A method for improving the efficiency of single-phase ground fault disposal of an electric cable

Abstract

The application provides a method for improving the single-phase grounding fault disposal efficiency of a cable, comprising the following steps: step S1, acquiring a three-phase current signal; step S2, performing feature extraction on the current signal and establishing a three-phase transient current feature map; step S3, performing calculation and judgment through a research and judgment algorithm; and step S4, if the judgment result is a grounding fault, issuing a fault remote signal. The method for improving the single-phase grounding fault disposal efficiency of a cable firstly processes the acquired current signal to establish a three-phase transient current feature map, performs preliminary judgment, then accurately and quickly judges the fault through a research and judgment algorithm, and finally issues a fault remote signal. The method has a wide practical range, high research and judgment accuracy, and rapid fault disposal efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of fault processing, and in particular to a method for improving the efficiency of single-phase grounding fault processing of a cable. BACKGROUND

[0002] Single-phase grounding fault is a common fault type in the operation of 10kV cable lines, accounting for more than 30% of cable faults, and the fault processing time is much longer than that of short circuit fault. At present, most areas use small current grounding mode (non-grounding and arc suppression coil grounding) for cables. After a grounding fault occurs, the voltage of the non-fault phase will rise to the line voltage. Long-time operation with fault can expand to phase-to-phase short circuit, and arc grounding will cause system resonance overvoltage, affecting the safety of the power system. At the same time, the fault current at the grounding point may endanger personal safety and needs to be processed quickly. At present, grounding fault processing mainly relies on three ways: one is manual trial pulling. This method is low in efficiency, causes non-fault line to be powered off, affects power supply reliability, and cannot determine the fault section after successful trial pulling. The second is to rely on the alarm of the line selection device in the substation, but the accuracy rate of different substations is different, and it cannot further shorten the fault section. The third is to rely on the zero sequence overcurrent judgment function of the existing distribution automation DTU, but false negatives often occur, and the accuracy rate is generally insufficient. The single-phase grounding fault cannot be accurately judged, resulting in low efficiency of fault processing, affecting the operation and maintenance of the distribution network, and reducing the power supply reliability.

[0003] A method for positioning single-phase grounding fault of distribution network is disclosed in Chinese patent document CN104049177B, which relates to the technical field of distribution network and solves the technical problem of reducing the cost of single-phase grounding fault identification and positioning. The method divides each outgoing line of the distribution network bus into multiple outgoing line sections, then simulates single-phase grounding fault for each outgoing line section one by one under non-fault condition, calculates the bus side negative sequence current fault simulation value and bus side zero sequence current fault simulation value of the outgoing line at the bus outlet when each outgoing line section simulates single-phase grounding fault; when single-phase grounding fault occurs in the distribution network, all outgoing lines of the distribution network bus are judged one by one according to the simulation value under non-fault condition, to realize the identification and positioning of single-phase grounding fault. The method provided by the present application is suitable for positioning single-phase grounding fault of distribution network. However, it does not involve a method for improving the efficiency of single-phase grounding fault and a specific judgment method. SUMMARY

[0004] The application solves the problems of low accuracy of steady-state research and disposal efficiency of single-phase grounding fault of power distribution automation, and provides a method for improving the disposal efficiency of cable single-phase grounding fault, which firstly acquires current signals, processes the current signals, establishes a three-phase transient current feature map, preliminarily judges according to the three-phase transient current feature map, then accurately and quickly judges the fault through a research algorithm, and finally sends a fault remote signal.

[0005] In order to achieve the above-mentioned purpose, the application adopts the following technical scheme: a method for improving the disposal efficiency of cable single-phase grounding fault, comprising the following steps:

[0006] S1, acquiring a three-phase current signal of a line;

[0007] S2, extracting features of the current signal and establishing a three-phase transient current feature map;

[0008] S3, calculating and judging through a research algorithm;

[0009] S4, if it is determined that there is a grounding fault, sending a fault remote signal.

[0010] In the application, features are extracted from the acquired three-phase current signal of the line, a three-phase transient current feature map is established according to the characteristic quantities, a preliminary judgment can be made according to the map, and then a research algorithm is used to research whether a single-phase grounding fault occurs, and a fault remote signal is output after the fault is determined, to complete the whole process. The application uses a three-phase current transient sudden change variable algorithm to research the fault, and uses the transient sudden change variable algorithm to make up for the shortcomings of low accuracy of steady-state research of existing power distribution automation, is suitable for different types of neutral grounding modes, the current sampling precision should meet 0.5S, has a transient high-speed recording function, uses three-phase current sudden change recording start conditions to extract fault characteristics of the current, uses a three-phase current fault transient sudden change variable research algorithm to research the fault, and outputs a fault remote signal. The application has high research accuracy and rapid fault disposal efficiency, and can be applied to the field of single-phase grounding fault.

[0011] Preferably, the step S3 comprises the following steps:

[0012] S31, the research algorithm is specifically: setting the total network ground capacitance current as I ∑c , the maximum single-line ground capacitance current as I L , multiplying the ratio of the total network capacitance current to the single-line capacitance current by a certain reliability coefficient to obtain a setting parameter:

[0013] K=(I ∑c / I L )*0.8;

[0014] S32, calculate the amplitude ratio P of the ground phase fault transient component to the intact phase;

[0015] S33, when P>K, and the fault phase is opposite to the non-fault phase, determine that it is a ground fault.

[0016] In the application, when a single-phase ground fault occurs in a small current grounding system, there is a ground capacitance current and a fault point ground current (the sum of the ground capacitance currents of non-fault lines) in the fault phase transient characteristic component of the fault line, and only a ground capacitance current in the three-phase transient characteristic current of the non-fault phase of the fault line and the non-fault line. In the judgment process, the ratio of the amplitude of the ground phase fault transient component to the amplitude of the intact phase is compared with the setting parameter, and the judgment accuracy is high.

[0017] As preferred, the step S2 is specifically: after the current signal is processed by the Bartlett window function, each feature quantity is extracted and stored, and a three-phase transient current feature map is established according to the feature quantity.

[0018] In the application, it can be obtained from the three-phase transient current feature map that in the non-fault line, the three-phase transient current amplitudes and phases are basically consistent, when a ground fault occurs in a certain phase of the three-phase, the transient current amplitude is the largest, the phase is opposite to the other two non-fault phases, and the transient amplitudes of the other two phases are smaller and the phases are the same. According to the three-phase transient current feature map, a preliminary judgment can be made.

[0019] As preferred, the line three-phase current signal in the step S1 is acquired by using a portable detection device or a detection device installed at a fixed point.

[0020] In the application, the acquisition mode of the three-phase current signal is determined according to the local power grid arrangement condition, and the flexibility is high.

[0021] As preferred, in the step S4, the fault remote signal is sent out, the fault current signal is encrypted and converted, and then sent to the power grid control center, and a historical fault current signal model is established.

[0022] In the application, the fault current signal data is converted into a binary signal after time domain transformation, and is encrypted by inserting a random time stamp, and then is decrypted after being transmitted to the power grid control center, so that the safety of the fault data can be ensured.

[0023] As preferred, the historical fault current signal model is specifically: training samples are generated according to the extracted current feature quantity, and a historical fault current signal model is established by using a deep learning method, and whether the current signal sent to the power grid control center is similar to the historical fault current signal model can be selected for fitting comparison, so as to provide a reference for fault processing.

[0024] In the application, when the current signal sent to the power grid control center is compared with the historical fault current signal model with deep learning function, if the error is within a certain range, the comparison is successful, and if the error exceeds a certain range, the steps S1-S4 of the application need to be re-operated. The false judgment of small probability is prevented from affecting the result.

[0025] The beneficial effects of the application are: firstly, the current signal is collected and processed to establish a three-phase transient current feature map for preliminary judgment, then the fault is accurately and quickly judged through the research and judgment algorithm, and finally the fault remote signaling signal is sent. The method has wide practical range, high research and judgment accuracy, and rapid fault disposal efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 is the flow chart of the application;

[0027] Figure 2 is the research and judgment algorithm logic diagram of the application;

[0028] Figure 3 is the non-fault line transient current feature map of the application;

[0029] Figure 4 is the fault line transient current feature map of the application. DETAILED DESCRIPTION

[0030] Embodiment:

[0031] The embodiment provides a method for improving the single-phase grounding fault disposal efficiency of a cable, referring to Figure 1 and Figure 2 , comprising the following steps: step S1, acquiring a three-phase current signal of a line; in this step, the current signal can be collected by a fixed-point installation detection device and a portable detection device, which are confirmed according to the local environment.

[0032] Step S2, feature extraction after processing the current signal, and establishment of a three-phase transient current feature map; in this step, the collected current signal is first processed by a Bartlett window function, and then the feature quantity required for each fault research and judgment is extracted and stored, and the three-phase transient current feature map is created based on the feature quantity.

[0033] Then, step S3, calculation and judgment by the research and judgment algorithm; the specific steps include the following three steps: step S31, first, the setting parameters are calculated according to the maximum value of the total network ground capacitance current and the single-line ground capacitance current:

[0034] K=(I ∑c / I L )*0.8;

[0035] In the above formula, I∑c represents the total network-to-ground capacitance current, I L represents the single line-to-ground capacitance current maximum value.

[0036] Step S32, P is obtained, P is the ratio of the amplitude of the ground phase fault transient component and the amplitude of the healthy phase.

[0037] Step S33, P and K are compared in size, if P>K, and the fault phase and the non-fault phase are opposite, it is judged to be a single-phase ground fault.

[0038] Step S4, if it is judged to be a ground fault, a fault remote signal is sent out, in this step, the fault current signal is encrypted at the same time, after encryption and transformation, it is transmitted to the power grid control center, after decryption, the current characteristic quantity extracted from the current signal is used to generate a training sample, and a historical fault current signal model is created by using deep learning. Among them, the current signal sent to the power grid control center and the historical fault current signal model are compared in similarity, and a certain error range is set.

[0039] In the application, first, the collected line three-phase current signal is used for feature extraction, a three-phase transient current feature image is established according to the characteristic quantity, a preliminary judgment can be made according to the image, and then whether a single-phase ground fault occurs is judged according to the judgment algorithm, and the whole process is completed after the fault remote signal is output if it is judged to be a fault. The application uses three-phase current transient sudden change variable algorithm to judge the fault, and uses transient sudden change variable algorithm to judge to make up for the shortcomings of the existing distribution automation steady-state judgment accuracy, which is suitable for different types of neutral point grounding mode, the current sampling precision should meet 0.5S, has transient high-speed recording function, uses three-phase current sudden change recording starting condition to extract the fault characteristics of the current, uses three-phase current fault transient sudden change variable judgment algorithm to judge the fault, and outputs the fault remote signal. The application has high judgment accuracy, rapid fault disposal efficiency, and can be applied to the field of single-phase ground fault.

[0040] In the application, when a single-phase ground fault occurs in a small current grounding system, there is a ground capacitance current and a fault point-to-ground current (the sum of the non-fault line-to-ground capacitance current) in the fault phase transient characteristic component of the fault line, and only the ground capacitance current in the three-phase transient characteristic current of the non-fault phase of the fault line and the non-fault line. In the judgment process, the ratio of the amplitude of the ground phase fault transient component and the amplitude of the healthy phase is compared with the setting parameter, the judgment accuracy is high, and the value range of the setting parameter is approximately between 3-8.

[0041] Reference Figure 3 and Figure 4From the three-phase transient current feature map, in the non-fault line, the three-phase transient current amplitude and phase are basically consistent, when a ground fault occurs in one phase of the three-phase, in the figure, the A phase has a ground fault, it can be obviously seen that the transient current amplitude is the largest, the phase is opposite to the remaining two non-fault B phase and C phase, the transient amplitudes of the remaining two phases are smaller, and the phases are the same. According to the three-phase transient current feature map, preliminary judgment can be made.

[0042] In the application, the acquisition mode of the three-phase current signal is determined according to the local power grid arrangement condition, and the flexibility is high.

[0043] In the application, the fault current signal data is converted into a binary signal after time domain transformation, and is encrypted by cooperating with the random insertion of a time stamp, and is decrypted after being transmitted to the power grid control center, so that the safety of the fault data can be ensured.

[0044] In the application, when the current signal transmitted to the power grid control center is compared with the historical fault current signal model with the deep learning function, if the error is within a certain range, the comparison is successful, if the error exceeds a certain range, the steps S1-S4 of the application need to be re-operated. The misjudgment with a small probability is prevented from affecting the result.

[0045] The principle of the three-phase current fault transient mutation variable research and judgment algorithm in the application is as follows: when the normal power distribution network is operated, the composition of each line phase current is:

[0046] I=I L +I C

[0047] Among them, I L is the phase load current of the line, I C is the line-to-ground capacitance current;

[0048] When a single-phase ground fault occurs in the system, since the amplitude and phase of the three-phase line current remain unchanged after the fault, the phase load current of the transmission line does not change before and after the fault, that is:

[0049] I L =I’ L ,

[0050] Then for each phase of the non-fault line, the non-fault phase of the fault line and the downstream section of the fault phase of the fault line, the difference of the phase current before and after the fault (that is, the phase current mutation variable) is mainly reflected in the change of the ground capacitance current, that is:

[0051]

[0052] Among them, Δ u is the phase voltage mutation variable of the line, and ΔI cC is the ground capacitance value of the line for the phase; and for the upstream section of the fault line, the phase current jump variable consists of the change of the ground capacitance current itself and the ground current flowing through the grounding point, i.e.:

[0053]

[0054] wherein I K is the ground current,

[0055] In combination with the characteristics of the phase voltage jump variable, the characteristics of the phase current jump variable of the single-phase grounding fault of the distribution network can be summarized as follows:

[0056] 1. For the non-fault line and the downstream of the fault line, since the phase voltage jump variables of the three phases are consistent, the corresponding phase current jump variables of each phase are also consistent, i.e. the amplitudes of the three-phase current jump variables are equal and the waveforms are consistent;

[0057] 2. For the upstream of the fault line, since the ground current flows to the ground through the fault point on the fault phase, the phase current jump variables of the fault phase and the other two non-fault phases are significantly different.

[0058] In summary, the positioning of the single-phase grounding fault of the distribution network is performed by using the above phase current jump variable characteristics, i.e. comparing the three-phase phase current jump variables of each line, the three-phase phase current jump variables of the downstream section of the fault point or the non-fault line are almost consistent, and the phase current jump variable of one phase and the other two phases of the upstream section of the fault line are significantly different, thereby realizing the section positioning of the single-phase grounding fault.

[0059] The above embodiments are further elaboration and description of the present application, so as to facilitate understanding, and are not any limitation of the present application, and any modification, equivalent replacement and improvement within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for improving the efficiency of handling single-phase grounding faults in cables, characterized in that, Includes the following steps: S1, acquire the three-phase current signal of the line; S2, extract features from the current signal and establish a three-phase transient current feature map; S3 is calculated and judged through an analysis algorithm; the ratio of the transient component amplitude of the grounded phase fault to the amplitude of the healthy phase is compared with the setting parameters. If the transient component amplitude of the grounded phase fault is greater than the setting parameters, it is judged as a ground fault; the setting parameters are: the ratio of the total network capacitive current to the single-line capacitive current, multiplied by a reliability coefficient; S3 includes: The analysis algorithm is as follows: The total network-to-ground capacitance current is set as... The maximum single-line ground capacitance current is The setting parameters are obtained by multiplying the ratio of the total network capacitor current to the single-line capacitor current by the reliability coefficient. Specifically, the setting parameters... K equal Divide by Multiply by 0.8, and the reliability coefficient is 0.8; Calculate the ratio of the transient component amplitude of the grounded phase fault to the amplitude of the healthy phase. P ; when P > K Furthermore, if the phase of the faulty phase is opposite to that of the non-faulty phase, it is determined to be a ground fault. S4. If the fault is determined to be a ground fault, a fault remote signaling is issued. At the same time as issuing the fault remote signaling, the fault current signal is encrypted and converted before being sent to the power grid control center, and a historical fault current signal model is established.

2. The method for improving the efficiency of handling single-phase grounding faults in cables according to claim 1, characterized in that, Step S2 specifically involves: processing the current signal using a Bartlett window function, extracting and storing various feature quantities, and establishing a three-phase transient current feature map based on the feature quantities.

3. The method for improving the efficiency of handling single-phase grounding faults in cables according to claim 1, characterized in that, The acquisition of the three-phase current signal of the line in step S1 is performed using a portable detection device.

4. The method for improving the efficiency of handling single-phase grounding faults in cables according to claim 1, characterized in that, The acquisition of the three-phase current signal of the line in step S1 is achieved using a detection device installed at a fixed point.

5. The method for improving the efficiency of handling single-phase grounding faults in cables according to claim 1, characterized in that, In step S4, the fault current signal data is converted into a binary signal after time-domain transformation, encrypted with random insertion of timestamps, and then decrypted after being transmitted to the power grid control center.

6. The method for improving the efficiency of handling single-phase grounding faults in cables according to claim 5, characterized in that, The historical fault current signal model is specifically designed as follows: training samples are generated based on the extracted current features, and a historical fault current signal model is established using deep learning. The current signal sent to the power grid control center can be selected to be compared with the historical fault current signal model for similarity fitting, providing a reference for fault handling.

7. A method for improving the efficiency of handling single-phase grounding faults in cables according to claim 6, characterized in that, When comparing the current signal sent to the power grid control center with the historical fault current signal model with deep learning function, the comparison is successful if the error is within a certain range. If the error exceeds a certain range, the steps S1-S4 of this invention need to be repeated.

Images