Single-phase earth fault feature recognition and phase selection method

Abstract

The invention discloses a single-phase earth fault feature recognition and phase selection method, which comprises the following steps of: in a start judgment process of a single-phase earth fault, constructing a zero-sequence voltage forward sequence to obtain a first similarity of the zero-sequence voltage forward sequence between two branches; comparing the zero-sequence voltages of the corresponding branches in pairs to obtain a zero-sequence voltage form difference value to construct a fault line zero-sequence voltage variation value, and judging whether a certain branch is started or not according to the zero-sequence voltage variation value; accurate starting conditions are obtained under the condition that the fault current of the single-phase earth fault is difficult to capture, and the accuracy of the subsequent fault identification and phase selection process is further improved.

Description

Technical Field

[0001] The invention belongs to the field of single-phase grounding fault processing, and in particular relates to a single-phase grounding fault feature recognition and phase selection method. Background Art

[0002] The distribution network low current grounding system refers to a three-phase system in which the medium and low voltage distribution network adopts the neutral point ungrounded, the neutral point grounded through the arc suppression coil, and the neutral point grounded through the high resistance. When a single-phase grounding fault occurs, the three-phase line voltage of the line remains basically unchanged, and the fault current is small, which is often difficult to identify and handle.

[0003] At present, the academic community has proposed many methods for selecting small current grounding lines, such as zero-sequence component method, transient power direction method, injection method, traveling wave method, etc. However, these methods have some limitations: for example, they are affected by human experience, and are also affected by changes in the topology of the distribution network and environmental factors, and the accuracy of single-phase grounding fault identification is low. The methods introduced by emerging artificial intelligence technology in the field of single-phase grounding detection also rely on a large amount of historical data training, which is not very practical at the current technical level.

[0004] In addition, the prior art determines the operating status of the power grid by calculating the similarity of the current in the distribution line. However, in a single-phase grounding fault, due to the short occurrence time of the fault current, the small change range and the objective current collection limitations, it is impossible to accurately identify the characteristics of a single-phase grounding fault by judging the similarity of the current. Summary of the invention

[0005] In order to solve the problems of low accuracy of line selection for single-phase grounding faults in the current power distribution network and insufficient ability to identify characteristics of minor faults, the present invention provides a single-phase grounding fault feature identification and phase selection method that can accurately identify single-phase grounding faults and precisely locate the faulty phase and improve the accuracy of grounding line selection.

[0006] A single-phase grounding fault feature recognition and phase selection method comprises the following steps: S1: Obtain a zero-sequence voltage mutation amount; the zero-sequence voltage is a self-generated zero-sequence voltage; S2: Obtaining the effective value of the zero-sequence component; the effective value of the zero-sequence component includes the effective value of the self-generated zero-sequence voltage and current or the effective value of the external zero-sequence voltage and current; S3: construct a zero-sequence voltage positive sequence based on the absolute value of the self-generated zero-sequence voltage or the external zero-sequence voltage of each branch of the distribution network line, so as to obtain the first similarity of the zero-sequence voltage positive sequence between two branches, and compare the zero-sequence voltages of the corresponding branches in pairs to obtain the zero-sequence voltage shape difference value to construct the zero-sequence voltage change value of the fault line, and determine whether a branch is grounded and the line selection is started according to the zero-sequence voltage change value; S4: After the start is determined, when the line state is normal operation; according to the preset fixed value, determine whether the zero-sequence voltage mutation or the zero-sequence voltage effective value exceeds the limit to start; calculate the zero-sequence reactive power; determine whether it is an internal fault by the polarity of the zero-sequence reactive power; confirm the start in the case of an internal fault, and start the alarm or trip delay; among which, the zero-sequence voltage effective value is the self-produced zero-sequence voltage effective value or the external zero-sequence voltage effective value; After the start-up, if the line selection status is not running normally, find the reference point from the start-up point, calculate the mutation amount point by point for the three-phase current sampling data of the half sampling cycle length after the reference point, and select the phase with the largest cycle-to-cycle mutation amount as the grounded phase; S5: When the action conditions are met, output alarm or trip signal and phase selection result.

[0007] Further, in S2, the external zero-sequence voltage and current effective value includes an external zero-sequence current effective value and an external zero-sequence voltage effective value, the external zero-sequence current effective value is directly collected by a zero-sequence current transformer CT, and the external zero-sequence voltage effective value is directly collected by a zero-sequence voltage transformer PT.

[0008] Further, the self-produced zero-sequence voltage and current effective value in S2 includes the self-produced zero-sequence voltage effective value and the self-produced zero-sequence current effective value; The sampling value of the zero-sequence voltage is obtained by adding the sampling values ​​of the three-phase voltage, and then the sampling value of the zero-sequence voltage is processed by the full-cycle Fourier differential filtering algorithm to obtain the effective value of the self-generated zero-sequence voltage; The sampling value of the zero-sequence current is obtained by adding the sampling values ​​of the three-phase current, and then the sampling value of the zero-sequence current is processed by a full-cycle Fourier differential filtering algorithm to obtain the self-generated zero-sequence current effective value.

[0009] Furthermore, S3 is specifically: S31: Construct the zero-sequence voltage forward sequence: Obtain the self-generated zero-sequence voltage or external zero-sequence voltage of each branch of the distribution line at the current sampling time, take the absolute value and store it in a two-dimensional array in a loop. Each row of the two-dimensional array represents the absolute value of the zero-sequence voltage of a branch at different sampling times, that is, the positive sequence of the zero-sequence voltage; each column of the two-dimensional array represents the absolute value of the zero-sequence voltage of each branch at the same time; S32: Obtain the first similarity between any two branches in the zero-sequence voltage forward sequence: The similarity between the positive sequences of zero-sequence voltages of any two branches is recorded as the first similarity; S33: Calculate the difference in zero-sequence voltage shape between any two branches: After cyclically storing the two-dimensional array, the difference of the absolute value of the zero-sequence voltage at the same sampling time is calculated based on the zero-sequence voltage of any two branches, and the number of occurrences of each difference value is counted, and the difference value with low frequency is discarded, and the sum of all remaining differences is marked as the zero-sequence voltage shape difference value of the two branches; S34: Calculate the zero-sequence voltage variation value of each branch: expressed as: ; Where: is the zero-sequence voltage variation value of the i-th branch; n is the total number of branches of the distribution line; is the first similarity between the zero-sequence voltages of the i-th branch and the k-th branch; is the difference in zero-sequence voltage shape between the ith branch and the kth branch; To prevent decimals with a denominator of 0; S35: Confirm start: Based on the zero-sequence voltage variation value of each branch, the average value is taken according to the data of the branch. When the zero-sequence voltage variation value of a branch exceeds the limit and the corresponding average value is the largest, it is determined that the line grounding selection is started.

[0010] Furthermore, in S4, the starting constant of the zero-sequence voltage effective value takes a typical value of 8V, and the constant of the zero-sequence voltage mutation amount takes a typical value of 5V.

[0011] Furthermore, in S4, whether it is an intra-zone fault is determined by the zero-sequence reactive power polarity, specifically: The zero-sequence reactive power is calculated at the moment of grounding. When the zero-sequence reactive power is greater than 0, it is determined as an internal fault. When the zero-sequence reactive power is less than 0, it is determined as an external fault.

[0012] Furthermore, in step S4, the reference point is found from the starting point, the mutation amount is calculated point by point for the three-phase current sampling data of the half sampling period length after the reference point, and the phase with the largest cycle-to-cycle mutation amount among the three phases is selected as the grounded phase. Specifically, the process is as follows: S41: Data storage: After the grounding line selection start confirmation is detected, n sampling points are pushed forward from the start point, and the three-phase current sampling values ​​of the previous 2T are stored in the array for standby use with the sampling point as the reference point, where T represents the sampling period; S42: Calculation of three-phase current mutation: Calculate the three-phase current sampling values ​​of T / 2 length after the reference point point by point; first obtain the current point sampling value, then obtain the sampling value 1T ago and the sampling value 2T ago from the array; calculate the difference between the current point mutation amount and the point mutation amount 1T ago, that is, the cycle mutation amount; S43: Get the maximum value of the three-phase current mutation: In the data T / 2 after the reference point, the cycle-to-cycle mutation of the three-phase currents A, B, and C is calculated point by point, and the maximum cycle-to-cycle mutation of each phase is obtained; S44: Select the grounding phase: Compare the maximum values ​​of the cycle-to-cycle mutations of each phase and select the phase corresponding to the maximum value as the grounded phase.

[0013] Furthermore, the calculation formula for the mutation amount of phase A in S43 is as follows: ; ; ; Where: is the sampling value of phase A current at the current point; is the sampling value of phase A current 1T ago; is the sampling value of phase A current 2T ago; is the modulus of the current point mutation value of phase A current; is the modulus of the point mutation quantity of phase A current 1T ago; is the weekly mutation amount of the current point, when When it is less than 0, Assign the value 0.

[0014] The beneficial effects of the present invention are:

[0015] The present invention obtains the first similarity of the zero-sequence voltage positive sequences between two branches by constructing a zero-sequence voltage positive sequence in the process of starting judgment of a single-phase grounding fault, and compares the zero-sequence voltages of the corresponding branches in pairs to obtain the zero-sequence voltage shape difference value to construct the zero-sequence voltage change value of the fault line, and determines whether a branch is started according to the zero-sequence voltage change value; in the case where the fault current of a single-phase grounding fault is difficult to capture, accurate starting conditions are obtained, and it is not affected by line topology changes and environmental factors, which greatly improves the power supply reliability and further improves the accuracy of subsequent fault identification and phase selection processes.

[0016] In actual tests, the scheme of the present invention has passed the grounding waveform inversion and dynamic model test. In particular, in the dynamic model test, this method has been tested hundreds of times for various test scenarios such as stability faults, intermittent faults, intermittent arc grounding faults, PT line breaks, inconsistent CT ratios, and reverse CT polarity in overhead lines and cable lines in neutral point ungrounded systems, arc suppression coil grounding systems, and low resistance grounding systems. The accuracy of action and phase selection is above 98%. The effect of the present invention is significantly better than the currently common single-phase grounding line selection method. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1It is a flow chart of a single-phase grounding fault feature identification and phase selection method of the present invention. DETAILED DESCRIPTION

[0018] The following are only preferred specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a technician familiar with the technical field within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. The embodiments described below are only used to explain the present invention and cannot be interpreted as limitations on the present invention. The protection scope of the present invention should be based on the protection scope of the claims. The embodiments of the present invention are described in detail below. In order to facilitate the description of the present invention and simplify the description, the technical terms used in the specification of the present invention should be interpreted in a broad sense, including but not limited to conventional replacement schemes not mentioned in this application, and also including direct implementation methods and indirect implementation methods.

[0019] Combination Figure 1 This embodiment describes a single-phase grounding fault feature recognition and phase selection method disclosed in this embodiment, including the following steps:

[0020] like Figure 1 As shown, first determine whether the grounding line selection function is enabled. The program determines this based on the status of the soft pressure plate and the control word. When both are enabled, the grounding line selection function is enabled. The second step is to determine whether the line selection function is blocked. The device is blocked by PT resonance and PT disconnection. If the blocking condition is not met, the grounding line selection function can be started normally.

[0021] Determine whether the device is in the startup state. When it detects that the zero-sequence voltage mutation exceeds the limit, the zero-sequence voltage exceeds the limit, or the grounding selection has been started, the program confirms that it is in the startup state, and can perform fault feature identification, phase selection, action, return, etc. Otherwise, the program confirms that it is in the normal operation state and returns after initialization.

[0022] When judging the occurrence of a fault and selecting a phase, calculate the zero-sequence reactive power (self-produced, external zero-sequence can be used through the control word) and the three-phase current mutation. When the polarity of the zero-sequence reactive power is positive, it is confirmed as an internal fault; when the polarity of the zero-sequence reactive power is negative, it is confirmed as an external fault; select the fault phase with the largest three-phase current mutation, and the phase selection result is output together with the tripping / alarm signal.

[0023] A single-phase grounding fault feature recognition and phase selection method, in one example, the method includes: S1: Obtain the zero-sequence voltage mutation amount; the zero-sequence voltage is the self-produced zero-sequence voltage; specifically: Synthesize self-generated zero-sequence voltage through three-phase voltage sampling channels: Where: is the self-produced zero-sequence voltage, is the A phase voltage sampling value, is the B phase voltage sampling value, is the C phase voltage sampling value.

[0024] Calculate the self-produced zero-sequence voltage mutation amount using the self-produced zero-sequence voltage: ; ; ; Where: is the current sampling value of the self-produced zero-sequence voltage, is the sampling value of the self-produced zero-sequence voltage 1T before, is the sampling value of the self-produced zero-sequence voltage 2T before, is the current point mutation value of the self-produced zero-sequence voltage, is the point mutation amount of the self-produced zero-sequence voltage 1T before, It is the weekly mutation amount of the self-generated zero-sequence voltage at the current point.

[0025] S2: Obtain the effective value of the zero-sequence component; the effective value of the zero-sequence component includes the effective value of the self-generated zero-sequence voltage and current or the effective value of the external zero-sequence voltage and current; specifically: The sampling value of the zero-sequence voltage is obtained by adding the sampling values ​​of the three-phase voltage, and then the sampling value of the zero-sequence voltage is processed by the full-cycle Fourier differential filtering algorithm to obtain the effective value of the self-generated zero-sequence voltage; The sampling value of the zero-sequence current is obtained by adding the sampling values ​​of the three-phase current, and then the sampling value of the zero-sequence current is processed by a full-cycle Fourier differential filtering algorithm to obtain the self-generated zero-sequence current effective value.

[0026] The effective value of the self-produced zero-sequence voltage is calculated and expressed as: ; ; ; Where: is the effective value of the self-produced zero-sequence voltage, is the A-phase component of the self-generated zero-sequence voltage, is the B-phase component of the self-generated zero-sequence voltage, is the C-phase component of the self-generated zero-sequence voltage, is the real part of the self-generated zero-sequence voltage, is the real part of the A-phase component, is the real part of the B-phase component, is the real part of the C-phase component, is the imaginary part of the self-generated zero-sequence voltage, is the imaginary part of the A-phase component, is the imaginary part of the B-phase component, is the imaginary part of the C-phase component; The calculation method for the effective value of self-produced zero-sequence current is the same as that for self-produced zero-sequence voltage. Please refer to the above formula.

[0027] The external zero-sequence voltage and current effective value includes an external zero-sequence current effective value and an external zero-sequence voltage effective value. The external zero-sequence current effective value is directly collected by a zero-sequence current transformer CT, and the external zero-sequence voltage effective value is directly collected by a zero-sequence voltage transformer PT.

[0028] S3: A zero-sequence voltage positive sequence is constructed based on the absolute value of the self-generated zero-sequence voltage or the external zero-sequence voltage of each branch of the distribution network line, so as to obtain the first similarity of the zero-sequence voltage positive sequence between two branches, and the zero-sequence voltages of the corresponding branches are compared pairwise to obtain the zero-sequence voltage morphology difference value to construct the zero-sequence voltage change value of the fault line, and whether a branch is grounded and the line selection is started is determined according to the zero-sequence voltage change value.

[0029] Furthermore, S3 is specifically: S31: Construct the zero-sequence voltage forward sequence: Obtain the self-generated zero-sequence voltage or external zero-sequence voltage of each branch of the distribution line at the current sampling time, take the absolute value and store it in a two-dimensional array in a loop. Each row of the two-dimensional array represents the absolute value of the zero-sequence voltage of a branch at different sampling times, that is, the positive sequence of the zero-sequence voltage; each column of the two-dimensional array represents the absolute value of the zero-sequence voltage of each branch at the same time; S32: Obtain the first similarity between any two branches in the zero-sequence voltage forward sequence: The similarity between the positive sequences of zero-sequence voltages of any two branches is recorded as the first similarity; S33: Calculate the difference in zero-sequence voltage shape between any two branches: After cyclically storing the two-dimensional array, the difference of the absolute value of the zero-sequence voltage at the same sampling time is calculated based on the zero-sequence voltage of any two branches, and the number of occurrences of each difference value is counted, and the difference value with low frequency is discarded, and the sum of all remaining differences is marked as the zero-sequence voltage shape difference value of the two branches; S34: Calculate the zero-sequence voltage variation value of each branch: expressed as: ; Where: is the zero-sequence voltage variation value of the i-th branch; n is the total number of branches of the distribution line; is the first similarity between the zero-sequence voltages of the i-th branch and the k-th branch; is the difference in zero-sequence voltage shape between the ith branch and the kth branch; To prevent decimals with a denominator of 0; S35: Confirm start: Based on the zero-sequence voltage variation value of each branch, the average value is taken according to the data of the branch. When the zero-sequence voltage variation value of a branch exceeds the limit and the corresponding average value is the largest, it is determined that the line grounding selection is started.

[0030] S4: After the start-up is determined, when the line selection state is normal operation; according to the preset set value, determine whether the zero-sequence voltage mutation amount or the zero-sequence voltage effective value exceeds the limit to start. When any of the two exceeds the limit or the grounding line selection function has been started, start the grounding determination, otherwise exit; the zero-sequence voltage start-up set value and the zero-sequence voltage mutation amount set value are both empirical values, respectively, the start-up set value of the zero-sequence voltage effective value takes a typical value of 8V, and the set value of the zero-sequence voltage mutation amount takes a typical value of 5V. When any of the three conditions, namely, the zero-sequence voltage mutation amount exceeds the limit, the zero-sequence voltage exceeds the limit, and the device has been started, it means that the prerequisites for grounding start-up have been met, and the start-up, action, etc. can be determined. When none of the three conditions are met, it means that the current normal operation can be directly exited. When the zero-sequence voltage effective value exceeds the limit, the action determination begins; the zero-sequence reactive power is calculated; the zero-sequence reactive power polarity is used to determine whether it is an in-zone fault; when there is an in-zone fault, the start-up is confirmed, and the alarm or tripping delay begins; among which, the zero-sequence voltage effective value is the self-produced zero-sequence voltage effective value or the external zero-sequence voltage effective value; Based on the characteristics of single-phase grounding fault, the zero-sequence voltage and zero-sequence current effective values ​​are selected as exceeding the limit as the basic condition, and the zero-sequence reactive power at the current point is calculated through the sampling channels of zero-sequence voltage and zero-sequence current.

[0031] The limit violation is expressed as: ; ; Where: is the effective value of self-produced / external zero-sequence voltage, is the effective value of self-generated / zero-sequence current, is the starting setting of zero-sequence voltage, It is the starting setting of zero-sequence current, and the zero-sequence current setting is 0.03A.

[0032] According to the polarity of zero-sequence reactive power, the fault inside and outside the zone is judged: when the zero-sequence reactive power is greater than 0, it is judged as an internal fault; when the zero-sequence reactive power is greater than 0, it is judged as an internal fault; in order to prevent false operation, a threshold value of 0.4Var is set for zero-sequence reactive power, Var represents the unit of reactive power, that is, var. When the zero-sequence reactive power calculated by four consecutive sampling points is greater than the threshold value, it is judged as an internal fault, the start is confirmed, the start flag is set, and the delayed alarm or tripping begins.

[0033] If the zero-sequence voltage does not exceed the limit, the judgment will be returned.

[0034] Furthermore, after S4 determines that the line selection state is not operating normally after starting, it finds the reference point from the starting point, calculates the mutation amount point by point for the three-phase current sampling data of T / 2 length after the reference point, and selects the phase with the largest mutation amount among the three phases as the grounded phase; specifically: S41: Data storage: After the grounding line selection start is confirmed, since the condition that the zero-sequence reactive power of 4 consecutive sampling points is greater than 0.4Var is used to determine the grounding line selection start confirmation, 4 points are pushed forward from the starting point, and the starting point is confirmed as the reference point, and the three-phase current sampling values ​​2T in front of it are stored in the array for standby use.

[0035] S42: Calculation of three-phase current mutation: In a single-phase grounding fault, we assume that the selected reference point is in a stable operating state within 2T time. Then, for the sampling value within T / 2 time after the reference point, the calculated mutation amount of the previous point of the cycle is bound to be close to 0, and the mutation amount of the current point should be much larger than the mutation amount of the point one week ago, and the body will show obvious single-phase grounding transient characteristics. Therefore, for the calculation of the cycle mutation amount, the calculation method of direct subtraction without modulus value is adopted, and the points less than 0 are directly assigned 0 and discarded.

[0036] S43: Get the maximum value of the three-phase current mutation: In the data T / 2 after the reference point, after calculating the week-to-week mutation of the three-phase currents A, B, and C point by point, for the fault within the zone, the mutation of the fault phase is much larger than that of the non-fault phase, while for the fault outside the zone, the magnitude and direction of the mutation of the three-phase current are basically the same. Therefore, after calculating the week-to-week mutation of the three phases at the current point, we can compare them first. If the mutation of the largest phase among the three phases is more than m times that of the other two phases (the typical value of m can be 2). It can be considered as typical fault data within the zone. The week-to-week mutation of the three-phase current at the current point is compared with the maximum week-to-week mutation of each channel, and the larger value of the two is left as the new maximum mutation of this channel. If the set conditions are not met, it is considered that the week-to-week mutation of the three phases at the current point is not typical, so it is discarded, and the calculation of the next point can be exited to finally obtain the maximum week-to-week mutation of each phase.

[0037] For example, the calculation formula for the mutation amount of phase A in S43 is as follows: ; ; ; Where: is the sampling value of phase A current at the current point; is the sampling value of phase A current 1T ago; is the sampling value of phase A current 2T ago; is the modulus of the current point mutation value of phase A current; is the modulus of the point mutation quantity of phase A current 1T ago; is the weekly mutation amount of the current point, when When it is less than 0, Assign the value 0. S44: Select the grounding phase: Compare the maximum values ​​of the cycle-to-cycle mutations of each phase and select the phase corresponding to the maximum value as the grounded phase.

[0038] S5: When the action conditions are met, output alarm or trip signal and phase selection result.

Claims

1. A single-phase grounding fault feature recognition and phase selection method, characterized in that: The steps include: S1: Obtain a zero-sequence voltage mutation amount; the zero-sequence voltage is a self-generated zero-sequence voltage; S2: Obtaining the effective value of the zero-sequence component; the effective value of the zero-sequence component includes the effective value of the self-generated zero-sequence voltage and current or the effective value of the external zero-sequence voltage and current; S3: construct a zero-sequence voltage positive sequence based on the absolute value of the self-generated zero-sequence voltage or the external zero-sequence voltage of each branch of the distribution network line, so as to obtain the first similarity of the zero-sequence voltage positive sequence between two branches, and compare the zero-sequence voltages of the corresponding branches in pairs to obtain the zero-sequence voltage shape difference value to construct the zero-sequence voltage change value of the fault line, and determine whether a branch is grounded and the line selection is started according to the zero-sequence voltage change value; S4: After the start is determined, when the line state is normal operation; according to the preset fixed value, determine whether the zero-sequence voltage mutation or the zero-sequence voltage effective value exceeds the limit to start; calculate the zero-sequence reactive power; determine whether it is an internal fault by the polarity of the zero-sequence reactive power; confirm the start in the case of an internal fault, and start the alarm or trip delay; among which, the zero-sequence voltage effective value is the self-produced zero-sequence voltage effective value or the external zero-sequence voltage effective value; After the start-up, if the line selection status is not running normally, find the reference point from the start-up point, calculate the mutation amount point by point for the three-phase current sampling data of the half sampling cycle length after the reference point, and select the phase with the largest cycle-to-cycle mutation amount as the grounded phase; S5: When the action conditions are met, output alarm or trip signal and phase selection result.

2. A single-phase grounding fault feature identification and phase selection method according to claim 1, characterized in that: In S2, the external zero-sequence voltage and current effective value includes an external zero-sequence current effective value and an external zero-sequence voltage effective value. The external zero-sequence current effective value is directly collected by a zero-sequence current transformer CT, and the external zero-sequence voltage effective value is directly collected by a zero-sequence voltage transformer PT.

3. A single-phase grounding fault feature identification and phase selection method according to claim 1, characterized in that: The self-produced zero-sequence voltage and current effective value in S2 includes the self-produced zero-sequence voltage effective value and the self-produced zero-sequence current effective value; The sampling value of the zero-sequence voltage is obtained by adding the sampling values ​​of the three-phase voltage, and then the sampling value of the zero-sequence voltage is processed by the full-cycle Fourier differential filtering algorithm to obtain the effective value of the self-produced zero-sequence voltage; The sampling value of the zero-sequence current is obtained by adding the sampling values ​​of the three-phase current, and then the sampling value of the zero-sequence current is processed by a full-cycle Fourier differential filtering algorithm to obtain the self-generated zero-sequence current effective value.

4. A single-phase grounding fault feature identification and phase selection method according to claim 1, characterized in that: S3 is specifically: S31: Construct the zero-sequence voltage forward sequence: Obtain the self-generated zero-sequence voltage or external zero-sequence voltage of each branch of the distribution line at the current sampling time, take the absolute value and store it in a two-dimensional array in a loop. Each row of the two-dimensional array represents the absolute value of the zero-sequence voltage of a branch at different sampling times, that is, the positive sequence of the zero-sequence voltage; each column of the two-dimensional array represents the absolute value of the zero-sequence voltage of each branch at the same time; S32: Obtain the first similarity between any two branches in the zero-sequence voltage forward sequence: The similarity between the positive sequences of zero-sequence voltages of any two branches is recorded as the first similarity; S33: Calculate the difference in zero-sequence voltage shape between any two branches: After cyclically storing the two-dimensional array, the difference of the absolute value of the zero-sequence voltage at the same sampling time is calculated based on the zero-sequence voltage of any two branches, and the number of occurrences of each difference value is counted, and the difference value with low frequency is discarded, and the sum of all remaining differences is marked as the zero-sequence voltage shape difference value of the two branches; S34: Calculate the zero-sequence voltage variation value of each branch: expressed as: ； Where: is the zero-sequence voltage variation value of the i-th branch; n is the total number of branches of the distribution line; is the first similarity between the zero-sequence voltages of the i-th branch and the k-th branch; is the difference in zero-sequence voltage shape between the ith branch and the kth branch; To prevent decimals with a denominator of 0; S35: Confirm start: Based on the zero-sequence voltage variation value of each branch, the average value is taken according to the data of the branch. When the zero-sequence voltage variation value of a branch exceeds the limit and the corresponding average value is the largest, it is determined that the line grounding selection is started.

5. A single-phase grounding fault feature identification and phase selection method according to claim 1, characterized in that: In S4, the starting setting value of the zero-sequence voltage effective value takes a typical value of 8V, and the setting value of the zero-sequence voltage mutation amount takes a typical value of 5V.

6. A single-phase grounding fault feature identification and phase selection method according to claim 1, characterized in that: In S4, the zero-sequence reactive power polarity is used to determine whether the fault is an internal fault, specifically: The zero-sequence reactive power is calculated at the moment of grounding. When the zero-sequence reactive power is greater than 0, it is determined as an internal fault. When the zero-sequence reactive power is less than 0, it is determined as an external fault.

7. A single-phase grounding fault feature identification and phase selection method according to claim 1, characterized in that: In step S4, the reference point is found from the starting point, the mutation amount is calculated point by point for the three-phase current sampling data of the half sampling period after the reference point, and the phase with the largest cycle-to-cycle mutation amount among the three phases is selected as the grounded phase. Specifically, the process is as follows: S41: Data storage: After the grounding line selection start confirmation is detected, n sampling points are pushed forward from the start point, and the three-phase current sampling values ​​of the previous 2T are stored in the array for standby use with the sampling point as the reference point, where T represents the sampling period; S42: Calculation of three-phase current mutation: Calculate the three-phase current sampling values ​​of T / 2 length after the reference point point by point; first obtain the current point sampling value, then obtain the sampling value 1T ago and the sampling value 2T ago from the array; calculate the difference between the current point mutation amount and the point mutation amount 1T ago, that is, the cycle mutation amount; S43: Get the maximum value of the three-phase current mutation: In the data T / 2 after the reference point, the cycle-to-cycle mutation of the three-phase currents A, B, and C is calculated point by point, and the maximum cycle-to-cycle mutation of each phase is obtained; S44: Select the grounding phase: Compare the maximum values ​​of the cycle-to-cycle mutations of each phase and select the phase corresponding to the maximum value as the grounded phase.

8. A single-phase grounding fault feature identification and phase selection method according to claim 7, characterized in that: The calculation formula for the mutation amount of phase A in S43 is as follows: ； ； ； Where: is the sampling value of phase A current at the current point; is the sampling value of phase A current 1T ago; is the sampling value of phase A current 2T ago; is the modulus of the current point mutation value of phase A current; is the modulus of the point mutation quantity of phase A current 1T ago; is the weekly mutation amount of the current point, when When it is less than 0, Assign the value 0.

Images