault judgment module S103 and a display alarm device S104.

[0054] The signal acquisition module S101 is distributed at the grid-connected end and the inverter side of the photovoltaic DC system to collect voltage or current signals in real time;

[0055] The signal processing unit S102 receives the collected signal, performs line spectrum interference suppression, frequency band decomposition and signal reconstruction;

[0056] The fault judgment module S103 judges whether there is a fault arc by extracting statistical characteristic values ​​(such as root mean square value, variance, peak factor, etc.) and combining them with preset thresholds;

[0057] The display alarm device S104 is used to display the system operation status and fault information in real time, and trigger an alarm when a fault occurs.

[0058] Furthermore, the signal acquisition module S101 monitors the voltage signal in the photovoltaic DC system in real time, uses a voltage sensor to obtain the signal of the key position of the system (such as the inverter input terminal and the grid-connected line), and converts it into f 0 =100kHz sampling frequency digital processing to ensure the capture of high-frequency fault characteristics.

[0059] Further, the signal processing process of the signal processing unit S102 is as follows: Figure 2 As shown,

[0060] Perform short-time Fourier transform (STFT) on the collected voltage signal, with a window length of N win is 512, the length of the window overlap is Noverlap is 256, the number of fast Fourier transform (FFT) points N FFT is 1024, and the frequency resolution is as follows:

[0061]

[0062] The peak extraction is to compare the amplitudes of adjacent frequency points of the time-frequency spectrum matrix Y1 frame by frame, extract the local peak frequency, and obtain the peak frequency matrix Y4.

[0063] According to the frequency stability N(f), the starting frequency point S'(f) is preliminarily selected, and the offset process of the starting frequency point S'(f) on the time axis is dynamically planned to obtain the main frequency path P(t,f). The total deviation constraint Cost(f t )as follows:

[0064] Cost(f t )=α·|f t -f t-1 |+β·|f t -f start |

[0065] Among them, |f t -f t-1 | is the frequency offset cost between time frames, |f t -f start | is the global offset cost between the current path frequency point and the starting frequency point, α and β are cost weight parameters, both set to 0.5.

[0066] Calculate the path offset length L of the main frequency path P(t,f) path =C(f,t) and node energy entropy H, according to the path offset length threshold L 0 and energy entropy threshold H 0 Identify line spectrum frequencies.

[0067] Among them, the path offset length reflects the frequency stability of the line spectrum frequency, and the node energy entropy reflects the periodicity and energy stability of the frequency component.

[0068] The collected voltage signal is subjected to fast Fourier transform (FFT) to obtain a single-sided spectrum X(f).

[0069] The frequency band division boundary is extracted according to the minimum value point of the envelope on X(f), and the adjacent frequency band intervals containing line spectrum frequency points are merged into one frequency band interval.

[0070] The signal is decomposed into multiple modal components (IMFs) using the empirical wavelet transform (EWT), and the non-linear modal components are superimposed and reconstructed to obtain the characteristic signal x without linear spectrum interference. r (t).

[0071] Furthermore, the fault judgment module S103 extracts and analyzes statistical features of the reconstructed signal to judge the fault arc.

[0072] Statistical features such as the root mean square value, variance, kurtosis, peak factor, pulse factor and waveform factor of the signal are extracted. These features can reflect the signal's energy level, waveform sharpness and instantaneous pulse characteristics.

[0073] The possible arc fault in the system is determined according to the fault characteristic threshold. If it is determined that the circuit has an arc fault, the display alarm device S104 is triggered to alarm. If it is determined that the circuit is normal, the detection continues.

[0074] Furthermore, the display alarm device S104 displays the operating status and fault information of the photovoltaic DC system in real time, and triggers an alarm when a fault arc is detected.

[0075] The display alarm device S104 receives the fault judgment result and related signal characteristic data in real time through the communication interface with the fault judgment module S103, and displays the current system operation status, including system status, related characteristic values, and fault information, on the display screen of the device.

[0076] Finally, it should be noted that the above is only a preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and changes. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A photovoltaic DC system fault detection method based on line spectrum suppression, characterized in that: The method comprises: Initialize the arc fault detection device and collect fault arc signal data through the signal acquisition module; Performing fast Fourier transform (FFT) and short-time Fourier transform (STFT) on the fault arc signal to generate a single-sided spectrum and a time-frequency spectrum of the signal; Detecting local peak frequencies of the time-frequency spectrum and dynamically planning and generating a main frequency path; Performing line spectrum recognition on the main frequency path with a threshold index, and determining the frequency band division boundary according to the line spectrum frequency and the envelope minimum value on the unilateral spectrum; According to the frequency band division boundary, the signal is decomposed by empirical wavelet transform (EWT) frequency band, and the non-linear spectrum mode is reconstructed to obtain the fault characteristic signal; An arc fault is determined according to the statistical characteristic value threshold.

2. The method according to claim 1, characterized in that In the local peak frequency detection, the time-frequency spectrum of the signal is analyzed, and in each time frame, the local peak frequency with an amplitude higher than that of its adjacent frequency points is identified by comparing the amplitude changes in the frequency dimension.

3. The method according to claim 1, characterized in that The dynamic programming is implemented by the following steps: The main frequency path is obtained by dynamic programming in the peak frequency matrix using the starting frequency point. The total deviation constraint of dynamic programming is as follows: Cost(f t )=α·|f t -f t-1 |+β·|f t -f start | Among them, f start is the initial frequency point, f t is the current path frequency point, α and β are cost weight parameters. The obtained main frequency path is subjected to line spectrum recognition using a threshold indicator.

4. The method according to claim 3, characterized in that The threshold indicators are path offset length and node energy entropy. The path offset length is the total deviation constraint sum of the main frequency path, indicating that the line spectrum has frequency stability in the process of changing over time. The node energy entropy evaluates the energy concentration of each frequency point in a specific frequency path, and the implementation process is as follows: Among them, E i is the energy value of the node on the main frequency path.

5. The method according to claim 1, characterized in that The implementation steps of the empirical wavelet transform (EWT) are as follows: According to the upper envelope minimum of the single-sided spectrum as the initial boundary, the adjacent frequency bands containing line spectrum frequency points are merged to obtain the frequency band division boundary. For each frequency band, construct a bandpass filter W i (f) Perform modal decomposition on the acquired signal: in, and Represents the Fourier transform and inverse transform of the signal, IMF i (t) is the decomposed modal component.

6. The method according to claim 1, characterized in that The statistical characteristics include energy variation, waveform characteristics and transient characteristics, which are used to reflect the energy level and fluctuation, waveform sharpness and transient pulse characteristics of the fault signal.

7. A photovoltaic DC system fault detection device based on line spectrum suppression, characterized in that: The device includes: A signal acquisition module, used for collecting voltage signals from a photovoltaic DC system; A signal processing unit for performing line spectrum interference suppression, frequency band decomposition and signal reconstruction; A fault judgment module is used to extract the statistical characteristic value of the reconstructed signal and judge whether there is an arc fault; Display alarm equipment is used to display the system operation status and fault information in real time, and trigger an alarm when a fault is detected.

Images