Qualitative module for identifying edge of latent fault of network configuration cable compatible with multiple types of dtu

Abstract

The application discloses a latent fault edge identification and qualification module compatible with multiple DTUs of a distribution network cable, comprising a module measurement data access unit, a constant value configuration unit, a data sampling and preprocessing unit, a fault identification and qualification unit and a signal transmission unit; the module measurement data access unit transforms the loop-in loop-out interval and branch interval of each bus, including modifying the zero sequence current loop, the phase current loop and the voltage loop, and installing a temperature and humidity sensor and a voiceprint sensor, and the current measurement, voltage measurement and voiceprint measurement adopt the same sampling frequency; the constant value configuration unit sets three sets of constant values through self-adaptation or remote configuration. The latent fault edge identification and qualification module compatible with multiple DTUs of the distribution network cable solves the problems of low sampling frequency, high pressure of the main station and high data communication security risk of the existing technology in the background technology.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of distribution network cable fault detection, and particularly relates to a distribution network cable latent fault edge identification and qualitative module compatible with multiple DTUs. BACKGROUND

[0002] The existing distribution network cable latent fault identification technology mainly uses a DTU (distribution terminal unit) to configure a special protection signal to obtain data of the latent fault development to a certain stage, and then realizes comprehensive research and judgment after the data is uploaded to a master station for analysis, so as to preliminarily identify the fault. However, the technical scheme has the following three significant problems:

[0003] 1. The measurement sampling frequency of the national standard DTU is low, and the zero sequence protection setting value is relatively fixed, so it is difficult to effectively obtain the characteristic signal of the initial stage of the latent fault, resulting in untimely fault identification.

[0004] 2. The master station needs to obtain a large amount of waveform and signal data, which is easy to cause communication channel congestion, and the pressure of unified operation of the master station is huge, so that the real-time performance of fault identification cannot be guaranteed.

[0005] 3. The power grid company has strict safety requirements for the communication module of the DTU, so it is impossible to obtain real-time data from the DTU and feed back relevant information to the DTU through conventional data communication means, which limits the application of the fault identification technology.

[0006] Therefore, there is an urgent need for a latent fault edge identification and qualitative module compatible with the existing DTU safe access requirements to solve the above technical problems. SUMMARY

[0007] The main purpose of the application is to provide a distribution network cable latent fault edge identification and qualitative module compatible with multiple DTUs to solve the problems of low sampling frequency, high pressure of the master station and high data communication safety risk of the existing technology proposed in the background technology.

[0008] To achieve the above purpose, the application provides a distribution network cable latent fault edge identification and qualitative module compatible with multiple DTUs, which comprises a module measurement data access unit, a setting value configuration unit, a data sampling and preprocessing unit, a fault identification and qualitative unit and a signal transmission unit.

[0009] The module measurement data access unit reforms the ring-in ring-out interval and branch interval of each busbar.

[0010] The setting value configuration unit sets three sets of setting values through self-adaptation or remote configuration to capture the initial deterioration data characteristics of the latent fault.

[0011] The data sampling and preprocessing unit monitors high-frequency zero sequence current sampling in intervals, and when the zero sequence current sampling of an interval exceeds a predicted value, relevant current and voltage data are intercepted and retained, and whether the current and voltage fluctuation law is a latent fault is identified through a pre-trained feature recognition algorithm;

[0012] The fault identification and qualification unit first locates the insulation deterioration range, then grades the fault severity in combination with relevant parameters, and uploads the location to the edge substation through a wireless network;

[0013] The signal transmission unit transmits the fault severity and classification result to the DTU public telesignaling signal, which is sent to the operation and maintenance personnel for defect elimination after being uploaded to the public master station by the DTU.

[0014] As a further preferred technical solution of the above technical solution, in the module measurement data access unit,

[0015] The current loop uses the original zero sequence CT and phase CT or installs a high-frequency current transformer, and the current loop of the module is connected to the existing or modified measurement loop in the form of expanding terminals in the DTU cabinet, and is connected to AC0 respectively, to realize the access of current measurement data, and the ring-in and ring-out intervals are marked as HW and the branch intervals are marked as FZ during access;

[0016] The voltage loop distinguishes the line voltage by phase and marks it as UCB and UAB, and calculates the time sequence UA and UC with a 30-degree angle difference.

[0017] As a further preferred technical solution of the above technical solution, the three sets of setting values in the setting value configuration unit are:

[0018] Setting value 1 is the setting value calculated based on the length of the line, setting value 2 is the setting value calculated based on the length of the line combined with the longest line of the tie line, and setting value 3 is the setting value calculated based on the edge.

[0019] The adaptive judgment mode is to take setting value 1 if the monitored amplitude is within the preset range of setting value 1, take setting value 2 if the monitored amplitude is within the preset range of setting value 2, and define setting value 3 as the transient zero sequence current amplitude of the line-to-ground multiplied by the sensitivity if the monitored amplitude does not meet the above two conditions.

[0020] As a further preferred technical solution of the above technical solution, the data sampling and preprocessing unit retains the waveform data of the preset interval at the time when the zero sequence current exceeds Ty, identifies whether the current and voltage fluctuation law is a latent fault through a pre-trained feature recognition algorithm, judges the number of zero sequence current exceeding points Cd in the interval and the average zero sequence current amplitude Fi at the time Ty based on the identification, calculates the average angle difference at the time of insulation breakdown, and calculates the simultaneous rate Tsl of the time Ty and the voiceprint mutation.

[0021] The judgment mode is to first judge whether the current Ia, Ic at the Ty moment is distorted, if yes, the current distortion phase is judged as the insulation failure phase Phx, if Ia, Ic are not distorted, the insulation failure is the b phase; Ua, Uc are calculated through the angle difference 30 degrees of Uba, Ubc, and Ub is calculated through the angle difference 150 degrees; the average value Jx of the absolute value of the angle difference between the distorted phase voltage at all Ty moments and the nearest high voltage angle is calculated.

[0022] As a further preferred technical solution of the above technical solution, the insulation deterioration range positioning in the fault identification and qualitative unit is specifically:

[0023] The number of intervals Jg with zero sequence current out of limit at the same moment is judged, if it is 1, it is judged as the bus fault of the station; if it is 2 or more, the Max (Fi) of each interval out of limit is taken as MaxF, each interval MaxF is compared with the preset value (MaxF*0.5) and classified, the power side interval is taken as the reference, and the interval with the nearest MaxF and power side interval (zero sequence current amplitude) is taken as the judgment object, the position of the rear end fault is judged, if the judgment object contains the branch interval FZ, it is judged as the load side fault of the branch interval, if the judgment object only contains the ring network interval HW, it is judged as the load side fault of the main line interval.

[0024] As a further preferred technical solution of the above technical solution, the fault severity grading in the fault identification and qualitative unit is specifically:

[0025] The average value Jx is used for grading, Lv1 level;

[0026] Lv2 level;

[0027] Lv3 level;

[0028] Lv4 level;

[0029] Lv5 level;

[0030] The higher the Lv1-Lv5 level is, the higher the defect elimination priority is, Lv1 is attention, Lv2-Lv3 is carried out in cooperation with power failure work, Lv4-Lv5 suggests emergency power failure for defect elimination;

[0031] And the fault identification and qualitative unit identifies whether there are multiple lines with zero sequence out of limit at the same moment through the wireless communication module under the reasonable coding mode of the data sampling and preprocessing unit, if yes, the discharge line caused by insulation deterioration is judged through the average amplitude Fi, and the line with the largest average amplitude Fi is signaled.

[0032] As a further preferred technical solution of the above technical solution, the signal transmission unit transmits the 3 classifications and the risk levels of Lv1-Lv5 to the DTU public teletransmission through a hard node teletransmission signal.

[0033] As a further preferred technical solution of the above technical solution, the power supply unit obtains power supply from a power supply module with a backup power supply of the DTU, and supplies power to the modules installed in the ring network chamber and the ring network box of the power distribution cable line.

[0034] The present application has the following beneficial effects:

[0035] (1) The module of the present application is compatible with various existing DTUs through secondary circuit modification, can perform high-frequency monitoring on zero sequence current signals, effectively obtains latent fault initial characteristic signals, and solves the problem that the sampling frequency of the existing national standard DTU is not high and it is difficult to obtain initial signals.

[0036] (2) The present application completes fault identification, qualitative analysis and positioning at the edge through adaptive fixed value configuration and edge lightweight algorithm, does not need to upload a large amount of data to the master station for operation, greatly reduces the operation pressure of the master station, avoids communication channel blockage, and guarantees the real-time performance of fault identification.

[0037] (3) The present application transmits the fault judgment result to the DTU through the secondary teletransmission circuit hard node output signal mode, does not need to use data communication means, completely meets the safety access requirements of the power grid company on the DTU communication module, realizes information transmission under the condition of zero data communication, and has no safety risk. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 is a schematic diagram of the present application. DETAILED DESCRIPTION

[0039] The following description is used to disclose the present application so that those skilled in the art can implement the present application. The preferred embodiments in the following description are only used as examples, and other obvious modifications can be thought of by those skilled in the art. The basic principles of the present application defined in the following description can be applied to other embodiments, modifications, improvements, equivalents and other technical solutions without departing from the spirit and scope of the present application.

[0040] In the preferred embodiments of the present application, those skilled in the art should note that the DTU and the like involved in the present application can be regarded as prior art.

[0041] Preferred embodiments.

[0042] The application discloses a network cable latent fault edge identification and qualitative module compatible with multiple DTUs, comprising a module measurement data access unit, a constant value configuration unit, a data sampling and preprocessing unit, a fault identification and qualitative unit and a signal transmission unit.

[0043] The module measurement data access unit transforms the ring-in ring-out interval and branch interval of each bus, including modifying the zero sequence current loop, the phase current loop and the voltage loop, and installing a temperature and humidity sensor and a voiceprint sensor at the station, and the current measurement, voltage measurement and voiceprint measurement adopt the same sampling frequency (the common CT sampling frequency is 8000 Hz, and the sampling frequency is 1 MHz when the high-frequency current transformer is additionally installed);

[0044] The constant value configuration unit sets three sets of constant values for capturing latent fault initial deterioration data characteristics (the constant value purpose is to capture corona discharge, surface discharge, water tree branch and other fault initial deterioration data characteristics, the initial zero sequence current constant value is set as the line length * unit conductor theoretical capacitance current value * sensitivity, and three sets of constant values are set), and the adaptive mode is determined by the zero sequence current monitoring, and the constant value area is selected by identifying the transient zero sequence current amplitude of the non-line transient grounding;

[0045] The data sampling and preprocessing unit monitors the high-frequency zero sequence current sampling in the interval Jg, and when the zero sequence current sampling of an interval exceeds the predicted constant value, the related current and voltage data are intercepted and retained, the current and voltage fluctuation law is identified whether it is a latent fault (sharp wave feature) through a pre-trained feature recognition algorithm, the insulation damage phase, damage times, zero sequence current amplitude and average angle difference are judged based on the identification, and the voiceprint mutation and temperature and humidity conditions are simultaneously synchronized;

[0046] The fault identification and qualitative unit adopts a pre-trained analytic hierarchy process algorithm or a lightweight machine learning decision tree algorithm, first performs insulation deterioration range positioning, then grades the fault severity in combination with related parameters, and uploads the positioning to the edge substation through a wireless network.

[0047] The signal transmission unit transmits the fault severity and classification results to the DTU public telesignaling signal in the form of a secondary telesignaling loop hard node signal, and sends the signal to the operation and maintenance personnel for defect elimination after the signal is sent to the public main station by the DTU.

[0048] Specifically, in the module measurement data access unit,

[0049] Current loop uses the original zero sequence CT and phase CT (KHz level) or installs high-frequency current transformer (MHz level, which can find earlier faults) to measure the current loop. Through the expansion of the terminal in the DTU cabinet, the current loop of the module is connected in series to the existing or modified measurement loop, and AC0 is connected respectively to realize the access of current measurement data, and the access interval is marked as HW, and the branch interval is FZ;

[0050] Voltage loop distinguishes line voltage by phase and marks as UCB, UAB. The time sequence UA and UC is calculated by 30 degree angle difference (the power distribution network generally uses 2 groups or 1 group of line voltage as voltage measurement input).

[0051] More specifically, the three sets of fixed values in the fixed value configuration unit are:

[0052] Fixed value 1 is the fixed value calculated by the length of the line, fixed value 2 is the fixed value calculated by the longest line combined with the contact, and fixed value 3 is the edge calculation fixed value.

[0053] (Due to the security access requirements of the State Grid, the edge device cannot access the distribution automation master station to obtain the operation mode, so an adaptive switching method is adopted) The adaptive judgment method is that if the monitoring amplitude is within the positive and negative preset range (30%) of fixed value 1, fixed value 1 is taken, and if it is within the positive and negative preset range (30%) of fixed value 2, fixed value 2 is taken, if it does not meet the above two conditions, define fixed value 3 as the transient zero sequence current amplitude of the line grounding * sensitivity (sensitivity value range is 1.2-1.5, adjusted according to the aging degree of the line), and enable fixed value 3.

[0054] Further, the data sampling and preprocessing unit retains the waveform data of the preset interval (300ms before and 500ms after) of the zero sequence current overrun Ty moment, identifies whether the current voltage fluctuation rule is a latent fault (sharp wave feature) through a pre-trained feature recognition algorithm, judges the zero sequence current overrun measurement point number Cd (insulation damage number) in the interval, the average zero sequence current amplitude Fi at Ty moment, calculates the average angle difference between the insulation damage moment (zero sequence overrun measurement point) and the maximum voltage moment (90 degrees, 270 degrees) of Ua, Ub, Uc, calculates the simultaneous rate Tsl of Ty moment and voiceprint mutation, Tsl = the number of measurement points with synchronous voiceprint mutation at Ty moment / the total number of measurement points at Ty moment;

[0055] The judgment mode is to first judge whether the current Ia and Ic at the Ty moment are distorted, if yes, the current distortion phase is judged as the insulation failure phase Phx, if Ia and Ic are not distorted, the insulation failure is the b phase; Ua and Uc are calculated through the angle difference of 30 degrees of Uba and Ubc, and Ub is calculated through the angle difference of 150 degrees; the average value Jx of the absolute value of the angle difference between the distorted phase voltage at all Ty moments and the nearest high voltage angle is calculated (and Jg, Phx, Cd, Fi, Jx, the synchronous voiceprint mutation situation Tsl, the synchronous temperature and humidity situation are input to the fault identification and qualitative unit).

[0056] Further, the insulation deterioration range positioning in the fault identification and qualitative unit is specifically:

[0057] The number of intervals Jg with zero sequence current out of limit at the same moment is judged, if it is 1, it is judged that the bus of the station is faulty; if it is 2 or more, the Max (Fi) of each interval out of limit is taken as MaxF, each interval MaxF is compared with a preset value (MaxF*0.5) and classified, the interval on the power side is taken as the reference, and the interval with the nearest MaxF and the interval on the power side (zero sequence current amplitude) is taken as the judgment object, the position of the rear end fault is judged, if the judgment object contains the branch interval FZ, it is judged that the load side of the branch interval is faulty, if the judgment object only contains the ring network interval HW, it is judged that the load side of the main line interval is faulty.

[0058] Preferably, the fault severity grading in the fault identification and qualitative unit is specifically:

[0059] The average value Jx is used for grading, as Lv1 level;

[0060] as Lv2 level;

[0061] as Lv3 level;

[0062] as Lv4 level;

[0063] as Lv5 level;

[0064] The higher the Lv1-Lv5 level is, the higher the defect elimination priority is, Lv1 is attention, Lv2-Lv3 is for cooperation with power outage work, Lv4-Lv5 suggests emergency power outage for defect elimination;

[0065] And the fault identification and qualitative unit, through the wireless communication module, under the reasonable coding mode (such as substation bus-line-station three-layer coding mode) of the data sampling and preprocessing unit, identifies whether there are multiple lines with zero sequence over-limit condition at the same time in the same substation bus, if so, through the average amplitude Fi to judge the discharge line caused by insulation deterioration, and the signal of the line with the largest average amplitude Fi is opened.

[0066] Preferably, the signal transmission unit opens and closes the hard node remote signaling signal of the risk level of 3 categories (bus, branch load side, main line load side) and Lv1-Lv5 to the DTU public remote signaling, and each bus uses multiple self-defined openings to transmit signals (such as signal 1, signal 8 hard node connection represents that there is a Lv5 high-risk signal on the bus).

[0067] Preferably, it further includes a power supply unit, which obtains power supply from the power module with backup power supply of the DTU, and supplies power to the modules installed in the ring network room and ring network box of the power distribution cable line.

[0068] It is worth mentioning that the technical features of the DTU and other technologies involved in the present application should be regarded as prior art, and the specific structure, working principle and possible control method and spatial arrangement method of these technical features can be selected by the person skilled in the art, which should not be regarded as the invention point of the present application, and the present application will not be further expanded and detailed.

[0069] For those skilled in the art, the technical solutions recorded in the foregoing embodiments can still be modified, or some technical features can be replaced, and any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A distribution network cable latent fault end-point identification and qualitative module compatible with multiple types of DTUs, characterized in that, It includes a module measurement data access unit, a setpoint configuration unit, a data sampling and preprocessing unit, a fault identification and characterization unit, and a signal transmission unit; The module measurement data access unit modifies the loop-in / loop-out interval and branch interval of each busbar segment; The setpoint configuration unit sets three sets of setpoints through adaptive or remote configuration to capture the early deterioration data characteristics of latent faults. The three sets of set values ​​in the set value configuration unit are as follows: Fixed value 1 is the fixed value used to calculate the length of this line, fixed value 2 is the fixed value used to calculate the length of this line combined with the longest connecting line, and fixed value 3 is the fixed value used to calculate the end. The adaptive judgment method is as follows: if the monitored amplitude is within the positive and negative preset range of value 1, then value 1 is taken; if it is within the positive and negative preset range of value 2, then value 2 is taken; if the above two conditions are not met, value 3 is defined as the transient zero-sequence current amplitude of this line grounding * sensitivity, and value 3 is enabled. The data sampling and preprocessing unit monitors the high-frequency zero-sequence current sampling at intervals. When the zero-sequence current sampling exceeds the predicted value at a certain interval, the relevant current and voltage data are intercepted and stored. The pre-trained feature recognition algorithm is used to identify whether the current and voltage fluctuation pattern is a latent fault. The fault identification and characterization unit first locates the range of insulation degradation, then classifies the severity of the fault based on relevant parameters, and uploads the information to the side terminal station via a wireless network. The signal transmission unit transmits the fault severity and classification results to the DTU's public remote signaling signal, which is then sent to the public master station by the DTU and sent to the maintenance personnel for troubleshooting.

2. The distribution network cable latent fault end identification and characterization module compatible with multiple types of DTUs according to claim 1, characterized in that, In the module measurement data access unit The current loop uses the existing zero-sequence CT and phase CT or the current measurement loop with added high-frequency current transformer. By expanding the terminals in the DTU cabinet, the current loop of the module is connected in series to the existing or modified measurement loop and connected to AC0 respectively to realize the current measurement data access. When connecting, the loop in and loop out interval is marked as HW and the branch interval is FZ. The voltage loops are marked with line voltages UCB and UAB according to phase, and the timing sequence UA and UC is calculated with a 30-degree angle difference.

3. The distribution network cable latent fault end identification and characterization module compatible with multiple types of DTUs according to claim 2, characterized in that, The data sampling and preprocessing unit retains waveform data of a preset interval at the time when the zero-sequence current exceeds the limit Ty. It identifies whether the current and voltage fluctuation pattern is a latent fault through a pre-trained feature recognition algorithm. Based on the recognition, it determines the number of zero-sequence current exceeding the limit measurement points Cd in the interval, the average zero-sequence current amplitude Fi at the time Ty, calculates the average angle difference at the time of insulation failure, and calculates the simultaneity rate Tsl between the time Ty and the acoustic text change. The determination method is as follows: First, determine whether the currents Ia and Ic at time Ty are distorted. If so, the distorted phase is determined to be the insulation failure phase Phx. If Ia and Ic are not distorted, the insulation failure is phase b. Calculate Ua and Uc using the angle difference of 30 degrees between Uba and Ubc, and calculate Ub using the angle difference of 150 degrees. Calculate the average value Jx of the absolute value of the angle difference between the distorted phase voltages at all times Ty and the nearest high voltage angle.

4. The distribution network cable latent fault end identification and characterization module compatible with multiple types of DTUs according to claim 3, characterized in that, The specific method for locating the insulation degradation range in the fault identification and qualitative unit is as follows: Determine the number of intervals Jg that simultaneously have zero-sequence current exceeding the limit. If there is 1, it is determined to be a busbar fault in this station. If there are 2 or more, take the Max(Fi) of each interval that exceeds the limit as MaxF. Compare the MaxF of each interval with the preset value and classify them. Take the power supply side interval as the benchmark and the interval with the closest MaxF to the power supply side interval as the judgment object. Determine the location of the back-end fault. If the judgment object includes branch interval FZ, it is determined to be a branch interval load side fault. If the judgment object only includes ring network interval HW, it is determined to be a main line interval load side fault.

5. The distribution network cable latent fault end identification and characterization module compatible with multiple types of DTUs according to claim 4, characterized in that, The fault severity classification in the fault identification and qualitative unit is specifically as follows: Classification based on average value Jx Level 1; Level 2; Level 3; Level 4; Level 5; The higher the level (Lv1-Lv5), the higher the priority of defect elimination. Lv1 is "attention", Lv2-Lv3 is to be carried out in conjunction with power outage work, and Lv4-Lv5 is recommended to be carried out by emergency power outage. Furthermore, the fault identification and characterization unit, through the wireless communication module, identifies whether multiple lines on the same substation busbar have zero-sequence over-limit conditions at the same time, under the reasonable encoding method adopted by the data sampling and preprocessing unit. If so, it determines the discharge line caused by insulation deterioration by the average amplitude Fi, and sends a signal to the line with the largest average amplitude Fi.

6. The distribution network cable latent fault end identification and characterization module compatible with multiple types of DTUs according to claim 5, characterized in that, The signal transmission unit sends the three categories and risk levels Lv1-Lv5 to the DTU common remote signaling via hard node remote signaling. Each bus section uses multiple custom outputs for signal transmission.

7. The distribution network cable latent fault identification and characterization module compatible with multiple types of DTUs according to claim 1, characterized in that, It also includes a power supply unit, which obtains power from the power module with backup power supply of the DTU to power the modules installed in the ring network room and ring network box of the power distribution cable line.

Images