CN110673024B - A high-voltage circuit breaker fault early warning method based on the power Internet of things - Google Patents

Abstract

The invention relates to the technical field of power system maintenance, in particular to a high-voltage circuit breaker fault early warning method based on a power internet of things, which comprises the following steps of: A) acquiring historical detection data; B) artificially setting a fault source, enabling the high-voltage circuit breaker to work until a fault occurs, and periodically detecting to obtain detection data; C) obtaining sample data, training a fault studying and judging neural network model; D) extracting precursor reference data; E) if the difference between the detection data and the precursor reference data is smaller than a set threshold, sending out a fault early warning; F) and inputting detection data into the fault early warning neural network model, and if a fault exists, sending out a fault alarm. The substantial effects of the invention are as follows: the high-voltage circuit breaker can be rapidly subjected to fault study and judgment through detection data; the high-voltage circuit breaker fault early warning system prompts maintenance personnel to perform targeted maintenance, and ensures that the maintenance of the high-voltage circuit breaker is more reliable.

Description

A high-voltage circuit breaker fault early warning method based on power Internet of things

technical field

The invention relates to the technical field of power system maintenance, in particular to a high-voltage circuit breaker fault early warning method based on the power Internet of Things.

Background technique

The high-voltage circuit breaker is responsible for the dual tasks of control and protection in the power system, and its performance is directly related to the safe operation of the power system. Among them, the mechanical characteristic parameter is one of the important parameters to judge the performance of the circuit breaker. However, there are many testing items for high-voltage circuit breakers, the testing process is time-consuming and labor-intensive, and the efficiency is low. Seriously affect the operation and maintenance of high-voltage circuit breakers. Although some technologies have appeared at present, the detection efficiency of high-voltage circuit breakers has been accelerated and the detection time has been reduced. However, the judgment of the detected data still relies on manual judgment. The accuracy of manual data research and judgment depends on the experience quality of the manual itself, and the reliability is poor. And manual research and judgment of data can only find data that is obviously abnormal. For abnormal data that is not obvious enough, manual judgment is very difficult. It is impossible to realize the early warning of the failure of the high-voltage circuit breaker.

For example, Chinese patent CN203707635U, published on July 9, 2014, discloses a high-voltage circuit breaker early warning signal circuit. This high-voltage circuit breaker early warning signal circuit consists of signal screen bus XPM, circuit breaker trip position relay normally closed contact TWJ2, circuit breaker close Gate position relay normally closed contact HWJ2, AC contactor 3KA normally open contact 3KA, resistor R1, pre-alarm bus YBM1 in series, including signal screen bus XPM, circuit breaker trip position relay normally closed contact TWJ2, circuit breaker closing position relay normally The closed contact HWJ2 is also connected to the DC negative power supply-KM1 after being connected in series with the AC contactor 3KA coil, a pair of parallel limit normally open contacts, and the switch QF1. The pair of parallel limit normally open contacts are controlled by the limit The normally open contact SW and the limit normally open contact YW are connected in parallel. However, it cannot solve the technical problem that the fault of the high-voltage circuit breaker cannot be accurately evaluated and warned.

SUMMARY OF THE INVENTION

The technical problem to be solved by the present invention is: the technical problem of currently lacking a method for fault assessment and early warning of a high-voltage circuit breaker. A high-voltage circuit breaker fault early warning system based on power Internet of Things is proposed to quickly assess faults and perform fault early warning.

In order to solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a high-voltage circuit breaker fault early warning method based on the power Internet of things, comprising the following steps: A) obtaining all historical detection data of the high-voltage circuit breaker; B) obtaining the same model The high-voltage circuit breaker is designed to artificially set the fault source under laboratory conditions, so that the high-voltage circuit breaker will work in the presence of the fault source until the fault occurs, and at the same time, it will be tested periodically to obtain the test data; C) N times before the fault occurs The detection data is associated with the corresponding fault type to form sample data, and the sample data is used to train the fault judgment neural network model; D) Extract the detection data between M and N times before the fault occurs, M>N, as the precursor reference data; E) Compare the detection data of the high-voltage circuit breaker with the precursor reference data, if the difference between the detection data and the precursor reference data is less than the set threshold, a fault warning will be issued; F) Input the detection data of the high-voltage circuit breaker into the fault obtained in step D) Early warning neural network model, if the fault early warning neural network model outputs the fault type, a fault alarm will be issued. By artificially setting the fault source, the correlation between the detection data and the fault type is higher, and the accuracy of the fault judgment neural network model and the fault early warning neural network model is higher. The fault judgment neural network model can quickly judge the fault of the high-voltage circuit breaker through the detection data. The fault early warning neural network model can give an early warning to the failure of the high-voltage circuit breaker, prompting the maintenance personnel to carry out targeted maintenance, so as to make the maintenance of the high-voltage circuit breaker more reliable. The detection data includes closing time, opening time, Just closing speed, just dividing speed, different phases of three phases, different phases of the same phase, gold short time, no-current time, maximum speed of moving contact, average speed of moving contact, action time of moving contact, bounce time, number of bounces , Maximum bounce amplitude, opening and closing stroke, current waveform curve of opening and closing process, dynamic curve of time and speed stroke in the opening and closing stroke of moving contact, opening distance and contact resistance. By obtaining various data of the high-voltage circuit breaker, the state data of the high-voltage circuit breaker is more comprehensive, which helps to improve the accuracy of fault judgment and provides conditions for the discovery of insignificant abnormal data.

Preferably, in step B), the method for artificially setting a fault source includes the following steps: B11) performing several tests on the high-voltage circuit breaker; B12) selecting a maintenance requirement in turn to make it substandard according to the maintenance requirements of the high-voltage circuit breaker, After performing several live opening and closing actions, perform several inspections; B13) Select two maintenance requirements in turn so that they do not meet the standard, and perform several live opening and closing actions, and then perform several inspections; B14) Use liquid nitrogen or dry ice Quickly cool the high-voltage circuit breaker, carry out several mechanical characteristic tests, and obtain the test data of the mechanical characteristic test. After cooling with liquid nitrogen or dry ice, the lubricating performance of the lubricant or lubricating oil is reduced, thereby simulating a jammed state. After the test is completed, the lubricating performance of the lubricant or lubricating oil is restored, thereby simulating the jamming of mechanical parts without damage. Fault type, get the status data under the fault type. Mechanical parts jam in nature because of poor lubrication or the ingress of dust particles.

Preferably, in step B, the method for establishing a fault judgment model for a high-voltage circuit breaker includes: B21) Obtaining all detection data, associating the detection data with the corresponding fault type, as sample data; B22) Preprocessing the sample data and normalizing it process, train the neural network model, and use the trained neural network model as the fault judgment model. The normalized data can improve the convergence speed of the neural network model and improve the training efficiency and accuracy of the neural network model.

Preferably, in step B21), the method for associating the detection data with the corresponding fault type includes: B211) obtaining the detection data in step B11) as historical detection data; B212) combining step B12) and step B13), several Several groups of detection data obtained by secondary detection are compared with historical detection data successively, if the difference between detection data and historical detection data is greater than the preset threshold, then this group of detection data is associated with the maintenance requirements that do not meet the standard; B213) will step B14) in) Several sets of test data obtained from several mechanical characteristic tests are compared with historical test data. If the difference between test data and historical test data is greater than a preset threshold, the set of test data is associated with the jamming fault of mechanical components. The detection data is associated with the fault type to form sample data, and the neural network model is trained.

Preferably, the method for judging whether the difference between the detection data and the historical detection data is greater than a preset threshold value includes: B31) performing segmentation processing on the detection data and the numerical value in the historical detection data, using the segment interval as the name, and converting the numerical value into a state B32) Convert the state quantities in the detection data and historical detection data into Boolean quantities, and use {0, 1} to represent false and true respectively; B32) Treat the Boolean quantities of the processed historical detection data as numerical values and calculate the mean value , the mean value is rounded to an integer, and the obtained integer is regarded as a Boolean quantity again, and the processed detection data and historical detection data are sorted according to the settings to form a detection vector and a historical detection vector respectively; B33) Calculate the difference between the detection vector and the historical detection vector The distance is compared with the preset distance threshold. If the distance is greater than the preset distance threshold, it is determined that the distance between the detection data corresponding to the detection vector and the historical detection data is greater than the preset threshold. Otherwise, the detection data corresponding to the detection vector and the historical detection data are determined. The distance of the data is not greater than the preset threshold. Data can be simplified by converting state quantities to Boolean quantities.

Preferably, in step B31), the method for performing segmentation processing on the numerical value in the detection data and the historical detection data includes: B311) selecting a numerical value, obtaining all the values of the numerical value in the historical detection data, and pressing The numerical values are arranged in order, denoted as the set Ki, and the minimum value in the set Ki is km _min and the maximum value is km _max ; B312) assign the initial value of the partition starting point k _s to k _min , and the partition end point _ke to assign the initial value to k _max , the investigation value _{km = k s +} n×Δk, Δk is the manually set step size, n is a positive integer, and the initial value of _n is 1;

Among them, the function N(x, y) represents the set Ki, and the number of data whose data values are in the numerical interval (x, y), then ( _{2km -k s )} is taken as the interval division point and added to the division point set Km, and ( 2km -k _s ) is assigned to _{k s ,} and _n continues to increment by 1 until km >km _max ; _B314 ) Add km and _km to the set Km, and use the value in Km as the dividing point, Divide the numerical value data into numerical intervals; B315) select the next numerical value, repeat steps B311) to B314) until all the numerical quantities are divided into interval segments; B316) detection data adopts the corresponding numerical value in the historical detection data. interval division. The numerical division can simplify the data, improve the efficiency of data processing, and improve the convergence speed of the neural network model.

分别作为对应数值区间的状态名；若历史检测数值量的数据，落入区间[n_m(d)，n_m(d+1)]，d∈[1，k-1]，则将状态名

作为该数值量的取值，完成数值量数据转化为状态量数据。Preferably, in step B31), using the segmented interval as the name, the method for converting a numerical quantity into a state quantity includes the following steps: dividing the numerical quantity data into several intervals, [n _m(1) , n _m(2) ], [n _m(2) , n _m(3) ]…[n _m(k-1) , n _m(k) ], where n _m(1) and n _m(k) are the starting points of the numerical interval, respectively and the end point, n _m(2) ～n _m(k-1) is the middle dividing point of the numerical interval, the

respectively as the state name of the corresponding numerical interval; if the historically detected numerical data falls into the interval [n _m(d) , n _m(d+1) ], d∈[1, k-1], the state name will be

As the value of the numerical quantity, the conversion of numerical quantity data into state quantity data is completed.

Preferably, in step B), a non-contact displacement sensor is installed on each mechanical moving part of the normal high-voltage circuit breaker, and under the condition of power failure, the opening and closing test of the high-voltage circuit breaker is continuously repeated until the high-voltage circuit breaker is disconnected. If the mechanical parts of the circuit breaker are damaged, record the opening and closing times N during the test, and the displacement data of each mechanical moving part during the opening and closing process as the historical displacement data; A non-contact displacement sensor is installed on each mechanical moving part, and the high-voltage circuit breaker to be judged is opened and closed once to obtain the displacement data measured by the non-contact displacement sensor, and compare it with the historical displacement data to obtain the closest displacement data. The number of opening and closing tests n corresponding to the historical displacement data is taken as (N-n) as the remaining service life of the high-voltage circuit breaker to be judged. The non-contact displacement detection does not interfere with the operation of the high-voltage circuit breaker, is easy to deploy, and has higher displacement detection accuracy.

The non-contact displacement sensor includes laser transmitter, current limiting resistor, photoresistor, power supply module, reflective sticker, voltage sensor and communication module. The laser transmitter is fixedly installed in the casing of the high-voltage circuit breaker and is aligned with the mechanical moving parts along the normal direction. An alignment point on the outer surface is adjusted so that the emitted light of the laser transmitter has an included angle with the normal direction of the outer surface of the mechanical moving part. During the stroke of the mechanical moving part, the alignment point of the laser transmitter moves along the outer surface of the mechanical moving part, forming a Moving range, the reflective sticker is attached to the mechanical moving part and covers the moving range of the alignment point. The reflective sticker has several high-reflection areas arranged at equal intervals along the stroke of the mechanical moving part, and the adjacent high-reflection areas are Low-reflection area, the width of high-reflection area is equal to the width of low-reflection area, the spot diameter of the laser emitter is equal to the integral multiple of the interval width, the photoresistor is installed on the other side of the normal symmetry with the laser emitter about the outer surface of the mechanical moving part, One end of the photoresistor is grounded, and the other end is connected to the power supply module through the current limiting resistor. The voltage sensor collects the voltage at the connection point between the photoresistor and the current limiting resistor, and the voltage sensor is connected to the communication module.

As an alternative, in step B), a non-contact displacement sensor is installed on each mechanical moving part of the normal high-voltage circuit breaker; Under the conditions, the opening and closing test of the high-voltage circuit breaker is repeated continuously until the mechanical parts of the high-voltage circuit breaker are damaged; the displacement data of each mechanical moving part during the opening and closing process during the test is recorded and the corresponding maintenance requirements are not up to standard. Correspond to the fault, then repair the high-voltage circuit breaker, and carry out the next test that the maintenance requirements are not up to standard; in step C), install non-contact displacement sensors on each mechanical moving part of the high-voltage circuit breaker to be judged. The high-voltage circuit breaker to be judged is opened and closed once to obtain the displacement data measured by the non-contact displacement sensor, and the displacement data measured by the non-contact displacement sensor of the high-voltage circuit breaker to be judged is compared with the historical displacement data to obtain The number of opening and closing tests n corresponding to the closest historical displacement data is taken as (N-n) as the remaining service life of the high-voltage circuit breaker to be judged. The remaining service life of the high-voltage circuit breaker can be obtained, which can provide a reference for the maintenance of the high-voltage circuit breaker.

The substantial effect of the invention is: by artificially setting the fault source, the correlation between the detection data and the fault type is higher, so that the accuracy of the fault judgment neural network model and the fault early warning neural network model is higher; through the fault judgment neural network model It can quickly judge the fault of the high-voltage circuit breaker through the detection data; through the fault early warning neural network model, the fault of the high-voltage circuit breaker can be warned, prompting the maintenance personnel to carry out targeted maintenance, making the maintenance of the high-voltage circuit breaker more reliable.

Description of drawings

FIG. 1 is a flow chart of a method for early warning of a high-voltage circuit breaker fault in the first embodiment.

FIG. 2 is a flowchart of a method for manually setting a fault source according to an embodiment.

FIG. 3 is a schematic structural diagram of a non-contact displacement sensor according to the first embodiment.

4 and 5 are schematic diagrams of the measurement of the non-contact displacement sensor according to the first embodiment.

Among them: 1. Linear reflection sticker, 2. Laser transmitter, 3. Cylindrical reflection sticker, 4. Cam, 5. Cylindrical end surface reflection sticker, 6. Moving parts, 7. Alignment point trajectory, 8. Arc reflection sticker, 100, a voltage sensor, 200, a communication module.

Detailed ways

The specific embodiments of the present invention will be further described in detail below through specific embodiments and in conjunction with the accompanying drawings.

Example 1:

A high-voltage circuit breaker fault early warning method based on the power Internet of things, as shown in Figure 1, includes the following steps: A) Obtaining all historical detection data of the high-voltage circuit breaker. The detection data includes closing time, opening time, just closing speed, just dividing speed, three-phase different periods, different periods of the same phase, gold short time, no-current time, maximum speed of moving contact, average speed of moving contact, Moving contact action time, bounce time, bounce times, maximum bounce amplitude, opening and closing stroke, current waveform curve of opening and closing process, dynamic curve of time and speed stroke in moving contact opening and closing stroke, opening distance and contact resistance. By obtaining various data of the high-voltage circuit breaker, the state data of the high-voltage circuit breaker is more comprehensive, which helps to improve the accuracy of fault judgment and provides conditions for the discovery of insignificant abnormal data.

B) Obtain a high-voltage circuit breaker of the same type, and artificially set the fault source under laboratory conditions, so that the high-voltage circuit breaker works in the presence of the fault source until a fault occurs, and at the same time, it is periodically tested to obtain test data. As shown in Figure 2, the method for artificially setting a fault source includes the following steps: B11) testing the high-voltage circuit breaker several times; B12) selecting a maintenance requirement in turn according to the maintenance requirements of the high-voltage circuit breaker so that it does not meet the standard, and performing several After the electrified opening and closing action, carry out several tests; B13) Select two maintenance requirements in turn so that they do not meet the standard, and carry out several electrified opening and closing actions, and then carry out several inspections; B14) Use liquid nitrogen or dry ice for rapid cooling For high-voltage circuit breakers, perform several mechanical characteristic tests to obtain the test data of the mechanical characteristic test.

The method for establishing a fault judgment model for a high-voltage circuit breaker includes: B21) Obtaining all detection data, associating the detection data with the corresponding fault type, as sample data; B22) Preprocessing the sample data, normalizing processing, and training a neural network model , using the trained neural network model as the fault judgment model. The normalized data can improve the convergence speed of the neural network model and improve the training efficiency and accuracy of the neural network model.

In step B21), the method for associating the detection data with the corresponding fault type includes: B211) obtaining the detection data in step B11) as historical detection data; B212) obtaining the detection data in steps B12) and B13) for several times. Several groups of detection data are compared with the historical detection data in turn, if the difference between the detection data and the historical detection data is greater than the preset threshold, then the group of detection data is associated with the maintenance requirements that do not meet the standard; B213) several times in step B14) Several sets of detection data obtained from the characteristic test are compared with the historical detection data. If the difference between the detection data and the historical detection data is greater than the preset threshold, the set of detection data is associated with the jamming failure of the mechanical parts. The detection data is associated with the fault type to form sample data, and the neural network model is trained.

The method for judging whether the difference between the detection data and the historical detection data is greater than a preset threshold value includes: B31) performing segmentation processing on the numerical quantities in the detection data and the historical detection data, and converting the numerical quantities into state quantities with the segment interval as the name; B32 ) Convert the state quantities in the detection data and historical detection data into Boolean quantities, and use {0, 1} to represent false and true respectively; B32) Treat the Boolean quantities of the processed historical detection data as numerical values, and calculate the mean value. Round up to an integer, the obtained integer is regarded as a Boolean quantity again, and the processed detection data and historical detection data are sorted according to the settings to form a detection vector and a historical detection vector respectively; B33) Calculate the distance between the detection vector and the historical detection vector, and The preset distance threshold is compared. If the distance is greater than the preset distance threshold, it is determined that the distance between the detection data corresponding to the detection vector and the historical detection data is greater than the preset threshold. Otherwise, the distance between the detection data corresponding to the detection vector and the historical detection data is determined. not greater than the preset threshold. Data can be simplified by converting state quantities to Boolean quantities.

In step B31), the method that the numerical quantity in the detection data and the historical detection data is carried out segmentation processing comprises: B311) select a numerical quantity, obtain all the value values of this numerical quantity in the historical detection data, and sequentially according to the numerical value. Arrangement, denoted as the set Ki, the minimum value in the set Ki is km _min and the maximum value is km _max ; B312) assign the starting point k _s of the partition to the initial value of km _min , and the end point _ke of the partition to assign the initial value of km _max , and the investigation value k _m = k _s +n×Δk, Δk is the manually set step size, n is a positive integer, and the initial value of _n is 1;

Among them, the function N(x, y) represents the set Ki, and the number of data whose data values are in the numerical interval (x, y), then ( _{2km -k s )} is taken as the interval division point and added to the division point set Km, and ( 2km -k _s ) is assigned to _{k s ,} and _n continues to increment by 1 until km >km _max ; _B314 ) Add km and _km to the set Km, and use the value in Km as the dividing point, Divide the numerical value data into numerical intervals; B315) select the next numerical value, repeat steps B311) to B314) until all the numerical quantities are divided into interval segments; B316) detection data adopts the corresponding numerical value in the historical detection data. interval division. The numerical division can simplify the data, improve the efficiency of data processing, and improve the convergence speed of the neural network model.

分别作为对应数值区间的状态名；若历史检测数值量的数据，落入区间[n_m(d)，n_m(d+1)]，d∈[1，k-1]，则将状态名

作为该数值量的取值，完成数值量数据转化为状态量数据。In step B31), the method for converting a numerical quantity into a state quantity with the segmented interval as the name includes the following steps: dividing the numerical quantity data into several intervals, [n _m(1) , nm ₍₂₎ ], [ n _m(2) , n _m(3) ]…[n _m(k-1) , n _m(k) ], where n _m(1) and n _m(k) are the start and end points of the numerical interval, respectively, n _m(2) ~n _m(k-1) is the middle dividing point of the numerical interval, and the

respectively as the state name of the corresponding numerical interval; if the historically detected numerical data falls into the interval [n _m(d) , n _m(d+1) ], d∈[1, k-1], the state name will be

As the value of the numerical quantity, the conversion of numerical quantity data into state quantity data is completed.

A non-contact displacement sensor is installed on each mechanical moving part of the normal high-voltage circuit breaker. Under the condition of power failure, the opening and closing test of the high-voltage circuit breaker is repeated continuously until the mechanical parts of the high-voltage circuit breaker are damaged, and record During the test, the number of opening and closing times N and the displacement data of each mechanical moving part during the opening and closing process are used as historical displacement data.

As shown in Figure 3, the non-contact displacement sensor includes a laser transmitter 2, a current limiting resistor, a photoresistor, a power supply module, a reflective sticker, a voltage sensor 100 and a communication module 200. The laser transmitter 2 is fixedly installed on the casing of the high-voltage circuit breaker Inside, align an alignment point on the outer surface of the mechanical moving part 6 along the normal direction, and adjust so that the light emitted by the laser transmitter 2 has an included angle with the normal direction of the outer surface of the mechanical moving part 6. During the stroke of the mechanical moving part 6, the laser emits The alignment point of the device 2 moves along the outer surface of the mechanical moving part 6 to form a moving range, the reflective sticker is attached to the mechanical moving part 6 and covers the moving range of the alignment point, and the reflective sticker has several equal intervals along the stroke of the mechanical moving part 6 Arranged high-reflection areas, between adjacent high-reflection areas is low-reflection area, the width of high-reflection area is equal to the width of low-reflection area, the spot diameter of laser emitter 2 is equal to an integral multiple of the interval width, photoresistor installation and laser emission The device 2 is on the other side of the normal symmetry of the outer surface of the mechanical moving part 6, one end of the photoresistor is grounded, and the other end is connected to the power supply module through the current limiting resistor. The voltage sensor 100 collects the voltage at the connection point between the photoresistor and the current limiting resistor. 100 is connected to the communication module 200 . Fig. 3 shows a linear reflection sticker 1, and the detected mechanical moving parts 6 move along a straight line, such as a moving contact, an unlocking lock, and the like. As shown in Fig. 4, when performing non-contact displacement detection on rotating parts, such as the shaft and the cam 4, the cylindrical surface reflection sticker 3 can be attached to the outer surface of the shaft or the arc portion of the cam 4 with equal radius. The picture is blurry and the distance between the high and low reflection areas is distorted. When the arc part of equal radius of the cam 4 is also a working surface, the cylindrical end face reflection sticker 5 can be attached to the end face of the cam 4 . As shown in FIG. 5 , when the detected moving part 6 has complex plane motion, that is, it contains both translational motion and rotational motion, select an appropriate alignment point on the detected moving part 6 so that the alignment point within the stroke of the moving part 6 Always on the moving part 6, the alignment point track 7 will be an arc, and an adapted arc-shaped reflection sticker 8 is attached. The arc-shaped reflection sticker 8 arranges the high-reflection area and the low-reflection area along the arc interval, so that the high-reflection area is And the edge of the low reflection area can be perpendicular to the arc at the corresponding position. This embodiment provides a non-contact displacement sensor implementation. In the prior art, non-contact displacement sensors are well known for detecting vibration and displacement. Those skilled in the art can design other forms of non-contact displacement by themselves. sensor to complete the detection of displacement.

C) Associate the detection data N times before the fault occurs with the corresponding fault type to form sample data, and use the sample data to train the fault judgment neural network model. A non-contact displacement sensor is installed on each mechanical moving part of the high-voltage circuit breaker to be judged, and the high-voltage circuit breaker to be judged is opened and closed once, and the displacement data measured by the non-contact displacement sensor is obtained and compared with the history Comparing the displacement data, obtain the number of opening and closing tests n corresponding to the closest historical displacement data, and take (N-n) as the remaining service life of the high-voltage circuit breaker to be judged. The non-contact displacement detection does not interfere with the operation of the high-voltage circuit breaker, is easy to deploy, and has higher displacement detection accuracy.

D) Extract the detection data between M and N times before the failure occurs, M>N, as the precursor reference data.

E) Compare the detection data of the high-voltage circuit breaker with the precursor reference data. If the difference between the detection data and the precursor reference data is less than the set threshold, a fault warning will be issued.

F) Input the detection data of the high-voltage circuit breaker into the fault early warning neural network model obtained in step D). If the fault early warning neural network model outputs the fault type, a fault alarm is issued. By artificially setting the fault source, the correlation between the detection data and the fault type is higher, and the accuracy of the fault judgment neural network model and the fault early warning neural network model is higher. The fault judgment neural network model can quickly judge the fault of the high-voltage circuit breaker through the detection data. The fault early warning neural network model can give early warning to the fault of the high-voltage circuit breaker, prompting the maintenance personnel to carry out targeted maintenance, making the maintenance of the high-voltage circuit breaker more reliable.

Embodiment 2:

This embodiment is further improved on the basis of the first embodiment. Specifically, in step B), a non-contact displacement sensor is installed on each mechanical moving part of the normal high-voltage circuit breaker; according to the maintenance of the high-voltage circuit breaker In order to meet the requirements, select a maintenance requirement in turn to make it substandard. Under the condition of power failure, repeat the opening and closing test of the high-voltage circuit breaker until the mechanical parts of the high-voltage circuit breaker are damaged; record each mechanical moving part during the test. In the process of opening and closing, the displacement data is associated with the corresponding faults that the corresponding maintenance requirements are not up to standard, and then the high-voltage circuit breaker is repaired, and the next test that the maintenance requirements are not up to standard is carried out; in step C), in the high-voltage circuit breaker to be judged A non-contact displacement sensor is installed on each mechanical moving part of the device, and the high-voltage circuit breaker to be judged is opened and closed once, and the displacement data measured by the non-contact displacement sensor is obtained. The displacement data measured by the displacement sensor is compared with the historical displacement data, and the number of opening and closing tests n corresponding to the closest historical displacement data is obtained, and (N-n) is used as the remaining service life of the high-voltage circuit breaker to be judged. The remaining service life of the high-voltage circuit breaker can be obtained, which can provide a reference for the maintenance of the high-voltage circuit breaker. The remaining steps are the same as those in the first embodiment.

The above-mentioned embodiment is only a preferred solution of the present invention, and does not limit the present invention in any form, and there are other variations and modifications under the premise of not exceeding the technical solution recorded in the claims.

Claims (7)

1. a high-voltage circuit breaker fault early warning method based on the power Internet of things, is characterized in that, Include the following steps: A) Obtain all historical detection data of high-voltage circuit breakers; B) Obtain a high-voltage circuit breaker of the same type, and set the fault source artificially under laboratory conditions, so that the high-voltage circuit breaker works in the presence of the fault source until the fault occurs, and at the same time, it is periodically tested to obtain test data; C) Associate the detection data N times before the failure with the corresponding failure type to form sample data, and use the sample data to train the neural network model for failure research and judgment; D) Extract the detection data between M and N times before the failure, M>N, as the precursor reference data; E) Compare the detection data of the high-voltage circuit breaker with the precursor reference data. If the gap between the detection data and the precursor reference data is less than the set threshold, a fault warning will be issued; F) Input the detection data of the high-voltage circuit breaker into the fault early warning neural network model obtained in step D). If the fault early warning neural network model outputs the fault type, a fault alarm will be issued; Wherein, in step B), a non-contact displacement sensor is installed on each mechanical moving part of the normal high-voltage circuit breaker, and under the condition of power failure, the opening and closing test of the high-voltage circuit breaker is continuously repeated until the high-voltage circuit breaker is disconnected. If the mechanical parts of the device are damaged, record the opening and closing times N during the test, and the displacement data of each mechanical moving part during the opening and closing process as the historical displacement data; determine whether the difference between the detection data and the historical detection data is greater than the preset threshold. Methods include: B31) Perform segmentation processing on the detection data and the numerical quantities in the historical detection data, and convert the numerical quantities into state quantities with the segment interval as the name; B32) Convert the state quantities in the detection data and historical detection data into Boolean quantities, and use {0,1} to represent false and true respectively; B32) Treat the Boolean value of the processed historical detection data as a numerical mean value, round the mean value to an integer, and treat the obtained integer as a Boolean value again, sort the processed detection data and the historical detection data according to the settings, respectively constitute Detection vector and historical detection vector; B33) Calculate the distance between the detection vector and the historical detection vector, compare it with the preset distance threshold, if the distance is greater than the preset distance threshold, then determine that the distance between the detection data corresponding to the detection vector and the historical detection data is greater than the preset threshold, otherwise, then It is determined that the distance between the detection data corresponding to the detection vector and the historical detection data is not greater than a preset threshold. Among them, in step C), a non-contact displacement sensor is installed on each mechanical moving part of the high-voltage circuit breaker to be judged, the high-voltage circuit breaker to be judged is opened and closed once, and the measured value of the non-contact displacement sensor is obtained. The obtained displacement data is compared with the historical displacement data, and the number of opening and closing tests n corresponding to the closest historical displacement data is obtained, and (N-n) is used as the remaining service life of the high-voltage circuit breaker to be judged; In step B31), the method for converting a numerical quantity into a state quantity with the segment interval as the name includes the following steps: Divide the numerical data into several intervals,

,in

and

are the start and end points of the numerical interval, respectively.

is the middle dividing point of the numerical interval, and the

, respectively as the state name of the corresponding numerical range; If the historical detection value of the data falls within the range

, then the state name

As the value of the numerical quantity, the conversion of numerical quantity data into state quantity data is completed. 2. A kind of high-voltage circuit breaker fault early warning method based on power Internet of things according to claim 1, is characterized in that, The detection data includes closing time, opening time, just closing speed, just opening speed, three-phase different periods, same-phase different periods, gold short time, no-current time, maximum speed of moving contact, average moving contact Speed, moving contact action time, bounce time, number of bounces, maximum bounce amplitude, opening and closing stroke, current waveform curve during opening and closing, dynamic curve of time, speed and stroke in the opening and closing stroke of moving contact, opening distance and contact resistance. 3. A kind of high-voltage circuit breaker fault early warning method based on power Internet of things according to claim 1 or 2, is characterized in that, In step B), the method for artificially setting a fault source includes the following steps: B11) Test the high-voltage circuit breaker several times; B12) According to the maintenance requirements of the high-voltage circuit breaker, select one maintenance requirement in turn to make it not up to the standard, and perform several tests after several times of live opening and closing actions; B13) Select two maintenance requirements in turn so that they do not meet the standard, and perform several tests after several times of live switching operations; B14) Use liquid nitrogen or dry ice to quickly cool the high-voltage circuit breaker, conduct several mechanical characteristic tests, and obtain the test data of the mechanical characteristic test. 4. A kind of high-voltage circuit breaker fault early warning method based on power Internet of things according to claim 3, is characterized in that, In step B, the method for establishing a fault judgment model for a high-voltage circuit breaker includes: B21) Obtain all the detection data, and associate the detection data with the corresponding fault type as sample data; B22) Preprocess the sample data, normalize it, train the neural network model, and use the trained neural network model as the fault judgment model. 5. A kind of high-voltage circuit breaker fault early warning method based on power Internet of things according to claim 4, is characterized in that, In step B21), the method for associating the detection data with the corresponding fault type includes: B211) Obtain the detection data in step B11) as historical detection data; B212) In step B12) and step B13), several sets of detection data obtained by several times of detection are sequentially compared with historical detection data, if the difference between the detection data and the historical detection data is greater than a preset threshold, then the group of detection data is not up to standard associated with the maintenance requirements; B213) Compare several sets of detection data obtained from several mechanical characteristic tests in step B14) with the historical detection data, if the difference between the detection data and the historical detection data is greater than the preset threshold, then the set of detection data and the mechanical parts are stuck and faulty association. 6. A kind of high-voltage circuit breaker fault early warning method based on power Internet of things according to claim 1, is characterized in that, In step B31), the method for segmenting the detection data and the numerical value in the historical detection data includes: B311) Select a numerical value, obtain all the values of the numerical value in the historical detection data, arrange them in order of numerical value, and record it as the set Ki, and the minimum value in the set Ki is

and the maximum value is

; B312) will partition the starting point

assign initial value to

, the partition end point

assign initial value to

, the investigation value

,

for the manually set step size,

is a positive integer,

The initial value is 1; B313)

Continuously increment by 1, if the inspection value

The following conditions are met:

Among them, the function

Represents the set Ki, and the data values are in the numerical range

the number of data, then the

As the interval dividing point and adding the dividing point set Km, the

The value assigned to

, continue to order

keep incrementing by 1 until

; B314) will

and

Add the set Km, and use the value in Km as the dividing point to divide the numerical data into numerical intervals; B315) Select the next numerical value, and repeat steps B311) to B314) until all the numerical values are divided into interval segments; B316) The detection data is divided into intervals corresponding to the numerical values in the historical detection data. 7. A kind of high-voltage circuit breaker fault early warning method based on power Internet of things according to claim 1 or 2, it is characterized in that, In step B), a non-contact displacement sensor is installed on each mechanical moving part of the normal high-voltage circuit breaker; according to the maintenance requirements of the high-voltage circuit breaker, one maintenance requirement is selected in turn so that it does not meet the standard. Repeat the opening and closing test of the high-voltage circuit breaker until the mechanical parts of the high-voltage circuit breaker are damaged; record the displacement data of each mechanical moving part during the opening and closing process during the test and correlate the corresponding maintenance requirements to the corresponding faults that do not meet the standards. Then repair the high-voltage circuit breaker, and carry out the next test that the maintenance requirements are not up to standard; In step C), a non-contact displacement sensor is installed on each mechanical moving part of the high-voltage circuit breaker to be judged, the high-voltage circuit breaker to be judged is opened and closed once, and the measured value of the non-contact displacement sensor is obtained. Displacement data, compare the displacement data measured by the non-contact displacement sensor of the high-voltage circuit breaker to be judged with the historical displacement data, and obtain the number of opening and closing tests n corresponding to the closest historical displacement data, and take (N-n) as the waiting time. The remaining service life of the high-voltage circuit breaker is judged.

Images