Full peer-to-peer power distribution system line fault detection method and system based on transient power

Abstract

The invention discloses a full-peer power distribution system line fault detection method and system based on transient power, and belongs to the technical field of power grid fault detection, and the method comprises the steps: S1, collecting fault voltage and current data of a local side and an opposite side in a power distribution network line, and calculating the difference values of the voltage fault components of the local side and the opposite side to time; if the difference value of the voltage fault components of the local side and the opposite side to time is greater than a set threshold value, entering S2; s2, reconstructing the fault signal according to dual-tree complex wavelet transform; s3, screening the reconstructed signals according to a multi-frequency-point selection principle; and S4, carrying out Hilbert transformation on voltage data corresponding to the screened signals, and determining a fault position, the high-frequency transient reactive power polarity is extracted through a time-frequency algorithm, and various types of faults inside and outside the line area can be rapidly and accurately identified.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid fault detection, and in particular to a method and system for detecting line faults in a fully symmetrical power distribution system based on transient power. Background Art

[0002] When discussing the construction of a fully symmetrical distribution network, the core is to introduce flexible soft switches as key nodes to achieve flexible configuration of three-side or multi-side power supply. This network architecture is based on a honeycomb connection method, which realizes the deep integration of distribution network and microgrid, and supports the efficient mode of closed-loop operation. It is worth noting that the layout of the honeycomb distribution network is not limited to a strict regular hexagonal geographical structure, and can also adapt to a variety of different shapes and configurations such as quadrilaterals and pentagons. However, the large-scale access of distributed power sources to the fully symmetrical distribution network has made the fully symmetrical distribution network highly power-electronic, and also caused great changes in the distribution topology and operation control. The network morphology architecture of the fully symmetrical distribution network is different from that of the traditional distribution network, and its fault characteristics have changed significantly, mainly reflected in the change of the source-load structure, which causes the randomization of fault characteristics and the fuzzification of fault paths, making it difficult for traditional relay protection based on power frequency to adapt. Therefore, it is of great significance to study the protection scheme of the fully symmetrical distribution network adapted to the access of distributed resources to ensure its safe operation.

[0003] At present, the types of distribution network detection strategies are mainly divided into two categories: single-end quantity detection and double-end quantity detection. In terms of single-end quantity detection, the existing research on fault detection of distribution lines containing distributed power sources mainly focuses on the improvement of traditional fault detection principles based on photovoltaic fault current output characteristics. However, the topology of the distribution network is complex, the operation and control methods are diverse, and fault detection methods in different sections are difficult to adapt. Compared with the fault detection method based on single-end quantity, the fault detection method using electrical quantities at both ends of the line has stronger applicability in the new distribution network with multiple sources. In order to eliminate the influence of power electronic control, the frequency domain electrical quantity information in the transient natural response stage can be used to construct appropriate detection criteria to improve the adaptability of the double-end quantity fault detection method. Therefore, it is necessary to study the fast detection method of distribution network line faults based on double-end communication, and fully consider the influence of different fault types and fault conditions, noise interference and communication delay on fault feature extraction, so as to achieve faster and more accurate fault detection capabilities of distribution network lines. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present invention provides a method and system for detecting line faults in a fully symmetrical power distribution system based on transient power, so as to solve the defects of the existing fault detection method based on two-terminal quantities, such as poor ability to resist transient resistance, difficulty in extracting fault features or susceptibility to interference, and high requirements for information synchronization at both ends.

[0005] In order to achieve the above-mentioned invention object, the technical solution adopted by the present invention is: a line fault detection method of a full-equivalent distribution system based on transient power, comprising:

[0006] S1, collecting fault voltage and current data on the local side and the opposite side of the distribution network line, and calculating the differential value of the voltage fault component on the local side and the opposite side over time respectively;

[0007] In response to the time difference between the voltage fault component on the local side and the opposite side being greater than a set threshold, entering S2;

[0008] S2, reconstructing the fault signal according to the dual-tree complex wavelet transform;

[0009] S3, screening the reconstructed signal according to the multi-frequency point selection principle;

[0010] S4. Perform Hilbert transform on the voltage data corresponding to the filtered signal and determine the fault location.

[0011] Further: S2 includes:

[0012] S21. According to the dual-tree complex wavelet transform, the real and imaginary parts of the fault signal are simultaneously decomposed by wavelet to obtain the wavelet coefficient d T j (n) and the scale factor c T J (n);

[0013] S22, for wavelet coefficient d T j (n) and the scale factor c T J (n) Perform inverse transformation to extract the signal components of each original frequency band of the fault signal and complete the reconstruction of the fault signal.

[0014] Further: In S21, the expression for wavelet decomposition is:

[0015]

[0016]

[0017] in, are d Re j d Im j are the wavelet coefficients of the real part tree transform and the wavelet coefficients of the imaginary part tree transform, d Re j d Im j are the wavelet coefficients of the real part tree transform and the wavelet coefficients of the imaginary part tree transform, c Rej 、c Im j are the scale coefficients of the real part tree transform and the imaginary part tree transform respectively, j is the scale, J is the number of signal frequency divisions, and n is the translation amount.

[0018] Further: In S22, the reconstructed signal The expression is:

[0019]

[0020] Among them, ψ h , g are two real wavelets respectively, and k is the imaginary integral coefficient.

[0021] Further: In S3, the screening principle is:

[0022]

[0023] Where, C is the filter capacitor; L1 is the filter inverter inductor; C DC is the DC side capacitance, f is the target frequency for screening, f L is the lower frequency limit, f S is the sampling frequency, f R is the first resonant frequency, f PWM is the PWM carrier frequency, f H is 3000Hz, and min(.) means taking the minimum value.

[0024] Further: S4 includes:

[0025] S41, performing Hilbert transform on the voltage data corresponding to the filtered signal to calculate the instantaneous reactive power of the fault component;

[0026] S42. Determine the fault location according to the instantaneous reactive power of the fault components on the local side and the opposite side.

[0027] Further: In the S41, the expression for calculating the instantaneous reactive power q(t) of the fault components on both sides is:

[0028]

[0029] in, represents the voltage fault component u at the detection point at time t fi Hilbert transform of (t); i fi (t) represents the current fault component of the detection point at time t, and P is the main value of the integration.

[0030] Further: the S42 includes:

[0031] S421. Construct a distribution network line fault detection identifier λ by using the instantaneous reactive power of the fault component on the local side and the opposite side, and its expression is:

[0032]

[0033] Among them, t s Indicates that the protection time window is long, q M (τ) represents the instantaneous reactive power of the fault component at the detection point M at time τ, q N (τ) represents the instantaneous reactive power of the fault component at the detection point N at time τ;

[0034] S422. Determine the fault location according to the value of the distribution network line fault detection identifier λ:

[0035] When the value of λ is greater than 0, it is determined to be an internal fault of the distribution line;

[0036] When the value of λ is less than 0, it is determined to be an external fault of the distribution line.

[0037] Further: it also includes a lightning interference detection method: by reconstructing the high frequency segment decomposition modulus maximum value W of the current component in the signal HS I M , W HS I N The maximum value of the modulus of the decomposition quantity in the low frequency band W LS I M , W LS I N The ratio of is used to identify lightning interference.

[0038] The present invention also provides a distribution network line fault detection system based on transient power for implementing a full-equivalent distribution system line fault detection method based on transient power, comprising:

[0039] Voltage acquisition and calculation module: used to collect voltage data on the local and opposite sides of the distribution network line, and calculate the differential value of the voltage fault component on the local and opposite sides over time respectively; when the differential value of the voltage fault component over time is greater than the set threshold, the fault detection module is started on the local and opposite sides of the line respectively;

[0040] Current acquisition and calculation module: used to collect current data on the local and opposite sides of the distribution network line, and used for time-frequency analysis of high-frequency components;

[0041] Fault detection module: It is used to perform preprocessing using dual-tree complex wavelet transform to extract high-frequency voltage and current in a specific frequency band, and to quantify the difference in power transfer direction through the instantaneous reactive power values ​​of the high-frequency voltage and current at both ends based on the Hilbert transform algorithm to construct a fault detection identifier;

[0042] It is also used to screen the maximum high-frequency modulus and the maximum low-frequency modulus of the current to prevent lightning interference.

[0043] The beneficial effects of the present invention are:

[0044] 1. The present invention uses the transient reactive power product of high-frequency fault components to construct a detection criterion, which can quickly and accurately identify various types of faults inside and outside the line area;

[0045] 2. The present invention designs a fault detection start-up criterion and a data preprocessing method, and realizes a complete distribution network line fault detection method. The method has engineering and practical value for fault identification and detection of distribution networks;

[0046] 3. The present invention calculates and sets the fault initiation criteria and detection criteria modules in combination with the actual distribution network data, which can be directly loaded into the distribution network fault detection system. By using the collected fault data information, the occurrence of line faults can be accurately judged, and it has the characteristics of fast recognition speed, high accuracy, strong tolerance to noise and synchronization errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 The present invention is a flow chart of a line fault detection method for a full-equivalent power distribution system based on transient power;

[0048] Figure 2 A topological diagram of a distribution network in an embodiment;

[0049] Figure 3 A step diagram of a fault detection method according to an embodiment;

[0050] Figure 4 A schematic diagram of starting criterion values ​​for different fault locations in the area of ​​the embodiment;

[0051] Figure 5 This is a DTCWT signal reconstruction capability effect diagram of the embodiment;

[0052] Figure 6 1 is a diagram showing the withstand transition resistance under different types of faults of the embodiment;

[0053] Figure 7 This is a diagram showing the synchronization error tolerance under different types of faults of the embodiment. DETAILED DESCRIPTION

[0054] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0055] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0056] Embodiment 1:

[0057] like Figure 1 As shown, this embodiment provides a method for detecting line faults in a full-equivalent power distribution system based on transient power, including:

[0058] S1, collecting fault voltage and current data on the local side and the opposite side of the distribution network line, and calculating the differential value of the voltage fault component on the local side and the opposite side over time respectively;

[0059] In response to the time difference between the voltage fault component on the local side and the opposite side being greater than a set threshold, entering S2;

[0060] S2, reconstructing the fault signal according to the dual-tree complex wavelet transform;

[0061] S3, screening the reconstructed signal according to the multi-frequency point selection principle;

[0062] S4. Perform Hilbert transform on the voltage data corresponding to the filtered signal and determine the fault location.

[0063] Specifically, S1 includes:

[0064] S11, collecting voltage and current data on the distribution network line side and the opposite side;

[0065] The voltage and current data are the absolute values ​​of the voltage and current measured at the line port;

[0066] S12. Calculate the differential value of the voltage fault component over time. The formula is as follows:

[0067]

[0068] Where ΔU is the instantaneous value of the fault component voltage on either side, ΔU set is the threshold value for starting the judgment;

[0069] S13: setting a threshold value of the time difference value of the line voltage fault component under normal operation of the distribution network;

[0070] In this embodiment, the start criterion threshold is set according to the maximum value of the voltage change when the ground is connected to the ground through a 200Ω transition resistor outside the avoidance zone;

[0071] S14: Determine whether the time difference value of the voltage fault component collected on the local side and the opposite side exceeds the threshold value. If so, start the fault detection process and enter S2. Otherwise, do not start.

[0072] Specifically, S2 includes:

[0073] In S21, the expression for wavelet decomposition is:

[0074]

[0075] in, are d Re j d Im j are the wavelet coefficients of the real part tree transform and the wavelet coefficients of the imaginary part tree transform, d Re j d Im j are the wavelet coefficients of the real part tree transform and the wavelet coefficients of the imaginary part tree transform, c Re j 、c Im j are the scale coefficients of the real part tree transform and the imaginary part tree transform, respectively, j is the scale, J is the number of signal frequency divisions, and n is the translation amount;

[0076] S22, for wavelet coefficient d T j (n) and the scale factor c T J (n) Perform inverse transformation to extract the signal components of each original frequency band of the fault signal, and complete the reconstruction of the fault signal. The reconstructed signal The expression is:

[0077]

[0078] Among them, ψ h , g are two real wavelets respectively, and k is the imaginary integral coefficient.

[0079] Specifically, the voltage and current frequency band used as the detection criterion of S3 should meet the following selection principles:

[0080] The upper limit of the frequency needs to consider the following factors: According to the sampling theorem, the upper limit of the frequency should be less than the sampling frequency f S Half of the first resonant frequency f; In order to ensure that the circuit impedance characteristics are highly inductive during a fault, the selected frequency should be lower than the first resonant frequency f R ; To avoid the impact of high-frequency components of photovoltaic power output on detection, the selected frequency should be less than the PWM carrier frequency f PWM Considering that the signal attenuation is serious and the signal-to-noise ratio is low above 3000Hz, the selected frequency should not be higher than 3000Hz and is set to f H , so the screening principle of S3 can be expressed as:

[0081]

[0082] Where, C is the filter capacitor; L1 is the filter inverter inductor; C DC is the DC side capacitance;

[0083] Specifically, the specific process of S4 includes:

[0084] S41, performing Hilbert transform on the voltage data corresponding to the filtered signal to calculate the instantaneous reactive power of the fault component;

[0085]

[0086] in, represents the voltage fault component u at the detection point at time t fi Hilbert transform of (t); i fi (t) represents the current fault component of the detection point at time t, and P is the main value of the integral

[0087] S42, determining the fault location according to the instantaneous reactive power of the fault components on the local side and the opposite side;

[0088] S42 specifically includes:

[0089] S421. Construct a distribution network line fault detection identifier λ by using the instantaneous reactive power of the fault component on the local side and the opposite side, and its expression is:

[0090]

[0091] Among them, t s Indicates that the protection time window is long, q M (τ) represents the instantaneous reactive power of the fault component at the detection point M at time τ, q N (τ) represents the instantaneous reactive power of the fault component at the detection point N at time τ;

[0092] S422. Determine the fault location according to the value of the distribution network line fault detection identifier λ:

[0093] When the value of λ is greater than 0, it is determined to be an internal fault of the distribution line;

[0094] When the value of λ is less than 0, it is determined to be an external fault of the distribution line;

[0095] The frequency band selected based on theoretical analysis will make the equivalent network impedance of the distribution system appear inductive, so it will generally not be 0.

[0096] Preferably, this embodiment also provides a lightning interference detection method: by determining the maximum value W of the high frequency decomposition quantity modulus of the current component in the reconstructed signal HS IM , W HS I N The maximum value of the modulus of the decomposition quantity in the low frequency band W LS I M , W LS I N The ratio of can be used to identify lightning interference. The specific expression is:

[0097] α=α M +α N >α set

[0098]

[0099] Among them, α M is the current mode maximum ratio at the monitoring point M, α N is the current mode maximum ratio at the monitoring point N, α set is the lightning interference threshold, and α is the sum of the maximum values ​​of the current modulus at both ends.

[0100] Embodiment 2:

[0101] In order to enable those skilled in the art to better understand the principles of the present invention, the following is described in conjunction with the accompanying drawings and embodiments:

[0102] In this embodiment, Figure 2 The distribution network shown is an example, and the method proposed in the present invention is used for calculation and analysis. Figure 2 The distribution network system parameters shown are shown in Table 1:

[0103] Table 1 Distribution network system simulation parameters

[0104]

[0105]

[0106] The fault points are set to f1, f2 and f3 respectively, the transition resistance setting range is 50-150 ohms, the sampling frequency is 20kHz, and the sampling time window is 10ms. The method proposed by the present invention is tested and verified through a large amount of simulation data. The specific flow chart is as follows Figure 3 shown.

[0107] S1, collecting fault voltage and current data on the local side and the opposite side of the distribution network line, and calculating the differential value of the voltage fault component on the local side and the opposite side over time respectively;

[0108] In response to the time difference between the voltage fault component on the local side and the opposite side being greater than a set threshold, entering S2;

[0109] In this embodiment, three-phase short circuit fault, two-phase short circuit fault and single-phase ground fault are set at different positions of line MN. The fault positions are respectively set at 0.1km, 2km, 4km, 6km, 8km and 9.9km away from the detection element. The voltage fault component and the time difference value are as follows: Figure 4 As shown. Figure 4 It can be seen that when an intra-area fault occurs, the time difference value of the voltage fault component exceeds the threshold, the fault detection process is started, and a start signal is sent to the fault detection element on the opposite side of the line to determine whether the fault occurs on the line between the two ports in the subsequent process. When the fault detection element on this side receives the start signal from the opposite side, the fault detection process is also started.

[0110] S2, reconstructing the fault signal according to the dual-tree complex wavelet transform;

[0111] Specifically, S2 includes:

[0112] S21. According to the dual-tree complex wavelet transform, the real and imaginary parts of the fault signal are simultaneously decomposed by wavelet to obtain the wavelet coefficient d T j (n) and the scale factor c T J (n);

[0113]

[0114] in, are d Re j d Im j are the wavelet coefficients of the real part tree transform and the wavelet coefficients of the imaginary part tree transform, d Re j d Im j are the wavelet coefficients of the real part tree transform and the wavelet coefficients of the imaginary part tree transform, c Re j 、c Im j are the scale coefficients of the real part tree transform and the imaginary part tree transform, respectively, j is the scale, J is the number of signal frequency divisions, and n is the translation amount;

[0115] S22, for wavelet coefficient d T j (n) and the scale factor c T J (n) Perform inverse transformation to extract the signal components of each original frequency band of the fault signal, and complete the reconstruction of the fault signal. The reconstructed signal The expression is:

[0116]

[0117] Among them, ψ h , g are two real wavelets respectively, and k is the imaginary integral coefficient.

[0118] In this embodiment, taking the signal S(t)=sin(2π*(25*t+10*sin(1.5*t))) as an example, adding Gaussian white noise with a signal-to-noise ratio of 15dB, the signal is reconstructed using continuous wavelet transform (CWT) and dual-tree complex wavelet transform. Typical test results are as follows: Figure 5 As shown. Figure 5 It can be seen that the dual-tree wavelet transform algorithm has good signal reconstruction ability and can show strong robustness to Gaussian white noise.

[0119] S3, screening the reconstructed signal according to the multi-frequency point selection principle;

[0120] Specifically, the voltage and current frequency band used as the detection criterion of S3 should meet the following selection principles:

[0121] The upper limit of the frequency needs to consider the following factors: According to the sampling theorem, the upper limit of the frequency should be less than the sampling frequency f S Half of the first resonant frequency f; In order to ensure that the circuit impedance characteristics are highly inductive during a fault, the selected frequency should be lower than the first resonant frequency f R ; To avoid the impact of high-frequency components of photovoltaic power output on detection, the selected frequency should be less than the PWM carrier frequency f PWM Considering that the signal attenuation is serious and the signal-to-noise ratio is low above 3000Hz, the selected frequency should not be higher than 3000Hz and is set to f H , so the screening principle of S3 can be expressed as:

[0122]

[0123] Where, C is the filter capacitor; L1 is the filter inverter inductor; C DC is a DC side capacitor. In this embodiment, the upper limit of the frequency band is selected to be 3000 Hz and the lower limit is selected to be 853 Hz.

[0124] S4. Perform Hilbert transform on the voltage data corresponding to the filtered signal and determine the fault location.

[0125] S41, performing Hilbert transform on the voltage data corresponding to the filtered signal to calculate the instantaneous reactive power of the fault component;

[0126] S42, determining the fault location according to the instantaneous reactive power of the fault components on the local side and the opposite side;

[0127] S42 specifically includes:

[0128] S421. Construct a distribution network line fault detection identifier λ by using the instantaneous reactive power of the fault component on the local side and the opposite side, and its expression is:

[0129]

[0130] Among them, t s Indicates that the protection time window is long, q M (τ) represents the instantaneous reactive power of the fault component at the detection point M at time τ, q N (τ) represents the instantaneous reactive power of the fault component at the detection point N at time τ;

[0131] S422. Determine the fault location according to the value of the distribution network line fault detection identifier λ:

[0132] When the value of λ is greater than 0, it is determined to be an internal fault of the distribution line;

[0133] When the value of λ is less than 0, it is determined to be an external fault of the distribution line.

[0134] In this embodiment, the lightning interference detection method is used as S5. The specific process of S5 is: by using the high-frequency decomposition modulus maximum value W of the current component in the reconstructed signal HS I M , W HS I N The maximum value of the modulus of the decomposition quantity in the low frequency band W LS I M , W LS I N The ratio of can be used to identify lightning interference. The specific expression is:

[0135] α=α M +α N >α set

[0136]

[0137] Among them, α M is the current mode maximum ratio at the monitoring point M, α N is the current mode maximum ratio at the monitoring point N, α set is the lightning interference threshold, which is set to 28 in this embodiment, and α is the sum of the maximum values ​​of the current modulus at both ends.

[0138] In this embodiment, different types of faults are set at different locations on the line MN, and the fault detection criteria are as follows: Figure 4 shown.

[0139] Different types of faults are set at points f1, f2 and f3 of the line, and the start criteria and detection criteria are shown in Table 2:

[0140] Table 2 Protection criteria for different fault types

[0141]

[0142] It can be seen from Table 2 that when a short circuit fault occurs within the zone, the start criterion value exceeds the threshold value, and the detection can be started reliably; the detection criterion value is greater than 0, and the detection result is correct. When an inter-pole short circuit fault occurs outside the zone, the detection criterion value is less than 0, so it is detected as an outside-zone fault.

[0143] Furthermore, in order to fully verify the impact of photovoltaic capacity on the detection scheme and improve the photovoltaic penetration rate of the line, a three-phase short circuit fault is taken as an example for analysis. The simulation results are shown in Table 3.

[0144] Table 3 Protection status under different photovoltaic penetration rates

[0145]

[0146] Furthermore, in order to illustrate the superiority of the method proposed in the present invention, faults with different transition resistances at different fault locations were tested. The results under different fault types are shown in the figure. Figure 6 As shown, the results show that the method proposed in the present invention has a good anti-transition resistance effect.

[0147] In addition, considering the existence of double-end communication synchronization error in the application, the fault detection criteria under different synchronization error times were tested. The results are as follows: Figure 7 As shown, the results show that the method proposed in the present invention has a good effect of tolerating synchronization errors and is less affected by data window delays.

[0148] Finally, in order to analyze the impact of lightning interference, the reliability of the lightning protection criterion under simulated lightning interference was tested, and the simulation results of lightning interference are shown in Table 4. From Table 4, it can be seen that when lightning interference occurs, although the fault detection criterion is greater than 0, the lightning protection detection criterion far exceeds the set value. Therefore, the proposed detection can avoid false detection under lightning interference.

[0149] Table 4 Comparison of simulation results of lightning interference and single-phase grounding fault

[0150]

[0151] Embodiment 3:

[0152] This embodiment provides a distribution network line fault detection system based on transient power, which is used to implement the full-equivalent distribution system line fault detection method based on transient power described in Embodiments 1-2, including:

[0153] Voltage acquisition and calculation module: used to collect voltage data on the local and opposite sides of the distribution network line, and calculate the differential value of the voltage fault component on the local and opposite sides over time respectively; when the differential value of the voltage fault component over time is greater than the set threshold, the fault detection module is started on the local and opposite sides of the line respectively;

[0154] Current acquisition and calculation module: used to collect current data on the local and opposite sides of the distribution network line, and used for time-frequency analysis of high-frequency components;

[0155] Fault detection module: It is used to perform preprocessing using dual-tree complex wavelet transform to extract high-frequency voltage and current in a specific frequency band, and to quantify the difference in power transfer direction through the instantaneous reactive power values ​​of the high-frequency voltage and current at both ends based on the Hilbert transform algorithm to construct a fault detection identifier;

[0156] It is also used to screen the maximum high-frequency modulus and the maximum low-frequency modulus of the current to prevent lightning interference.

[0157] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for detecting line faults in a fully symmetrical power distribution system based on transient power, characterized in that: include: S1. Collect fault voltage and current data on the local and opposite sides of the distribution network line, and calculate the difference value of the voltage fault component on the local and opposite sides over time respectively; In response to the time difference between the voltage fault component on the local side and the opposite side being greater than the set threshold, entering S2; S2. Reconstruct the fault signal according to the dual-tree complex wavelet transform; S3, screening the reconstructed signal according to the multi-frequency point selection principle; S4. Perform Hilbert transform on the voltage data corresponding to the filtered signal and determine the fault location.

2. The method for detecting line faults in a fully symmetrical power distribution system based on transient power according to claim 1, characterized in that: The S2 includes: S21. Perform wavelet decomposition on the real and imaginary parts of the fault signal simultaneously according to the dual-tree complex wavelet transform to obtain the wavelet coefficient d T j (n) and scale factor c T J (n); S22, wavelet coefficient d T j (n) and scale factor c T J (n) Perform inverse transformation to extract the signal components of each original frequency band of the fault signal and complete the reconstruction of the fault signal.

3. The method for detecting line faults in a fully symmetrical power distribution system based on transient power according to claim 2, characterized in that: In S21, the expression for wavelet decomposition is: in, respectively are the wavelet coefficients of the real part tree transform and the wavelet coefficients of the imaginary part tree transform, are the wavelet coefficients of the real part tree transform and the wavelet coefficients of the imaginary part tree transform, are the scale coefficients of the real part tree transform and the imaginary part tree transform, j is the scale, J is the number of signal frequency divisions, and n is the translation amount.

4. The method for detecting line faults in a fully symmetrical power distribution system based on transient power according to claim 3, characterized in that: In S22, the reconstructed signal The expression is: Among them, ψ h , ψ g are two real wavelets, and k is the imaginary integral coefficient.

5. The method for detecting line faults in a fully symmetrical power distribution system based on transient power according to claim 1, characterized in that: In S3, the screening principle is: Among them, C is the filter capacitor; L1 is the filter inverter side inductor; C DC is the DC side capacitance, f is the target frequency for screening, f L is the lower frequency limit, f S is the sampling frequency, f R is the first resonant frequency, f PWM is the PWM carrier frequency, f H is 3000Hz, min(.) means taking the minimum value.

6. The method for detecting line faults in a fully symmetrical power distribution system based on transient power according to claim 1, characterized in that: The S4 includes: S41, performing Hilbert transform on the voltage data corresponding to the filtered signal to calculate the instantaneous reactive power of the fault component; S42. Determine the fault location based on the instantaneous reactive power of the fault components on the local side and the opposite side.

7. The method for detecting line faults in a fully symmetrical power distribution system based on transient power according to claim 6, characterized in that: In S41, the expression for calculating the instantaneous reactive power q(t) of the fault components on both sides is: in, represents the voltage fault component u at the detection point at time t fi Hilbert transform of (t); i fi (t) represents the current fault component of the detection point at time t, and P is the main value of the integration.

8. The method for detecting line faults in a fully symmetrical power distribution system based on transient power according to claim 7, characterized in that: The S42 includes: S421. Construct a distribution network line fault detection identifier λ based on the instantaneous reactive power of the fault component on the local side and the opposite side. The expression is: Among them, t s Indicates the length of the protection time window, q M (τ) represents the instantaneous reactive power of the fault component at the detection point M at time τ, q N (τ) represents the instantaneous reactive power of the fault component at the detection point N at time τ; S422. Determine the fault location according to the value of the distribution network line fault detection identifier λ: When the value of λ is greater than 0, it is determined to be an internal fault of the distribution line; When the value of λ is less than 0, it is determined to be an external fault of the distribution line.

9. The method for detecting line faults in a fully symmetrical power distribution system based on transient power according to claim 1, characterized in that: It also includes a lightning interference detection method: the high-frequency decomposition modulus maximum value W of the current component in the reconstructed signal is obtained. HS I M 、W HS I N and the maximum value of the modulus of the low-frequency decomposition quantity W LS I M 、W LS I N The ratio of is used to identify lightning interference.

10. A distribution network line fault detection system based on transient power, used to implement the full-equivalent distribution system line fault detection method based on transient power according to any one of claims 1 to 9, characterized in that: include: Voltage acquisition and calculation module: used to collect voltage data on the local and opposite sides of the distribution network line, and calculate the differential value of the voltage fault component on the local and opposite sides over time; when the differential value of the voltage fault component over time is greater than the set threshold, the fault detection module on the local and opposite sides of the line is activated respectively; Current acquisition and calculation module: used to collect current data on the local and opposite sides of the distribution network line for time-frequency analysis of high-frequency components; Fault detection module: This module uses dual-tree complex wavelet transform for preprocessing to extract high-frequency voltage and current in a specific frequency band. It then uses the Hilbert transform algorithm to quantify the difference in power transfer direction using the instantaneous reactive power values ​​of the high-frequency voltage and current at both ends, thereby constructing a fault detection identifier. It is also used to screen the maximum high-frequency modulus and the maximum low-frequency modulus of the current to prevent lightning interference.

Images