Fault positioning method and device based on zero sequence current waveform slope similarity

Abstract

The invention discloses a fault positioning method and device based on zero-sequence current waveform slope similarity. The method comprises the following steps: collecting zero-sequence current data of each detection point in a power distribution network; performing local weighted linear fitting on the collected zero-sequence current data to obtain a zero-sequence current fitting curve; deriving the function expression of the zero-sequence current fitting curve to obtain a corresponding slope curve; calculating a Pearson correlation coefficient of a zero-sequence current fitting curve slope curve of the detection points on the two sides of each section; and determining a fault section according to the minimum value of the Pearson correlation coefficient. Compared with common linear fitting, the local weighted linear fitting algorithm has higher accuracy and can effectively relieve the problem of under-fitting, so that the local weighted linear fitting algorithm is adopted to process zero-sequence current sampling data, the fitting accuracy is higher, and meanwhile, the requirement for a fault detection device can be reduced. And the method has high detection sensitivity under a high-resistance fault. And the method is also applicable to conditions of multi-branch lines and line tail end faults.

Description

Technical Field

[0001] The present application relates to the technical field, and in particular to a fault location method and device based on zero-sequence current waveform slope similarity. Background Art

[0002] Because the reliability of distribution network power supply directly impacts users' electricity experience and quality of life, it has always been a key concern in distribution network operational safety. Distribution networks cover a wide area, feature complex line layouts, and vary in operating modes, making ground faults frequent. However, when a single-phase ground fault occurs, the line voltage remains symmetrical in magnitude and phase, allowing continued operation for one to two hours, maintaining high power supply reliability. However, if the fault is not cleared promptly, it may lead to a two-phase ground short circuit, expanding the scope of the accident.

[0003] At present, although the fault location devices used in power operation and maintenance work can meet certain needs, due to the large number of distribution line branches and the access of high-density distributed power sources, the network topology will become more complex, resulting in low positioning accuracy of the current positioning devices. Therefore, there is an urgent need to develop a high-accuracy fault location method. Summary of the Invention

[0004] The present application provides a fault location method and device based on zero-sequence current waveform slope similarity to solve the above-mentioned problem.

[0005] In one aspect, the present application provides a fault location method based on zero-sequence current waveform slope similarity, the method comprising the following steps:

[0006] Collect zero-sequence current data of each detection point in the distribution network;

[0007] Performing local weighted linear fitting on the collected zero-sequence current data to obtain a zero-sequence current fitting curve;

[0008] Derivative the function expression of the zero-sequence current fitting curve to obtain the corresponding slope curve;

[0009] Calculate the Pearson correlation coefficient of the slope curve of the zero-sequence current fitting curve of the detection points on both sides of each section;

[0010] The fault section is determined based on the minimum value of the Pearson correlation coefficient.

[0011] In one implementation of the present application, the local weighted linear fitting adopts a kernel ridge regression method, which specifically includes:

[0012] Perform kernel ridge regression fitting on the zero-sequence current sampling sequence to obtain the zero-sequence current fitting curve;

[0013] The fluctuation caused by noise is suppressed by the kernel ridge regression method to obtain a smooth slope curve.

[0014] In one implementation of the present application, the specific steps of the kernel ridge regression method include:

[0015] Define a nonlinear mapping function to map the measurement data in low-dimensional space to high-dimensional space;

[0016] Use ridge regression method to establish a linear model between measurement information and state information in high-dimensional space;

[0017] The high-dimensional operation is mapped to the low-dimensional kernel space through the kernel function to calculate the optimal solution of the weight parameters.

[0018] In one implementation of the present application, the calculation formula of the Pearson correlation coefficient is:

[0019]

[0020] in, and They represent the means of X and Y respectively; the larger the absolute value of ρ, the stronger the correlation between X and Y; when X and Y tend to increase or decrease at the same time, ρ is positive; when the value of one variable in X or Y increases while the other tends to decrease, ρ is negative.

[0021] In one implementation of the present application, the criteria for determining the fault section are:

[0022] If the Pearson correlation coefficient of the zero-sequence current slope curve of the detection points on both sides of a section is less than the preset threshold, the section is determined to be a fault section;

[0023] If the Pearson correlation coefficient is greater than or equal to the preset threshold, the section is determined to be a non-fault section.

[0024] In one implementation of the present application, the method is applicable to the detection of a high-resistance fault, and specifically includes:

[0025] In the case of high-resistance fault, the amplitude of the slope curve of zero-sequence current is much larger than the amplitude of zero-sequence current, and the fault characteristics are more obvious;

[0026] The detection sensitivity of high-resistance faults can be effectively improved through similarity analysis of slope curves.

[0027] In one implementation of the present application, the method is applicable to detecting multi-branch line and line end faults, and specifically includes:

[0028] For multi-branch lines, the zero-sequence current slope curve of each branch line is calculated separately, and the fault section is determined by the Pearson correlation coefficient;

[0029] For line end faults, the fault location can be accurately identified through similarity analysis of the slope curves.

[0030] In one implementation of the present application, the method further includes fault feature extraction, specifically:

[0031] After a fault occurs, the zero-sequence current on both sides of each section is sampled;

[0032] Perform local weighted linear fitting on the sampled data to obtain the zero-sequence current fitting curve;

[0033] Derivative the function expression of the fitting curve to obtain the slope curve.

[0034] In one implementation of the present application, the method further includes a fault location step:

[0035] Calculate the Pearson correlation coefficient of the zero-sequence current slope curve on both sides of each section;

[0036] If the Pearson correlation coefficient on both sides of a section is less than the preset threshold, the section is determined to be a fault section;

[0037] If the Pearson correlation coefficient is greater than or equal to the preset threshold, the section is determined to be a non-fault section.

[0038] The present application also provides a fault location device based on zero-sequence current waveform slope similarity, the device comprising:

[0039] at least one processor; and,

[0040] a memory communicatively connected to the at least one processor; wherein,

[0041] The memory stores instructions that can be executed by the at least one processor. The instructions are executed by the at least one processor to enable the at least one processor to complete the aforementioned fault location method based on zero-sequence current waveform slope similarity.

[0042] The present application provides a fault location method and device based on zero-sequence current waveform slope similarity, which has the following beneficial effects:

[0043] (1) Compared with ordinary linear fitting, the local weighted linear fitting algorithm has higher accuracy and can effectively alleviate the underfitting problem. Therefore, the local weighted linear fitting algorithm is used to process zero-sequence current sampling data, which has higher fitting accuracy and reduces the requirements for fault detection devices.

[0044] (2) The amplitude of the slope curve is much larger than the amplitude of the zero-sequence current. The similarity of the slope curve waveform of the zero-sequence current fitting curve is used to identify the fault section, and the fault characteristics are more obvious.

[0045] (3) It has high detection sensitivity under high resistance fault conditions.

[0046] (4) This method is also applicable to multi-branch lines and line end faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0048] Figure 1 A flow chart of a fault location method based on zero-sequence current waveform slope similarity provided in an embodiment of the present application;

[0049] Figure 2 A fitting curve diagram of the zero-sequence current provided in an embodiment of the present application;

[0050] Figure 3 The zero-sequence current sampling points and their fitting curves provided in the embodiment of the present application;

[0051] Figure 4 A slope curve diagram of a zero-sequence current fitting curve provided in an embodiment of the application;

[0052] Figure 5 A logic diagram of a fault location algorithm based on zero-sequence current waveform slope similarity provided in an embodiment of the present application;

[0053] Figure 6 A schematic diagram of a fault location device based on zero-sequence current waveform slope similarity provided in an embodiment of the present application. DETAILED DESCRIPTION

[0054] To make the purpose, technical solutions, and advantages of this application more clear, the technical solutions of this application will be clearly and completely described below in conjunction with the specific embodiments of this application and the corresponding drawings. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0055] The embodiment of the present application provides a fault location method and device based on the similarity of the slope of the zero-sequence current waveform. The technical solution proposed in the embodiment of the present application is described in detail below with reference to the accompanying drawings.

[0056] Figure 1 This is a flow chart of a fault location method based on the similarity of zero-sequence current waveform slope provided in an embodiment of the present application. Figure 1As shown, the method mainly includes the following steps:

[0057] Step S1: collecting zero-sequence current data of each detection point in the distribution network;

[0058] Step S2: performing local weighted linear fitting on the collected zero-sequence current data to obtain a zero-sequence current fitting curve;

[0059] Step S3: deriving the function expression of the zero-sequence current fitting curve to obtain a corresponding slope curve;

[0060] Step S4: Calculate the Pearson correlation coefficient of the slope curve of the zero-sequence current fitting curve of the detection points on both sides of each section;

[0061] Step S5: Determine the fault section according to the minimum value of the Pearson correlation coefficient.

[0062] In the embodiment of the present application, since the steady-state component of grounding is usually very weak, it will cause the line selection device to be difficult to operate correctly, thereby affecting the accuracy of fault detection. In addition, the steady-state method requires a continuous steady-state short-circuit process after the fault occurs, and is not suitable for systems with intermittent arc grounding or neutral point grounding through an arc suppression coil. In comparison, the transient method has a higher success rate and wider adaptability. However, the transient method is more complicated than the steady-state method in the detection process. When it is necessary to filter out the power frequency component, it requires higher data acquisition and processing capabilities of the positioning device. Since the slope calculated by simple derivation will be affected by irregular distortion caused by noise and is prone to fluctuations, a local weighted linear fitting algorithm is first used to process the zero-sequence current sampling sequence separately to smooth out such fluctuations.

[0063] This solution uses data-based kernel ridge regression to solve the overfitting and underfitting problems that are prone to occur in previous solutions.

[0064] For the ordinary linear regression model: X k =w k Z k , where X k is the state prediction value, w k is the weight parameter, Z k is the node measurement information matrix, and the least squares method is used to obtain the weight parameter w k When the objective function of the least squares linear regression is set to: J(w)=||Xw k Z k || 2. In the formula, λ is added to balance the regularization term and the loss function. To minimize the loss function J(w), the weight parameter is derived and set to 0.

[0065] Also because:

[0066]

[0067] Also because:

[0068] (X k -wZ k ) T (X k -wZ k )=(X T -wZ T,K )(X k -Z k w)

[0069] =X T,k Z k -S T,k Z k ww T Z T X k +w T Z T Z k w

[0070] Therefore, the equivalent is:

[0071]

[0072] Take 0 and simplify to get the parameter w k The optimal solution of .

[0073]

[0074] Where I is the unit matrix and the optimal solution is It is the sum of a semi-positive definite matrix and a diagonal matrix, which can solve the problem of matrix irreversibility caused by insufficient measurement information, resulting in overfitting of the model. However, the nonlinear relationship between the measurement data and the state data will greatly reduce the accuracy of the linear ridge regression. Considering that the nonlinear data in the low-dimensional space can be mapped to the high-dimensional space to achieve the linear separability of the data, the embodiment of the present application introduces a mapping function to map the measurement information to high dimensions, and uses the ridge regression method to establish a linear model between the measurement and state information. In order to use the kernel function in the ridge regression, the optimal solution of the parameters is To perform the transformation, first use the matrix inversion theorem:

[0075] (U -1 +V T W -1 V) -1 V T W -1 =UV T (VUV T +W) -1

[0076] because V=Z k ,W=I, which is equivalent to:

[0077]

[0078] At this time, the weight parameter becomes:

[0079]

[0080] In order to map the measurement data from low-dimensional space to high-dimensional space, it is necessary to define a nonlinear mapping function φ(·) and map the measurement data Z to φ(Z). At this time, the objective function of the ridge regression model is converted to:

[0081]

[0082] The optimal solution is:

[0083]

[0084] Generally speaking, the nonlinear mapping function φ(Z) is in a high-dimensional space and is abstract. Usually, its inner product cannot be directly calculated. However, the kernel function can map the high-dimensional operation to a low-dimensional kernel space to obtain its inner product. At this time, we define the kernel function K.

[0085] K(S i ,S j )=[φ(Z k ) T ]=(φ(S j )φ T (S j ))

[0086] make:

[0087] K=K(Z K ,Z K )=φ(Z k )φ(Z k ) T

[0088] a=(φ(Z K )φ(Z K ) T +λI) -1 X

[0089] At this time, the intermediate variable a becomes:

[0090] a=(K+λI) -1 X k

[0091] Then the weight parameter after kernel function processing becomes:

[0092]

[0093] Then the current estimation value is obtained as follows:

[0094]

[0095] Kernel ridge regression can be seen as a low-pass filtering method, which controls the w of the sampled data by adjusting ε i , which can effectively suppress high-frequency information. In order to intuitively demonstrate the superiority of LWLR, the zero-sequence current sampling data of the healthy line are fitted by least squares polynomial fitting and kernel ridge regression respectively. The fitting curves Isf and Ire are shown in Figure 2 shown.

[0096] As can be seen, Isf is a straight line that cannot effectively fit the changing trend of the zero-sequence current data, a phenomenon known as underfitting. In contrast, Ire has higher accuracy, effectively alleviating the underfitting problem and better fitting the actual changing trend of the data. Therefore, using a local weighted linear fitting algorithm to process the zero-sequence current sampling sequence can effectively eliminate the fluctuations caused by external factors such as noise while preserving the characteristics of the sampled data. This results in smoother and more accurate slope data, providing strong support for subsequent analysis and judgment.

[0097] Furthermore, suppose that the function y = f(x) is defined in a neighborhood around the point x0. When the independent variable x increments at x0 and (x0 + Δx) is also in this neighborhood, the corresponding function increment Δy = f(x0 + Δx) - f(x0). If a limit exists for the ratio of Δy to Δx when Δx → 0, then the function y = f(x) is said to be differentiable at x0, and this limit is called the derivative of the function y = f(x) at x0, denoted by:

[0098]

[0099] f′(x0) is actually a local linear approximation of the function through the concept of limit. Its geometric meaning is the tangent slope of the function at x0, which reflects the speed of change of y at x0. When an SPG fault occurs in a low-current grounding system, the transient zero-sequence current is approximately:

[0100]

[0101] The slope curve calculated by direct derivation fluctuates greatly. In order to smooth out this fluctuation, the local weighted linear fitting algorithm is first used to process the zero-sequence current sampling sequence. When a phase A grounding fault occurs, the zero-sequence current sampling points M and N upstream of the fault point and the zero-sequence current sampling points P and Q downstream of the fault point and their local weighted fitting curves are as follows: Figure 3 shown.

[0102] analyze Figure 3 It can be seen that the amplitude of the zero-sequence current fitting curves at detection points M and N upstream of the fault point is greater than that of detection points P and Q downstream of the fault point. There are significant differences between the zero-sequence current waveforms of M and N and those of P and Q, and the zero-sequence current waveforms of M and N (or P and Q) are highly similar. The fitting results fully reflect the changing characteristics of the zero-sequence current at each detection point and have higher fitting accuracy than conventional polynomial linear fitting. Because the zero-sequence current fitting curves of detection points on both sides of the fault point differ significantly, the zero-sequence current fitting curves of detection points on the same side upstream (or downstream) of the fault point are highly similar. Therefore, the slope curves of the zero-sequence current fitting curves upstream and downstream of the fault point are significantly different, and the slope curves of the zero-sequence current fitting curves of detection points on the same side of the fault point are highly similar.

[0103] respectively Figure 3 The function expression of the zero-sequence current fitting curve is derived to obtain the slope curves of the zero-sequence current fitting curves of the upstream detection points M and N of the fault point and the downstream detection points P and Q of the fault point, as shown in Figure 4 shown.

[0104] Depend on Figure 4 It can be seen that after taking the derivative of the function expression for the zero-sequence current fitting curve, the resulting slope curve amplitude is much greater than the zero-sequence current amplitude, indicating a more pronounced fault signature. Furthermore, the slope curve amplitude of the zero-sequence current fitting curve for detection points M and N upstream of the fault is much greater than that for detection points P and Q downstream of the fault. Furthermore, the slope curve waveforms of M and N differ significantly from those of P and Q, and the slope curve waveforms of M and N (or P and Q) are highly similar. Therefore, the similarity of the slope curve waveforms of the zero-sequence current fitting curve can be considered for identifying faulty sections. This metric can better reflect the difference in zero-sequence current between faulty and non-faulty sections.

[0105] Furthermore, the transient zero-sequence currents upstream and downstream of the fault point have significant differences in amplitude and phase, so the fault can be located by comparing the waveforms of the transient zero-sequence currents at adjacent monitoring points. If the waveform similarity is high, it is determined to be a non-fault section; if the waveform similarity is low, it is determined to be a fault section. To characterize the degree of waveform similarity, the Pearson Product-Moment Correlation Coefficient (PPMCC) is introduced to measure the linear correlation between two variables, with a value between -1 and 1. The definition of PPMCC is as shown below.

[0106]

[0107] Where: and denote the means of X and Y, respectively. The larger the absolute value of ρ, the stronger the correlation between X and Y. ρ is positive when X and Y tend to increase or decrease simultaneously; it is negative when one of X and Y increases while the other tends to decrease.

[0108] Furthermore, the present embodiment uses a local weighted linear fitting algorithm to process the sampled data, reducing the requirements for the fault detection device. Therefore, the PPMCC coefficient of the zero-sequence current waveform slope curve on both sides of each segment can be calculated for low-current ground fault location, simplifying the location principle while improving fault location accuracy.

[0109] Assuming that the detection points on both sides of the segment [i, j] are represented by i and j respectively, the zero-sequence current fitting curve is and The slopes of the curves are k 0i (n) and k 0j (n), the PPMCC coefficient calculation formulas for the two are as follows:

[0110]

[0111] Where: Correlation coefficient ρ ij Reflects the slope curve Δk of the two zero-sequence current fitting curves 0i (n) and Δk 0j (n) similarity. ρ ij The value of is between -1 and 1. ij When the value approaches the range boundary, it indicates that the linear correlation between variables is good; when the linear correlation between variables is poor, ρ ij tends to 0. Generally speaking, it is believed that when |ρ ij When |>0.8, there is a strong linear correlation between the variables.

[0112] Therefore, the slope curves of the zero-sequence current fitting curves on both sides of the non-fault section are highly similar, and the correlation coefficient is close to 1; the slope curves of the zero-sequence current fitting curves on both sides of the fault section are quite different, and the correlation coefficient is small.

[0113] Based on the above fault feature analysis, in order to solve the problem that the existing high-resistance fault location method has low location accuracy and reliability needs to be improved, this application proposes a small current grounding fault location method based on the Pearson correlation of the zero-sequence current waveform slope based on the analysis of the zero-sequence current distribution and slope curve characteristics of small current grounding faults. The Pearson correlation coefficient ρ of the slope curve of the local weighted linear fitting curve of the zero-sequence current on both sides of each section is used. ij As a fault judgment criterion, the fault section is identified.

[0114] First, perform local weighted linear fitting on the zero-sequence current of each detection point, then derive the function expression of the zero-sequence current fitting curve to obtain the corresponding slope curve, and finally determine the fault section based on the similarity relationship between the slope curves of the zero-sequence current fitting curves upstream and downstream of the fault point. Calculate the Pearson correlation coefficient of the detection points on both sides of each section respectively. If there is a ρ ij ≤0, that is, there is a poor correlation or negative correlation between the slope curves of the zero-sequence current fitting curves on both sides of a certain section, then the section is judged to be a fault section; if ρ ij →1, that is, the slopes of the zero-sequence current fitting curves on both sides of the section are highly correlated, and it is judged to be a non-fault section.

[0115] For a distribution network that combines overhead and cable lines, the specific steps for locating a small current ground fault based on the Pearson correlation of the zero-sequence current waveform slope are as follows:

[0116] (1) Collect zero-sequence voltage U0.

[0117] (2) Fault detection: When U0>kUN (UN is the rated phase voltage, k is usually between 0.1 and 0.15), the protection device starts.

[0118] (3) Extract fault characteristic information. After a fault occurs, the zero-sequence current on both sides of each section is sampled.

[0119] (4) Obtain the zero-sequence current fitting curve. Perform local weighted linear fitting on the sampled data to obtain the zero-sequence current fitting curve.

[0120] (5) Calculate the slope curve of each fitting curve. By taking the derivative of the function expression of the zero-sequence current fitting curve, the slope curve of the zero-sequence current fitting curve on both sides of each section is obtained.

[0121] (6) Calculate the Pearson correlation coefficient ρ of the slope curve on both sides of each segment ij If there exists ρ ij ≤0, the fault is judged to have occurred in the segment [i,j].

[0122] The embodiment of the present application provides a logic diagram of an overall fault location algorithm based on the similarity of the zero-sequence current waveform slope, such as Figure 5 shown.

[0123] The above is a fault location method based on the similarity of the zero-sequence current waveform slope provided by the embodiment of the present application. Based on the same inventive concept, the present application also provides a fault location device based on the similarity of the zero-sequence current waveform slope, such as Figure 6As shown, the device includes: at least one processor 601; and a memory 602 communicatively connected to the at least one processor 601; wherein the memory 602 stores instructions that can be executed by the at least one processor 601, and the instructions are executed by the at least one processor 601 to enable the at least one processor 601 to complete the aforementioned fault location method based on the similarity of the zero-sequence current waveform slope.

[0124] The various embodiments in this application are described in a progressive manner. Similar parts between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences from other embodiments. In particular, the device embodiments are generally similar to the method embodiments, so the description is relatively simple. For relevant parts, refer to the partial description of the method embodiments.

[0125] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0126] The foregoing is merely an 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 changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.

Claims

1. A fault location method based on zero-sequence current waveform slope similarity, characterized in that: The method comprises the following steps: Collect zero-sequence current data of each detection point in the distribution network; Performing local weighted linear fitting on the collected zero-sequence current data to obtain a zero-sequence current fitting curve; Derivative the function expression of the zero-sequence current fitting curve to obtain the corresponding slope curve; Calculate the Pearson correlation coefficient of the slope curve of the zero-sequence current fitting curve of the detection points on both sides of each section; The fault section is determined based on the minimum value of the Pearson correlation coefficient.

2. A fault location method based on zero-sequence current waveform slope similarity according to claim 1, characterized in that: The local weighted linear fitting adopts the kernel ridge regression method, which specifically includes: Perform kernel ridge regression fitting on the zero-sequence current sampling sequence to obtain the zero-sequence current fitting curve; The fluctuation caused by noise is suppressed by the kernel ridge regression method to obtain a smooth slope curve.

3. A fault location method based on zero-sequence current waveform slope similarity according to claim 2, characterized in that: The specific steps of the kernel ridge regression method include: Define a nonlinear mapping function to map the measurement data in low-dimensional space to high-dimensional space; Use ridge regression method to establish a linear model between measurement information and state information in high-dimensional space; The high-dimensional operation is mapped to the low-dimensional kernel space through the kernel function to calculate the optimal solution of the weight parameters.

4. The fault location method based on zero-sequence current waveform slope similarity according to claim 1, characterized in that: The calculation formula of the Pearson correlation coefficient is: in, and They represent the means of X and Y respectively; the larger the absolute value of ρ, the stronger the correlation between X and Y; when X and Y tend to increase or decrease at the same time, ρ is positive; when the value of one variable in X or Y increases while the other tends to decrease, ρ is negative.

5. The fault location method based on zero-sequence current waveform slope similarity according to claim 1, characterized in that: The criteria for determining the fault section are: If the Pearson correlation coefficient of the zero-sequence current slope curve of the detection points on both sides of a section is less than the preset threshold, the section is determined to be a fault section; If the Pearson correlation coefficient is greater than or equal to the preset threshold, the section is determined to be a non-fault section.

6. The fault location method based on zero-sequence current waveform slope similarity according to claim 1, characterized in that: The method is applicable to the detection of high-resistance faults, and specifically includes: In the case of high-resistance fault, the amplitude of the slope curve of zero-sequence current is much larger than the amplitude of zero-sequence current, and the fault characteristics are more obvious; The detection sensitivity of high-resistance faults can be effectively improved through similarity analysis of slope curves.

7. The fault location method based on zero-sequence current waveform slope similarity according to claim 1, characterized in that: The method is applicable to the detection of multi-branch line and line end faults, and specifically includes: For multi-branch lines, the zero-sequence current slope curve of each branch line is calculated separately, and the fault section is determined by the Pearson correlation coefficient; For line end faults, the fault location can be accurately identified through similarity analysis of the slope curves.

8. The fault location method based on zero-sequence current waveform slope similarity according to claim 1, characterized in that: The method further includes fault feature extraction, specifically: After a fault occurs, the zero-sequence current on both sides of each section is sampled; Perform local weighted linear fitting on the sampled data to obtain the zero-sequence current fitting curve; Derivative the function expression of the fitting curve to obtain the slope curve.

9. The fault location method based on zero-sequence current waveform slope similarity according to claim 1, characterized in that: The method further includes fault location, specifically: Calculate the Pearson correlation coefficient of the zero-sequence current slope curve on both sides of each section; If the Pearson correlation coefficient on both sides of a section is less than the preset threshold, the section is determined to be a fault section; If the Pearson correlation coefficient is greater than or equal to the preset threshold, the section is determined to be a non-fault section.

10. A fault location device based on zero-sequence current waveform slope similarity, characterized in that: The device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can complete the fault location method based on zero-sequence current waveform slope similarity as described in any one of claims 1-9.

Images