CN118362834B - Power grid fault judging method and system based on recording source end data mining - Google Patents

Abstract

The invention discloses a power grid fault judging method and system based on recording source end data mining, which relate to the technical field of power system relay protection and comprise the following steps: loading a recording data file, naming and initializing key element value variables of fault time sequence characteristics; determining a fault starting point, and judging and determining the fault starting point; determining a fault removal point, and judging whether the fault removal point is determined; determining a reclosing action point, and judging whether the reclosing action point is determined; determining a secondary failure starting point, and judging whether the secondary failure starting point is determined; and determining a secondary failure removal point, and judging whether the secondary failure removal point is determined. The power grid fault judging method based on the recording source end data mining provided by the invention can complete the whole process analysis of the power grid fault only by the basic sampling value information of the recording data, complete the rapid analysis of the power grid fault by fully mining the most basic information of the recording data, has no use scene and threshold limit, and has better low coupling property.

Description

Power grid fault judging method and system based on recording source end data mining

Technical Field

The invention relates to the technical field of relay protection of power systems, in particular to a power grid fault judging method and system based on recording source end data mining.

Background

When disturbance or fault occurs to the high-voltage-class power grid, the wave recording of the relevant relay protection equipment is triggered, and the wave recording data is uploaded to a master station system of the regulation and control mechanism in a file form through a dispatching data network. The method for judging the power grid faults by the master station system through the recording data comprises two methods: firstly, judging by means of position information of a primary disconnecting link and a circuit breaker and performing time sequence comparison verification by referring to fault time and recording data; and secondly, by means of a whole-network unified model, starting and ranging fixed value parameters of relay protection equipment for generating recorded wave data are called to carry out fault whole-process inversion reasoning. Both methods need to use a unified platform and a full-network model, so that the threshold for maintaining the unified platform and the full-network model is high, and the use scene of the recorded wave data is limited.

The storage format of the recording data file conforms to the COMTRADE (Common Format For TRANSIENT DATA Exchange, universal format for transient data Exchange of electric power system) standard, the current relay protection equipment requires 100% support of the COMTRADE standard, the initial version of the standard is IEEE Std C37.111-1991（"IEEE Standard Common Format For Transient Data Exchange（COMTRADE） For Power Systems"）, formulated and implemented in 1991, which defines analog sampling values and channel names for recording voltage/current information, state quantity sampling values and channel names for recording position information of protection action/primary knife gate and breaker, and serial numbers and corresponding moments of each sampling value, there is no requirement of network communication transmission and scheduling system analysis at the moment, so that the content defined by IEEE Std C37.111-1991 belongs to the simplest basic information. After the standard is introduced and started in the power industry of China, a plurality of standards are changed through expansion for a plurality of times in the past, and the current use is that the standard of 2013 edition is firstly expanded and defined on the basis of being compatible with IEEE Std C37.111-1991 in the aspect of releasing the general technical condition (DL/T553-2013) of a dynamic recording device of a power system in 2013, so that the recording data has the inversion reasoning condition of the whole fault process. The problems with the evolution described above are: 1) The wave recording data before 2013 does not have the unified model of the whole network and the parameter information of the ranging constant value; 2) The level of each manufacturer executing the new standard is different, so that the acquisition of the whole network unified model and the ranging fixed value parameter information is incomplete or inaccurate, and the accuracy and the reliability of the whole process inversion reasoning on the power grid faults are affected.

In summary, the current COMTRADE standard version is more, and the expanded information is diversified, so that the difference of the recording data formats actually conforming to the COMTRADE is large, and the compatibility is difficult. However, all versions of COMTRADE record data completely support IEEE Std C37.111-1991, that is, the record format of the most basic sampling value, channel name and sampling time information in the record data is clear and definite and unchanged, and the whole process of power grid fault analysis is carried out by mining source end data of the COMTRADE which follows the IEEE Std C37.111-1991, so that the method is the most general and most expected means.

Disclosure of Invention

The present invention has been made in view of the above-described problems.

Therefore, the technical problems solved by the invention are as follows: the method has the advantages that the judgment of the power grid fault is completed rapidly by excavating the most source voltage/current sampling value information of the IEEE Std C37.111-1991 part in the wave recording data, the time sequence characteristic type of the power grid fault and the key element value in the characteristic are determined, the judgment process does not need to confirm state information such as a knife switch and a breaker position, and the like, does not depend on inversion reasoning of any starting and ranging constant value parameters, and does not need to rely on the support of a unified platform and a whole network model. The method is not only suitable for current data, but also can trace back to any version of historical data in COMTRADE format in 1991, and has the characteristics of simplicity, practicability, reliability and universality.

In order to solve the technical problems, the invention provides the following technical scheme: a power grid fault judging method based on recording source end data mining comprises the following steps:

Loading a recording data file, naming and initializing key element value variables of fault time sequence characteristics; determining a fault starting point based on the key element value variable, determining a fault clearing point after determining the fault starting point, and judging whether the fault clearing point is determined or not; after determining the fault removal point, determining a reclosing action point, and judging whether the reclosing action point is determined; after determining the reclosing action point, determining a secondary fault starting point, and judging whether the secondary fault starting point is determined; after determining the secondary fault starting point, determining a secondary fault clearing point, and judging whether the secondary fault clearing point is determined; and determining the type of the fault time sequence characteristic and the corresponding key element based on the reclosing action point, the re-fault starting point and the judging result of the re-fault cutting point.

As a preferable scheme of the power grid fault judging method based on the recording source end data mining, the invention comprises the following steps: the loading recording data file comprises the recording data which follows the COMTRADE standard, the loading recording data obtains voltage/current sampling values and corresponding time information of each sampling value in the power grid fault process, and key element values of fault time sequence characteristics are determined;

the key element values comprise tf representing a fault starting point, tz representing a fault cutting point, tc representing a reclosing action point, tcf representing a secondary fault starting point and tcz representing a secondary fault cutting point;

The initialization includes unified initialization of variables to 0.

As a preferable scheme of the power grid fault judging method based on the recording source end data mining, the invention comprises the following steps: the fault starting point determination comprises calculating power frequency effective values of voltage and current phase by phase, and calculating a power grid fault starting point according to attenuation of the calculated voltage power frequency effective values and increase characteristics of the current power frequency effective values;

the starting point of calculating the power grid fault comprises the steps of setting a failure counter j and initializing j=0;

Sampling voltage and current sampling values phase by phase in the sequence of the voltage and current sampling values A, B, C; when the failure counter j=0, taking an A-phase voltage and current sampling value; when the failure counter j=1, taking a B-phase voltage and current sampling value; when the failure counter j=2, taking a C-phase voltage and current sampling value;

Let su be the voltage sampling value, si be the current sampling value, T be the sampling period, un be the voltage first cycle power frequency effective value, and in be the current first cycle power frequency effective value; traversing si along the T axis by taking tx as a variable, taking the step length as T/4, and initializing tx=0;

initializing tf=0, calculating a fault starting point tf based on tx;

Judging whether the determined mutation point is positioned behind the current tx point, wherein if tf is more than tx, letting uf be the voltage power frequency effective value at tf moment, if be the current power frequency effective value at tf moment, if5T be the maximum power frequency effective value of five cycle currents after tf moment;

judging whether a fault starting point prediction condition is met or not, wherein the fault starting point prediction condition comprises a first fault starting point prediction condition, a second fault starting point prediction condition, a third fault starting point prediction condition and a fourth fault starting point prediction condition, and the first fault starting point prediction condition is if >1.15 x in and uf <0.9 x un; the second failure start point prediction condition is if >1.3 x in and uf <0.95 x un; the third failure start point prediction condition is if >1.5 in and if > 0.5A; the fourth failure start point prediction condition is if >2.5 if 5T;

if the first fault starting point prediction condition, the second fault starting point prediction condition and the third fault starting point prediction condition meet 1 or more and meet the fourth fault starting point prediction condition, determining the mutation point tf as a fault starting point;

If the first fault starting point prediction condition, the second fault starting point prediction condition and the third fault starting point prediction condition meet 1 or more, but do not meet the fourth fault starting point prediction condition or do not meet the first fault starting point prediction condition, the second fault starting point prediction condition and the third fault starting point prediction condition, and the tx shifts to the right by T/4;

if tf is less than or equal to tx, tx is shifted right by T/4;

When tx moves by T/4 to the right, judging whether tx+2T after the right movement exceeds the end point of the sampling curve, and if so, judging that the failure counter +1 is expressed as j=j+1; judging whether the failure times reach 3 times or not, wherein j=3; if the fault starting point is not determined for 3 times, tf=0 is set, and if the fault starting point is not determined for 3 times, the voltage and current sampling values are sampled phase by phase in the sequence of A, B, C of the voltage and current sampling values again;

if the sampling curve end point is not exceeded, re-initializing tf=0, and calculating a fault starting point tf based on tx;

Searching a mutation point based on the current sampling value, calculating a power frequency effective value of voltage/current based on the mutation point position, and analyzing whether the power frequency effective value accords with fault characteristics;

When judging and determining a fault starting point, carrying out a fault removal point determining process;

If the fault starting point is not determined, the process is stopped in a failure mode.

As a preferable scheme of the power grid fault judging method based on the recording source end data mining, the invention comprises the following steps: determining a fault removal point comprises enabling si to be a current sampling value, T to be a sampling period, in to be a current first-cycle power frequency effective value, tf to be a fault starting point, traversing si along a T axis by taking tz as a variable, and initializing tz = tf;

Let iz be the effective value of the current power frequency at the tz moment, and iz5T be the effective value of the minimum power frequency of the five-cycle current after the tz moment;

Judging whether a fault clearing point prediction condition is met or not, wherein the fault starting point prediction condition comprises a first fault clearing point prediction condition, a second fault clearing point prediction condition, a third fault clearing point prediction condition and a fourth fault clearing point prediction condition, and the first fault clearing point prediction condition is that iz <0.01 x iz5T and iz <0.1A; the second fault cut point prediction condition is iz <0.5 x in and in <0.1A; the third fault clearing point prediction condition is iz <0.003A; the fourth fault cut-off point prediction condition is iz5T <0.003A;

If the first fault clearing point prediction condition, the second fault clearing point prediction condition and the third fault clearing point prediction condition meet 1 or more and meet the fourth fault clearing point prediction condition, determining tz as a fault clearing point;

If the first fault clearing point prediction condition, the second fault clearing point prediction condition and the third fault clearing point prediction condition meet 1 or more, but do not meet the fourth fault clearing point prediction condition or do not meet the first fault clearing point prediction condition, the second fault clearing point prediction condition and the third fault clearing point prediction condition, and the tz moves right by T/4; judging whether the tz+T after the right shift exceeds the end point of the sampling curve, if so, judging that the fault cut point is not determined, setting tz=0, and exiting in a failure mode.

As a preferable scheme of the power grid fault judging method based on the recording source end data mining, the invention comprises the following steps: the step of determining the reclosing action point comprises the steps of calculating the ratio of the current at the reclosing action reference point position to the normal load current before the fault and the multiple of the minimum current after the fault is removed, and performing feature analysis by combining the ratio and the multiple;

If the reclosing action point cannot be confirmed, confirming that the time sequence characteristic of the fault is class A, outputting tf and tz values and successfully exiting; if the reclosing operation point is confirmed, determining a re-fault starting point.

As a preferable scheme of the power grid fault judging method based on the recording source end data mining, the invention comprises the following steps: the determining of the secondary fault starting point comprises the steps of calculating the multiple of the current of the secondary fault reference point position and the normal load current before the fault and the ratio of the current to the maximum fault current, and performing feature analysis through the combination of the multiple and the ratio;

If the secondary fault starting point cannot be determined, confirming that the time sequence characteristic of the fault is class B, outputting values tf, tz and tc and successfully exiting; and if the secondary failure starting point is determined, determining a secondary failure cutting point.

As a preferable scheme of the power grid fault judging method based on the recording source end data mining, the invention comprises the following steps: determining the secondary fault removal point comprises determining that the time sequence characteristic of the fault is class B if the secondary fault removal point cannot be determined, outputting values tf, tz and tc and successfully exiting;

if the secondary fault removal point is determined, confirming that the time sequence characteristic of the fault is class C, outputting tf, tz, tc, tcf, tcz values and successfully exiting.

The invention also aims to provide a power grid fault judging system based on recording source end data mining, which can complete rapid analysis of power grid faults by fully mining information on the basis of recording data, has no use scene and threshold limit, has good low coupling, can independently run in a complete set, can be deployed to other platforms to bear the function of a fault analysis module, and meets the rapid diagnosis and analysis requirements of power grid faults of all levels of regulation mechanisms and power generation group centralized control centers.

In order to solve the technical problems, the invention provides the following technical scheme: a power grid fault judging system based on recording source end data mining comprises: the device comprises a data processing module, a fault positioning module, a reclosing identification module and a fault judging module; the data processing module is used for loading the recording data file, performing preprocessing operation, cleaning and formatting data, and initializing key element value variables of fault time sequence characteristics; the fault positioning module determines a starting point and an ending point of a fault by using an algorithm according to the preprocessing data of the data processing module, analyzes the current and voltage time sequence characteristics of the power grid, and identifies the accurate position of the fault; the reclosing identification module is used for identifying the moment of the reclosing operation of the power grid, and judging the action time of the reclosing and the starting point and the ending point of the subsequent secondary fault; the fault judging module is used for carrying out final fault analysis and judgment, providing fault type and cause analysis and providing preventive measures according to analysis results.

A computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the steps of the grid fault determination method based on record wave source end data mining as described above when the computer program is executed.

A computer readable storage medium having stored thereon a computer program which when executed by a processor implements the steps of a grid fault determination method based on record wave source side data mining as described above.

The invention has the beneficial effects that: the power grid fault judging method based on the recording source end data mining provided by the invention has a wide application range. The whole process analysis of the power grid fault can be completed only by the basic sampling value information of the recording data without depending on a unified platform and a full-network model support and without additional fixed value parameters, the power grid fault time sequence characteristic type and the corresponding key element value are given, and the requirements of all voltage class power grid systems of 110kV and above are met.

The applicability is strong. The method is applicable to all wave recording data which follow COMTRADE in 1991, including but not limited to wave recording data from a protection device, a concentrated/dispersed fault wave recording device and a traveling wave power frequency device.

Is convenient and quick. The rapid analysis of the power grid faults is completed by fully excavating the most basic information of the recording data, the use scene and threshold limit are avoided, the system has good low coupling, the system can independently run in a complete set, and the system can be deployed to other platforms to bear the functions of fault analysis modules, so that the rapid diagnosis and analysis requirements of the power grid faults of all levels of regulation and control mechanisms and power generation group centralized control centers are met.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 is an overall flowchart of a power grid fault determination method based on recording source end data mining according to an embodiment of the present invention.

Fig. 2 is a timing characteristic and key element diagram of a power grid fault according to a power grid fault determination method based on recording source end data mining according to an embodiment of the present invention.

Fig. 3 is a diagram of a determined fault start point of a power grid fault determination method based on recording source end data mining according to an embodiment of the present invention.

Fig. 4 is a deterministic fault cut-off diagram of a power grid fault determination method based on recording source end data mining according to an embodiment of the present invention.

Fig. 5 is a diagram of a determined reclosing action of a method for determining a power grid fault based on recording source end data mining according to an embodiment of the present invention.

Fig. 6 is a diagram of a determination restart fault start point of a power grid fault determination method based on recording source end data mining according to an embodiment of the present invention.

Fig. 7 is a determining secondary fault removal point diagram of a power grid fault determination method based on recording source end data mining according to an embodiment of the present invention.

Fig. 8 is a plot of sudden changes of sampling values based on tx judgment in a method for determining power grid faults based on recording source end data mining according to an embodiment of the present invention.

Fig. 9 is a determination mutation chart of a power grid fault determination method based on recording source end data mining according to an embodiment of the present invention.

Fig. 10 is a test sample screenshot of a power grid fault determination method based on recording source end data mining according to an embodiment of the present invention.

Fig. 11 is a cut-out diagram of an analysis result of a power grid fault determination method based on recording source end data mining according to an embodiment of the present invention.

Detailed Description

So that the manner in which the above recited objects, features and advantages of the present invention can be understood in detail, a more particular description of the invention, briefly summarized above, may be had by reference to the embodiments, some of which are illustrated in the appended drawings. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.

In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways other than those described herein, and persons skilled in the art will readily appreciate that the present invention is not limited to the specific embodiments disclosed below.

Example 1

Referring to fig. 1-9, for an embodiment of the present invention, a method for determining a power grid fault based on recording source end data mining is provided, including:

As shown in fig. 2, after a power grid fault, triggering a related relay protection device to start recording, wherein a recording data format complies with COMTRADE standard, and mainly comprises voltage and current sampling values before and after the power grid fault and corresponding time information, the sampling values are arranged along a time axis to form a waveform curve, and the characteristic meanings of each recording data in fig. 2 are as follows:

T: sampling period; su: a voltage sampling value; si: a current sampling value; tf: a fault starting point; tz: fault removal points; tc: reclosing action points; tcf: a second failure starting point; tcz: a fault removal point again; un: the first cycle power frequency effective value of the voltage; in: the current first cycle power frequency effective value; if: a power frequency effective value of the current at tf moment; and (3) iz: the power frequency effective value of the current at the tz moment; if5T: five-cycle current maximum power frequency effective value after tf moment; iz5T: five-cycle current minimum power frequency effective value after tz moment; ic: a current power frequency effective value at time tc; icf: a current power frequency effective value at the time tcf; icz: and the current power frequency effective value at the time of tcz.

In the above definition, un, in, if, iz, if T, iz and T, ic, icf, icz are both secondary values, the voltage power frequency effective value unit is V (volts), and the current power frequency effective value unit is A (amperes)

The fault time sequence characteristics of the power grid are divided into three types, and are distinguished according to A type/B type/C type, and key elements corresponding to the characteristics are explained as follows.

Class a fault timing feature:

The corresponding process is as follows: fault onset- > fault removal; the corresponding key elements are as follows: tf- > tz is mainly triggered by 220kV voltage class interphase faults, three-phase faults and faults occurring in the power transmission process;

Class B fault timing feature:

The corresponding process is as follows: fault start- > fault removal- > reclosing action; the corresponding key elements are as follows: tf- > tz- > tc, mainly triggered by single-phase faults, 110kV voltage class interphase faults and three-phase faults;

Class C fault timing feature:

The corresponding process is as follows: fault start- > fault removal- > reclosing action- > re-fault removal; the corresponding key elements are as follows: tf- > tz- > tc- > tcf- > tcz, is triggered by a single-phase fault of 220kV and above, a fault reclosing action of 110kV voltage class, or by a fault triggered again after the reclosing action is successful.

According to the method, whether the power grid fault is contained or not is confirmed through directly calculating and analyzing voltage and current sampling value information of the most basic part in the fault recording data file, and the time sequence characteristic type of the power grid fault and the corresponding key element value are further analyzed. The classification definition of the power grid fault time sequence characteristics, the fault process and the relation of corresponding key elements are shown in table 1.

TABLE 1 Fault time sequential characteristic type relationship Table

Fault timing feature type	Failure process	Corresponding key elements
			Class A	Failure onset- > failure excision	tf->tz
Class B	Fault start- > fault removal- > reclosing action	tf->tz->tc
			Class C	Fault start- > fault removal- > reclosing action- > re-fault removal	tf->tz->tc->tcf->tcz

Step 1: and loading a recording data file, naming and initializing key element value variables of fault time sequence characteristics.

The recording data is required to follow the COMTRADE standard, 100% of the current relay protection device in China supports the COMTRADE standard, and the voltage/current sampling value of the IEEE Std C37.111-1991 part, the corresponding serial number, the moment and other information in the standard are defined earliest in 1991, are unchanged all the time, belong to the basic information with the highest certainty, and are suitable for the COMTRADE standard of any subsequent version. Therefore, the loading wave recording data can acquire information such as voltage/current sampling values and corresponding moments of the sampling values in the power grid fault process, and the information is used for analysis and calculation in the subsequent steps.

The key element values of the fault time sequence characteristics to be calculated subsequently are as follows:

(1) Fault starting point. For convenience of description, the variable is named tf

(2) Fault removal points. For convenience of description, the variable is named tz

(3) Reclosing action point. For convenience of description, the variable is named tc

(4) The start point is again failed. For convenience of description, the variable is named tcf

(5) The fault cut point is again broken. For convenience of description, the variable is named tcz

The variables are uniformly initialized to 0.

Step 2: a fault starting point is determined.

The method comprises the steps of calculating power frequency effective values of voltage and current phase by phase, and calculating the starting point of the power grid fault through attenuation of the calculated voltage power frequency effective values and increase characteristics of the current power frequency effective values.

In order to improve analysis efficiency, a mode of searching for a mutation point based on a current sampling value, calculating a power frequency effective value of voltage/current based on the mutation point position, and finally analyzing whether the power frequency effective value accords with fault characteristics is adopted.

The background technology for searching the mutation points based on the current sampling value is shown in fig. 2, and the specific implementation method is shown in a flow chart 9. The method for searching the abrupt change point based on the current sampling value belongs to common means in industry, and is only taken as an example of a means for improving efficiency, and does not take the claims.

It is particularly clear that the step of judging the power frequency effective value of the voltage/current phase by phase is very critical, namely if the voltage sampling value is taken as the A phase, the current sampling value also has to be taken as the A phase; if the voltage sampling value takes the B phase, the current sampling value also needs to take the B phase; if the voltage samples take the C phase, the current samples must also take the C phase. The order of phase-by-phase determination is not critical, and the patent gives the order of the first A phase, then B phase and finally C phase, but does not exclude other non-repeated combinations of orders. Any phase in the sequence can directly and successfully exit after confirming the fault starting point, and the three phases still fail to confirm the fault starting point after all analysis is finished and then fail to exit.

The method of the patent gives a 1/4 sampling period for each time of the selected point step length, and the step length can be adjusted only as an implementation example. If the point selection priority target is to improve the precision, the step size can be reduced; if efficiency is improved, the step size can be increased.

Step 3: it is determined whether the failure start point (tf is greater than 0).

If step 2 fails to confirm the failure starting point, the failure exits. Otherwise, go to step 4.

Step4: a fault cut point is determined.

The effective value of the current power frequency after fault removal is 0, and the actual fault of the power grid is influenced by the induction current of the strong magnetic field, the current after fault removal is difficult to return to 0 immediately, so that the current is allowed to exist briefly at the first time of fault removal but must be small enough in order to ensure the sensitivity of an algorithm when the actual point selection of engineering is carried out. Therefore, the key point of the fault removal and point selection algorithm is as follows:

(1) The current after fault removal should be small enough not to be higher than 0.01 times the maximum fault current;

(2) The current at the fault clearing point must not be higher than 0.1A;

(3) The current at the fault clearing point must not be higher than 0.5 times the normal load current. If the normal load current is too small, the current at the fault cut-off point must not be higher than 0.003A;

(4) On the premise of meeting the three points, the minimum current in 5 cycles after fault removal must not be higher than 0.003A;

The method ensures that the fault current is strictly cut off, the calculated fault cut-off point position is very close to the actual waveform, and the risk of point selection lag caused by tiny tailing phenomena caused by common induced current and other factors is eliminated.

The method of the patent gives a 1/4 sampling period for each time of the selected point step length, and the step length can be adjusted only as an implementation example. If the point selection priority target is to improve the precision, the step size can be reduced; if efficiency is improved, the step size can be increased.

Step 5: it is determined whether the fault cut point (tz) is greater than 0.

If step 4 cannot confirm the fault clearing point, the failure exits. Otherwise, step 6 is entered.

Step 6: and determining a reclosing action point.

The single-phase fault of the power grid can automatically trigger reclosing action, current is recovered, and the amplitude level of the current is recovered from no to the amplitude level before the fault from the current sampling value. The difficulty in determining the reclosing action is that: firstly, the process from the reclosing action to the fault current recovery is longer, and the standard is used for determining that the reclosing action causes larger delay after the equal current is completely recovered; secondly, if the current recovery is insufficient, the reclosing action is confirmed, and erroneous judgment may be caused. The key algorithm provided by the patent is as follows: calculating the ratio of the current at the reclosing action reference point position to the normal load current before the fault and the multiple of the minimum current after the fault is removed, analyzing the ratio by the combination characteristic of the ratio and the multiple, and if the ratio is smaller, the multiple is required to be larger, and meanwhile, the current at the reclosing action reference point position is not lower than 0.003A.

The method ensures that the fault current is in the recovery process, and meanwhile, the given point selection position of the reclosing action point is very close to the actual waveform, and the risk of point selection lag caused by long time required for recovering the current to the normal load level after the common reclosing action is eliminated.

The method of the patent gives a 1/4 sampling period for each time of the selected point step length, and the step length can be adjusted only as an implementation example. If the point selection priority target is to improve the precision, the step size can be reduced; if efficiency is improved, the step size can be increased.

Step 7: it is determined whether or not the reclosing operation point (tc is greater than 0).

And (3) if the reclosing action point cannot be confirmed in the step (6), confirming that the time sequence characteristic of the fault is class A, outputting values tf and tz and successfully exiting. Otherwise, step 8 is entered.

Step 8: a secondary failure start point is determined.

The key point of selecting the starting point of the fault again is to balance the recovery degree of the current and the time of increasing the fault current. The key of the algorithm provided by the patent is as follows: and calculating the multiple of the current at the position of the secondary fault reference point and the normal load current before the fault and the ratio of the current to the maximum fault current, analyzing through the combined characteristics of the multiple and the ratio, and if the multiple is larger, requiring the ratio to be smaller, and meanwhile, the current at the position of the secondary fault reference point is not lower than 0.6 times of the maximum fault current.

The method of the patent gives a 1/4 sampling period for each time of the selected point step length, and the step length can be adjusted only as an implementation example. If the point selection priority target is to improve the precision, the step size can be reduced; if efficiency is improved, the step size can be increased.

Step 9: judging whether to determine the re-failure starting point (whether tcf is greater than 0)

If the starting point of the secondary fault cannot be determined in the step 8, confirming that the time sequence characteristic of the fault is class B, outputting values tf, tz and tc and successfully exiting. Otherwise, step 10 is entered.

Step 10: a secondary fault cut point is determined.

The point selection method of the secondary fault clearing point is similar to that of the fault clearing point, but the time for clearing the secondary fault is shorter in consideration of the protection mobility in the secondary fault, and three phases are cleared simultaneously, so that the current after the secondary fault clearing is only required to be confirmed to be small enough.

The method of the patent gives a 1/4 sampling period for each time of the selected point step length, and the step length can be adjusted only as an implementation example. If the point selection priority target is to improve the precision, the step size can be reduced; if efficiency is improved, the step size can be increased.

Step 11: it is determined whether the re-fault cut point (tcz is greater than 0) is determined.

If the secondary fault removal point cannot be determined in the step 10, confirming that the time sequence characteristic of the fault is class B, outputting values tf, tz and tc and successfully exiting. Otherwise, the timing characteristic of the fault is confirmed as class C, and each value is output tf, tz, tc, tcf, tcz and successfully exited.

As shown in fig. 3, for determining the fault starting point, the specific steps are as follows:

Step 2.1: setting a failure counter j, initializing j=0;

Step 2.2: the voltage and current samples are taken phase by phase in the order of their respective A, B, C samples. When the failure counter j=0, taking an A-phase voltage and current sampling value; when the failure counter j=1, taking a B-phase voltage and current sampling value; when the failure counter j=2, taking a C-phase voltage and current sampling value;

step 2.3: let su be the voltage sampling value, si be the current sampling value, T be the sampling period, un be the voltage first cycle power frequency effective value, and in be the current first cycle power frequency effective value. Traversing si along the T axis with tx as a variable (step size is T/4), initializing tx=0;

Step 2.4: initializing tf=0, calculating a fault starting point tf based on tx;

step 2.5: judging whether the mutation point determined in the process F is located after the current tx point, and indicating that tf > tx? If yes, enter step 2.6; otherwise, enter step 2.8;

Step 2.6: let uf be the tf moment voltage power frequency effective value, if be the tf moment current power frequency effective value, if5T be the tf moment after five cycle current maximum power frequency effective value;

Step 2.7: judging whether the following 4 conditions are satisfied: ① (if >1.15 in) and (uf <0.9 un); ② (if >1.3 in) and (uf <0.95 un); ③ (if >1.5 in) and (if > 0.5A); ④ (if >2.5 if 5T);

if the conditions ①②③ meet 1 or more and the conditions ④ are met, the step 2.13 is entered;

If condition ①②③ is satisfied for 1 or more, but condition ④ is not satisfied; or none of conditions ①②③ are satisfied; step 2.8 is entered;

Step 2.8: tx right shift by T/4;

Step 2.9: determine if the right shifted tx+2t exceeds the sampling curve endpoint? If yes, go to step 2.10; otherwise, returning to the step 2.4;

step 2.10: failure counter +1, denoted j=j+1;

step 2.11: judging whether the failure times reach 3 times or not, wherein j=3; if yes, go to step 2.12; otherwise, returning to the step 2.2;

Step 2.12: a fault starting point is not determined, tf=0 is set, and step 2.14 is entered;

step 2.13: determining the mutation point tf as a fault starting point, and entering step 2.14;

step 2.14: outputting tf to the main flow;

step 2.15: and exiting the process and entering the next step of the main process.

As shown in fig. 4, the specific steps for determining the fault removal point are as follows:

Step 4.1: let si be the current sampling value, T be the sampling period, in be the current first cycle power frequency effective value, tf be the fault starting point, take tz as the variable to traverse si along the T axis, initialize tz=tf;

Step 4.2: let iz be the effective value of the current power frequency at the tz moment, and iz5T be the effective value of the minimum power frequency of the five-cycle current after the tz moment;

Step 4.3: judging whether the following 4 conditions are satisfied: ① (iz <0.01 x iz 5T) and (iz < 0.1A); ② (iz <0.5 x in) and (in < 0.1A); ③iz<0.003A );④ (iz 5T < 0.003A);

If the conditions ①②③ meet 1 or more and the conditions ④ are met, the step 4.7 is entered;

If condition ①②③ is satisfied for 1 or more, but condition ④ is not satisfied; or none of conditions ①②③ are satisfied; step 4.4 is entered;

Step 4.4: tz right shift by T/4;

Step 4.5: determine if the right shifted tz + T exceeds the sampling curve endpoint? If yes, enter step 4.6; otherwise, returning to the step 4.2;

Step 4.6: determining no fault removal point, setting tz=0, and entering step 4.8;

Step 4.7: determining tz as a fault cut point, and entering step 4.8;

Step 4.8: outputting tz to a main flow;

Step 4.9: and exiting the process and entering the next step of the main process.

As shown in fig. 5, the specific steps for determining the reclosing action point are as follows:

Step 6.1: let si be the current sampling value, T be the sampling period, in be the current first cycle power frequency effective value, tz be the fault cut-off point, traverse si along the T axis with tc as the variable, initialize tc=tz+20t (after 400 ms);

Step 6.2: let ic be the current power frequency effective value at time tc, iz5T be the minimum power frequency effective value of five cycle current (i.e. the minimum current sampling value after fault removal) after time tz;

Step 6.3: judging whether the following 4 conditions are satisfied: ① (ic >0.1 x in) and (ic >10 x iz 5T); ② (ic >0.2 in) and (ic >5 iz 5T); ③ (ic >0.5 in) and (ic >3 iz 5T); ④ (ic > 0.003A);

if the conditions ①②③ meet 1 or more and the conditions ④ are met, the process proceeds to step 6.7;

If condition ①②③ is satisfied for 1 or more, but condition ④ is not satisfied; or none of the conditions ①②③ is satisfied, go to step 6.4;

Step 6.4: tc right shift by T/4;

Step 6.5: determining if the right shifted tc+t exceeds the sampling curve endpoint? If yes, go to step 6.6; otherwise, returning to the step 6.2;

Step 6.6: determining no reclosing action point, setting tc=0, and entering step 6.8;

Step 6.7: determining tc to be a reclosing action point, and entering a step 6.8;

step 6.8: outputting tc to the main flow;

Step 6.9: and exiting the process and entering the next step of the main process.

As shown in fig. 6, for determining the secondary failure start point, the specific steps are as follows:

Step 8.1: let si be the current sampling value, T be the sampling period, in be the current first cycle power frequency effective value, tc be the reclosing action point, traverse si along the T axis with tcf as the variable, initialize tcf=tc;

Step 8.2: let icf be the current power frequency effective value at the time tcf, if5T be the maximum power frequency effective value of the five-cycle current (namely the maximum fault current of the primary fault) after the time tf;

step 8.3: judging whether the following 3 conditions are satisfied: ① (icf >1.15 in) and (icf >0.9 if 5T); ② (icf >1.3 in) and (icf >0.8 if 5T); ③ (icf >1.5 in) and (icf >0.6 if 5T);

if the conditions ①②③ meet 1 or more, the step 8.7 is performed;

If none of the conditions ①②③ is satisfied, go to step 8.4;

Step 8.4: tcf right shift T/4;

Step 8.5: judging whether the right shifted tcf+t exceeds the sampling curve end point? If yes, go to step 8.6; otherwise, returning to the step 8.2;

Step 8.6: determining no secondary failure starting point, setting tcf=0, and entering step 8.8;

step 8.7: determining tcf as a secondary failure starting point, and entering step 8.8;

step 8.8: outputting tcf to the main flow;

Step 8.9: and exiting the process and entering the next step of the main process.

As shown in fig. 7, the specific steps for determining the secondary failure removal point are as follows:

step 10.1: let si be the current sampling value, T be the sampling period, in be the current first cycle power frequency effective value, tcf be the secondary fault starting point, traverse si along the T axis with tcz as the variable, initialize tcz=tcf;

Step 10.2: icz is the current power frequency effective value at the time tcz, if5T is the maximum power frequency effective value of the five-cycle current (namely the maximum fault current of the primary fault) after the time tf;

Step 10.3: judging whether the following 3 conditions are satisfied: ① (icz <0.01 if 5T) and (icz < 0.1A); ② ((icz <0.5 if 5T) and (icz < 0.1A); ③ (icz < 0.003A);

if the conditions ①②③ meet 1 or more, the step 10.7 is entered;

if none of the conditions ①②③ is satisfied, go to step 10.4;

step 10.4: tcz moves right by T/4;

step 10.5: judging whether the right shifted tcz+t exceeds the sampling curve endpoint? If yes, go to step 10.6; otherwise, returning to the step 10.2;

Step 10.6: determining no secondary fault removal point, setting tcz=0, and entering step 10.8;

Step 10.7: determining tcz as a secondary fault removal point, and entering step 10.8;

Step 10.8: outputting tcz to the main flow;

step 10.9: and exiting the process and entering the next step of the main process.

As shown in fig. 9, for calculating the failure start point, the specific steps are as follows:

step 2.13.1: let si be the current sampling value, T be the sampling period, IN be the rated current value, and tx be the position of the sampling value mutation point obtained by subsequent calculation along the T axis (tx is the sampling value mutation point position obtained by subsequent calculation, the sampling value mutation point satisfying the condition can be determined as the fault starting point), as shown IN fig. 8, IN fig. 8, indicates the rated current value, unit a (ampere), si indicates the current sampling value, T indicates the sampling period, tx indicates the position of the calculation mutation point, stx indicates the tx position current sampling value, unit a (ampere), s (tx+t) indicates the tx+t position current sampling value, unit a (ampere), s (tx+2t) indicates the tx+2t position current sampling value, unit a (ampere);

Step 2.13.2: let tf be the failure start point, initialize tx=0, tf=0;

step 2.13.3: let stx be the tx position current sample, s (tx+t) be the tx+t position current sample, s (tx+2t) be the tx+2t position current sample;

Step 2.13.4: judging whether the condition is satisfied: s (tx+2T) -s (tx+T) | -s (tx+T) -stx| > IN/50; if yes, go to step 2.13.8; otherwise, go to step 2.13.5;

step 2.13.5: tx right shift by T/4;

step 2.13.6: determine if the right shifted tx+t exceeds the sampling curve endpoint? If yes, go to step 2.13.7; otherwise, returning to the step 2.13.3;

step 2.13.7: if the sampling value mutation point tx is not determined, tf=0, and the step 2.13.9 is entered;

Step 2.13.8: determining a mutation point tx, and letting tf=tx+2t, and entering step 2.13.9;

step 2.13.9: outputting tf to the sub-process;

step 2.13.10: and exiting the process and entering the next step of the sub-process.

Example 2

For one embodiment of the present invention, a system for determining a power grid fault based on recording source end data mining is provided, including:

the device comprises a data processing module, a fault positioning module, a reclosing identification module and a fault judging module;

the data processing module is used for loading the recording data file, performing preprocessing operation, cleaning and formatting data, and initializing key element value variables of fault time sequence characteristics;

the fault positioning module determines a starting point and an ending point of a fault by using an algorithm according to the preprocessing data of the data processing module, analyzes current and voltage time sequence characteristics of a power grid, and identifies the accurate position of the fault;

The reclosing identification module is used for identifying the moment of the reclosing operation of the power grid, and judging the action time of the reclosing and the starting point and the ending point of the subsequent secondary fault;

The fault judging module is used for carrying out final fault analysis and judgment, providing fault type and cause analysis and providing preventive measures according to analysis results.

Example 3

One embodiment of the present invention, which is different from the first two embodiments, is:

The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, randomAccess Memory), a magnetic disk, an optical disk, or other various media capable of storing program codes.

Logic and/or steps represented in the flowcharts or otherwise described herein, e.g., a ordered listing of executable instructions for implementing logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.

More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). In addition, the computer readable medium may even be paper or other suitable medium on which the program is printed, as the program may be electronically captured, via, for instance, optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner, if necessary, and then stored in a computer memory.

It is to be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, may be implemented using any one or combination of the following techniques, as is well known in the art: discrete logic circuits having logic gates for implementing logic functions on data signals, application specific integrated circuits having suitable combinational logic gates, programmable Gate Arrays (PGAs), field Programmable Gate Arrays (FPGAs), and the like.

Example 4

Referring to fig. 10-11, for one embodiment of the present invention, a method for determining a power grid fault based on recording source end data mining is provided, and in order to verify the beneficial effects of the present invention, scientific demonstration is performed through experiments.

And (3) randomly selecting 893 groups for mixed testing on grid fault recording data (the data format of which conforms to the COMTRADE standard) recorded by the Yunnan power dispatching control center in 2003. A test sample screenshot is shown at 10.

The recording data is imported by adopting a centralized and automatic program, the data analysis and calculation are carried out by applying the method of the patent, the calculation result is automatically compared with the fault time sequence characteristics recorded in practice and output, 5ms (namely 1/4T sampling period step length given by the patent) is taken as the upper limit of the allowable deviation, the given analysis result is shown in a graph 11,

The graph shows that all the analysis of 893 groups of recording data is completed, and the analysis result report gives the following conclusion: the upper limit of fault time sequence characteristic deviation reaches the number of 2ms to be 0.

The test adopts field real data recorded by the actual operation of the Yunnan power dispatching control center, and batch samples are formed after 220kV and 500kV power grid fault recording data are randomly mixed and scattered, and the centralized analysis result of the samples proves the practicability and the accuracy of the method.

It should be noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention may be modified or substituted without departing from the spirit and scope of the technical solution of the present invention, which is intended to be covered in the scope of the claims of the present invention.

Claims (4)

1. A power grid fault judging method based on recording source end data mining is characterized by comprising the following steps:

loading a recording data file, naming and initializing key element value variables of fault time sequence characteristics;

Determining a fault starting point based on the key element value variable, determining a fault clearing point after determining the fault starting point, and judging whether the fault clearing point is determined or not;

after determining the fault removal point, determining a reclosing action point, and judging whether the reclosing action point is determined;

After determining the reclosing action point, determining a secondary fault starting point, and judging whether the secondary fault starting point is determined;

after determining the secondary fault starting point, determining a secondary fault clearing point, and judging whether the secondary fault clearing point is determined;

determining the type of the fault time sequence characteristic and corresponding key elements based on the reclosing action point, the re-fault starting point and the judging result of the re-fault removing point;

Determining the type of the fault time sequence characteristic and the corresponding key element value based on the point position;

The loading recording data file comprises the recording data which follows the COMTRADE standard, voltage and current sampling values in the power grid fault process and corresponding time information of each sampling value are obtained by loading the recording data, and key element values of fault time sequence characteristics are determined;

the key element values comprise tf representing a fault starting point, tz representing a fault cutting point, tc representing a reclosing action point, tcf representing a secondary fault starting point and tcz representing a secondary fault cutting point;

The initialization comprises unified initialization of variables to 0;

the fault starting point determination comprises calculating power frequency effective values of voltage and current phase by phase, and calculating a power grid fault starting point according to attenuation of the calculated voltage power frequency effective values and increase characteristics of the current power frequency effective values;

the starting point of calculating the power grid fault comprises the steps of setting a failure counter j and initializing j=0;

Sampling voltage and current sampling values phase by phase in the sequence of the voltage and current sampling values A, B, C; when the failure counter j=0, taking an A-phase voltage and current sampling value; when the failure counter j=1, taking a B-phase voltage and current sampling value; when the failure counter j=2, taking a C-phase voltage and current sampling value;

Let su be the voltage sampling value, si be the current sampling value, T be the sampling period, un be the voltage first cycle power frequency effective value, and in be the current first cycle power frequency effective value; traversing si along the T axis by taking tx as a variable, taking the step length as T/4, and initializing tx=0;

initializing tf=0, calculating a fault starting point tf based on tx;

Judging whether the determined mutation point is positioned behind the current tx point, wherein if tf is more than tx, letting uf be the voltage power frequency effective value at tf moment, if be the current power frequency effective value at tf moment, if5T be the maximum power frequency effective value of five cycle currents after tf moment;

Judging whether a fault starting point prediction condition is met or not, wherein the fault starting point prediction condition comprises a first fault starting point prediction condition, a second fault starting point prediction condition, a third fault starting point prediction condition and a fourth fault starting point prediction condition, and the first fault starting point prediction condition is if > 1.15 x in and uf < 0.9 x un; the second failure start point prediction condition is if > 1.3 x in and uf < 0.95 x un; the third failure start point prediction condition is if > 1.5 in and if > 0.5A; the fourth failure start point prediction condition is if > 2.5 if 5T;

if the first fault starting point prediction condition, the second fault starting point prediction condition and the third fault starting point prediction condition meet 1 or more and meet the fourth fault starting point prediction condition, determining the mutation point tf as a fault starting point;

If the first fault starting point prediction condition, the second fault starting point prediction condition and the third fault starting point prediction condition meet 1 or more, but do not meet the fourth fault starting point prediction condition or do not meet the first fault starting point prediction condition, the second fault starting point prediction condition and the third fault starting point prediction condition, and the tx shifts to the right by T/4;

if tf is less than or equal to tx, tx is shifted right by T/4;

When tx moves by T/4 to the right, judging whether tx+2T after the right movement exceeds the end point of the sampling curve, and if so, judging that the failure counter +1 is expressed as j=j+1; judging whether the failure times reach 3 times or not, wherein j=3; if the fault starting point is not determined for 3 times, tf=0 is set, and if the fault starting point is not determined for 3 times, the voltage and current sampling values are sampled phase by phase in the sequence of A, B, C of the voltage and current sampling values again;

if the sampling curve end point is not exceeded, re-initializing tf=0, and calculating a fault starting point tf based on tx;

Searching for a mutation point based on the current sampling value, calculating power frequency effective values of voltage and current based on the mutation point position, and analyzing whether the power frequency effective values accord with fault characteristics;

When judging and determining a fault starting point, carrying out a fault removal point determining process;

if the fault starting point is not determined, the process is stopped in a failure mode;

Determining a fault removal point comprises enabling si to be a current sampling value, T to be a sampling period, in to be a current first-cycle power frequency effective value, tf to be a fault starting point, traversing si along a T axis by taking tz as a variable, and initializing tz = tf;

Let iz be the effective value of the current power frequency at the tz moment, and iz5T be the effective value of the minimum power frequency of the five-cycle current after the tz moment;

Judging whether a fault clearing point prediction condition is met or not, wherein the fault starting point prediction condition comprises a first fault clearing point prediction condition, a second fault clearing point prediction condition, a third fault clearing point prediction condition and a fourth fault clearing point prediction condition, and the first fault clearing point prediction condition is that iz <0.01 x iz5T and iz <0.1A; the second fault cut point prediction condition is iz <0.5 x in and in <0.1A; the third fault clearing point prediction condition is iz <0.003A; the fourth fault cut-off point prediction condition is iz5T <0.003A;

If the first fault clearing point prediction condition, the second fault clearing point prediction condition and the third fault clearing point prediction condition meet 1 or more and meet the fourth fault clearing point prediction condition, determining tz as a fault clearing point;

If the first fault clearing point prediction condition, the second fault clearing point prediction condition and the third fault clearing point prediction condition meet 1 or more, but do not meet the fourth fault clearing point prediction condition or do not meet the first fault clearing point prediction condition, the second fault clearing point prediction condition and the third fault clearing point prediction condition, and the tz moves right by T/4; judging whether the tz+T after the right shift exceeds the end point of the sampling curve, if so, judging that the fault cut point is not determined, setting tz=0, and exiting in a failure mode;

The step of determining the reclosing action point comprises the steps of calculating the ratio of the current at the reclosing action reference point position to the normal load current before the fault and the multiple of the minimum current after the fault is removed, and performing feature analysis by combining the ratio and the multiple;

If the reclosing action point cannot be confirmed, confirming that the time sequence characteristic of the fault is class A, outputting tf and tz values and successfully exiting; if the reclosing action point is confirmed, determining a secondary fault starting point;

the determining of the secondary fault starting point comprises the steps of calculating the multiple of the current of the secondary fault reference point position and the normal load current before the fault and the ratio of the current to the maximum fault current, and performing feature analysis through the combination of the multiple and the ratio;

If the secondary fault starting point cannot be determined, confirming that the time sequence characteristic of the fault is class B, outputting values tf, tz and tc and successfully exiting; if the secondary failure starting point is determined, determining a secondary failure cutting point;

Determining the secondary fault removal point comprises determining that the time sequence characteristic of the fault is class B if the secondary fault removal point cannot be determined, outputting values tf, tz and tc and successfully exiting;

If the secondary fault removal point is determined, confirming that the time sequence characteristic of the fault is class C, outputting tf, tz, tc, tcf, tcz values and successfully exiting;

The time sequence characteristic of the fault is that the corresponding process of the class A is fault removal after the fault starts; corresponding key elements are tf and tz, and are triggered by 220kV voltage class interphase faults, three-phase faults and faults generated in the power transmission process;

the time sequence characteristics of the faults are that the corresponding process of class B is fault starting, fault removing and reclosing actions; corresponding key elements are tf, tz and tc respectively, and are triggered by single-phase faults, 110kV voltage class interphase faults and three-phase faults;

the time sequence characteristics of the faults are that the corresponding processes of the class C are fault start, fault removal, reclosing action, secondary fault and secondary fault removal; corresponding key elements are tf, tz, tc, tcf and tcz respectively, and the reclosing action is unsuccessful by single-phase faults with voltage classes of 220kV and above and faults with voltage classes of 110kV, or the reclosing action is triggered by faults again after success;

Determining a reclosing action point, wherein the specific steps comprise that si is a current sampling value, T is a sampling period, in is a current first-cycle power frequency effective value, tz is a fault cutting point, tc is used as a variable, si is traversed along a T axis, and tc=tz+20T is initialized;

Let ic be the current power frequency effective value at time tc, iz5T be the minimum power frequency effective value of the five-cycle current after time tz;

Judging whether a reclosing action point prediction condition is met or not, wherein the reclosing action point prediction condition comprises a first reclosing action point prediction condition, a second reclosing action point prediction condition, a third reclosing action point prediction condition and a fourth reclosing action point prediction condition; the first reclosing action point prediction condition is ic >0.1 x in and ic >10 x iz5t; the second reclosing action point prediction condition is ic >0.2 x in and ic >5 x iz5t; the third reclosing action point prediction condition is ic >0.5 x in and ic >3 x iz5t; the fourth reclosing action point prediction condition is ic >0.003A;

if the first reclosing action point prediction condition, the second reclosing action point prediction condition and the third reclosing action point prediction condition meet 1 or more and meet the fourth reclosing action point prediction condition, determining tc to be a reclosing action point, and outputting tc to a main flow;

If the first reclosing action point prediction condition, the second reclosing action point prediction condition and the third reclosing action point prediction condition meet 1 or more, but the fourth reclosing action point prediction condition is not met or the first reclosing action point prediction condition, the second reclosing action point prediction condition and the third reclosing action point prediction condition are not met, tc shifts to the right by T/4, whether tc+T after right shifting exceeds the end point of the sampling curve is judged, if yes, the reclosing action point is not determined, tc=0 is set, and tc is output to the main flow;

Determining a secondary fault starting point, wherein the specific steps comprise the steps of enabling si to be a current sampling value, T to be a sampling period, in to be a current first-cycle power frequency effective value, tc to be a reclosing action point, traversing si along a T axis by taking tcf as a variable, and initializing tcf=tc;

Let icf be the current power frequency effective value at the time of tcf, if5T be the maximum power frequency effective value of the five-cycle current after the time of tf;

Judging whether a secondary failure starting point prediction condition is met or not, wherein the secondary failure starting point prediction condition comprises a first secondary failure starting point prediction condition, a second secondary failure starting point prediction condition and a third secondary failure starting point prediction condition;

The first re-failure starting point prediction condition is icf >1.15 x in and icf >0.9 x if5t, the second re-failure starting point prediction condition is icf >1.3 x in and icf >0.8 x if5t, and the third re-failure starting point prediction condition is icf >1.5 x in and icf >0.6 x if5t;

if the first secondary failure starting point prediction condition, the second secondary failure starting point prediction condition and the third secondary failure starting point prediction condition meet 1 or more, determining tcf as a secondary failure starting point;

If the first secondary failure starting point prediction condition, the second secondary failure starting point prediction condition and the third secondary failure starting point prediction condition are not met, the tcf moves to the right by T/4; judging whether the right-shifted tcf+T exceeds the end point of the sampling curve, if so, not determining a secondary failure starting point, and setting tcf=0;

Determining a secondary fault removal point, wherein the specific steps comprise the steps of enabling si to be a current sampling value, T to be a sampling period, in to be a current first-cycle power frequency effective value, tcf to be a secondary fault starting point, traversing si along a T axis by taking tcz as a variable, and initializing tcz=tcf;

icz is the current power frequency effective value at the time tcz, if5T is the maximum power frequency effective value of the five-cycle current after the time tf;

judging whether a secondary fault clearing point prediction condition is met or not, wherein the secondary fault clearing point prediction condition comprises a first secondary fault clearing point prediction condition, a second secondary fault clearing point prediction condition and a third secondary fault clearing point prediction condition;

If the first secondary fault removal point prediction condition, the second secondary fault removal point prediction condition and the third secondary fault removal point prediction condition meet 1 or more, determining tcz as a secondary fault removal point;

If the first secondary fault removal point prediction condition, the second secondary fault removal point prediction condition and the third secondary fault removal point prediction condition do not meet tcz right shift T/4; judging whether the right-shifted tcz+T exceeds the end point of the sampling curve, if so, not determining a secondary fault removal point;

determining a fault starting point, wherein the specific steps comprise that si is a current sampling value, T is a sampling period, IN is a rated current value, and tx is used as a variable to traverse si along a T axis;

Let tf be the failure start point, initialize tx=0, tf=0;

Let stx be the tx position current sample, s (tx+t) be the tx+t position current sample, s (tx+2t) be the tx+2t position current sample;

Judging whether the ratio of (tx+2T) -s (tx+T) to s (tx+T) -stx is larger than IN/50 is satisfied;

if the determined mutation point tx is satisfied, letting tf=tx+2t;

If the tx right shift T/4 is not satisfied; judging whether the tx+T after the right shift exceeds the end point of the sampling curve, if so, determining no sampling value abrupt change tx, and if so, making tf=0 and outputting tf.

2. A system employing the grid fault determination method based on record wave source end data mining as set forth in claim 1, comprising:

the device comprises a data processing module, a fault positioning module, a reclosing identification module and a fault judging module;

The data processing module is used for loading the recording data file, performing preprocessing operation, cleaning and formatting data, and initializing key element value variables of fault time sequence characteristics;

the fault positioning module determines a starting point and an ending point of a fault by using an algorithm according to the preprocessing data of the data processing module, analyzes the current and voltage time sequence characteristics of the power grid, and identifies the accurate position of the fault;

The reclosing identification module is used for identifying the moment of the reclosing operation of the power grid, and judging the action time of the reclosing and the starting point and the ending point of the subsequent secondary fault;

The fault judging module is used for carrying out final fault analysis and judgment, providing fault type and cause analysis and providing preventive measures according to analysis results.

3. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the grid fault determination method based on record wave source end data mining according to claim 1.

4. A computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor realizes the steps of the grid fault determination method based on record wave source side data mining according to claim 1.