CN110794264B - Generator partial discharge type identification method based on time domain pulse characteristics - Google Patents
Generator partial discharge type identification method based on time domain pulse characteristics Download PDFInfo
- Publication number
- CN110794264B CN110794264B CN201910973832.4A CN201910973832A CN110794264B CN 110794264 B CN110794264 B CN 110794264B CN 201910973832 A CN201910973832 A CN 201910973832A CN 110794264 B CN110794264 B CN 110794264B
- Authority
- CN
- China
- Prior art keywords
- pulse
- discharge
- module
- generator
- time domain
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
- G01R31/1227—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R23/00—Arrangements for measuring frequencies; Arrangements for analysing frequency spectra
- G01R23/16—Spectrum analysis; Fourier analysis
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/34—Testing dynamo-electric machines
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Testing Relating To Insulation (AREA)
Abstract
The invention relates to a generator partial discharge type identification method based on time domain pulse characteristics, which comprises the following steps: 1) a VHF capacitive sensor is arranged on a three-phase outlet bus; 2) the VHF capacitive sensor is connected into the amplifying and filtering module through a high-frequency signal line; 3) a threshold filtering module is accessed behind the high-speed acquisition module; 4) connecting a windowed pulse identification module to a threshold filtering module; 5) and a classification recombination module is connected behind the windowed pulse identification module. The invention has the beneficial effects that: the method for identifying the partial discharge type of the generator based on the time domain pulse characteristics overcomes the limitations that the field interference is serious, and the interference types are different due to different environments of each field in the method for identifying the discharge spectrogram characteristics; the problems that misjudgment is easy to occur when the spectrogram of pulse type interference is close to the actual discharge spectrogram, misjudgment or misjudgment is easy to occur or missed judgment is easy to occur by using neural network algorithm identification when various discharges exist at the same time and are mutually overlapped are also solved.
Description
Technical Field
The invention relates to the technical field of insulation monitoring of large generators, in particular to a generator partial discharge type identification method based on time domain pulse characteristics.
Background
At present, insulation monitoring of a large generator is mainly based on partial discharge monitoring, the partial discharge is the most sensitive premonitory information before insulation breakdown failure of high-voltage power equipment, and specifically, under the action of an electric field, only partial area in an insulation system of the power equipment discharges without penetrating through conductors applying voltage (breakdown does not occur yet). The generator causes partial discharge for many reasons, and the general can be divided into two categories: firstly, in the production and transportation process of the generator, potential hidden dangers of the generator are caused by the process not reaching the standard, and the discharge of the end part of a generator coil bar is generally expressed; secondly, in the operation process, due to factors such as the temperature and humidity of the environment, mechanical force, friction force, dust particles, repeated and frequent starting, various transient overvoltage and the like, the insulation material in the generator is aged, air gaps are generated in the generator, partial parts fall off, partial discharge is aggravated under the action of an electric field, and the generator is finally damaged, which is generally represented as air gap discharge and groove discharge in a bar. Therefore, the generator partial discharge defect types can be divided into three main categories: internal air gap discharge of the bar, bar end discharge and slot discharge.
The retrieval finds that various methods for identifying the type of the partial discharge defect of the power equipment include: the method comprises the steps of carrying out identification on a SOM-BP neural network algorithm, a convolutional neural network algorithm, cloud computing, a fingerprint identification algorithm, a singular value decomposition algorithm, a vector machine identification algorithm, an SVM algorithm, a particle swarm neural network algorithm and the like, wherein the identification methods are based on the characteristics of a statistical discharge spectrogram (a scatter diagram is also called as a PRPS or PRPN diagram), generally selecting a plurality of to dozens of characteristic vectors of different discharge types, then training a sample library, and utilizing the different intelligent algorithms to carry out comparison identification on the discharge spectrogram tested on site and the discharge spectrogram of the sample library so as to achieve the purpose of distinguishing one discharge type. The invention discloses a method for identifying an automatic mode of a generator stator insulation partial discharge defect, which is found by searching, wherein the patent application number is 201210010000.0, the patent name is 'an automatic mode identification method of a generator stator insulation partial discharge defect', the main content of the method is that various software and hardware filtering is adopted to filter out interference, and finally, a discharge spectrogram is compared with the discharge spectrogram characteristics of a sample library, and the discharge type is identified by a neural network algorithm. The invention is characterized in that the method is provided with the patent number of 201010609194.7 and the patent name of a large motor insulation state online diagnosis and evaluation method, and the content mainly comprises the steps of building a neural network framework, training a sample library, and finally comparing a discharge spectrogram with typical characteristics of the discharge spectrogram of the sample library so as to judge which type of discharge is. The identification methods in the above two patents are the same.
In summary, the partial discharge of the generator is identified through the characteristics of the discharge spectrogram, so as to achieve the purpose of judging the type of the partial discharge defect, because the field interference is serious, the environments of each field are different, and the interference types are different, although the identification method can identify the discharge type to a certain extent, the identification method has certain limitation, when the spectrogram of the pulse type interference is close to the actual discharge spectrogram, the judgment error is easy to generate, and in addition, when multiple discharges exist simultaneously and are overlapped, the neural network algorithm is used for identifying, so that the misjudgment or the misjudgment is easy to generate. In order to solve the problems, the patent provides a time domain pulse characteristic analysis method based on deep analysis of time domain pulse characteristics and mechanisms of each partial discharge, and the purpose of identifying the partial discharge fault type is achieved.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provides a generator partial discharge type identification method based on time domain pulse characteristics.
The method for identifying the partial discharge type of the generator based on the time domain pulse characteristics specifically comprises the following steps:
step 1) installing a VHF capacitive sensor on a three-phase outlet bus, wherein the frequency band of the VHF capacitive sensor is 5 MHz-100 MHz; the signal is received by the VHF capacitive sensor on a three-phase outlet bus of the generator;
step 2) connecting the VHF capacitive sensor to an amplifying and filtering module through a high-frequency signal line; the high-speed acquisition module is connected after the amplifying and filtering module; the signal is amplified and filtered by the amplifying and filtering module and then transmitted into the high-speed acquisition module;
step 3) a threshold filtering module is accessed behind the high-speed acquisition module; threshold filtering is carried out on the data acquired by the high-speed acquisition module by using a threshold filtering module, and background noise is filtered; judging the basic discharge characteristics, if the basic discharge characteristics do not accord with the basic discharge characteristics, the discharge pulse is an interference pulse, and directly filtering the interference pulse; if the discharge pulse is consistent with the basic characteristics of the discharge pulse, the discharge pulse is a defect discharge pulse, and characteristic parameters of the discharge pulse are extracted; then combining to form a time domain pulse sequence;
step 4), after the windowed pulse identification module is connected to the threshold filtering module; inputting the time domain pulse sequences formed by combining in the step 3) into a windowed pulse identification module, and classifying the pulse sequences through the characteristic parameters and the discharge characteristics of the defect pulses which are already set in the windowed pulse identification module; if the discharge characteristics of the defect pulse are not consistent with those of the defect pulse, judging to be other discharge types;
step 5) a classification recombination module is connected behind the windowed pulse identification module; and classifying and recombining the determined discharge pulse sequences by a classifying and recombining module, and respectively forming different types of discharge spectrograms.
Preferably, the discharge characteristics of the defect pulse in step 4) include the following three types: time domain pulse characteristics of generator bar internal air gap discharge: the pulse width is between 150ns and 500ns, the rising edge is between 40ns and 100ns, the overshoot ratio is between 5 and 8, and the number of oscillation pulses is between 5 and 12; time domain pulse characteristics of generator bar end discharge: the pulse width is within T40 ns, the rising edge is less than T1 ns 15ns, the overshoot ratio is less than 3, and the number of oscillation pulses is less than 3; time domain pulse characteristics of generator wire rod groove discharge: the pulse width is more than 500ns, the rising edge is more than 100ns, the overshoot ratio is more than 8, and the number of oscillation pulses is more than 8.
The invention has the beneficial effects that:
the invention adopts a generator partial discharge type identification method based on time domain pulse characteristics, and overcomes the limitation that the field interference is serious in the identification method based on discharge spectrogram characteristics, and the interference types are different due to different environments of each field; the problems that misjudgment is easy to occur when the spectrogram of pulse type interference is close to the actual discharge spectrogram, misjudgment or misjudgment is easy to occur or missed judgment is easy to occur by using neural network algorithm identification when various discharges exist at the same time and are mutually overlapped are also solved. The method can effectively identify three discharge defect types of the starting motor, fills the technical defects of identifying the discharge type by only using discharge spectrogram characteristics, promotes the development of the partial discharge monitoring technology of the generator, and provides reliable judgment basis for preventing the occurrence of generator insulation breakdown accidents and the accurate maintenance of the generator.
Drawings
FIG. 1 is a diagram of a partial discharge on-line monitoring system of a generator;
FIG. 2 is a flow chart of a generator partial discharge type pulse feature identification process;
FIG. 3 is a diagram of a basic time domain pulse waveform of partial discharge of a generator;
FIG. 4 is a graph showing the effect of different square waves transmitted in a bar on the characteristics of the waveform;
FIG. 5 is a graph showing the effect of different pulse waves transmitted in a bar on the characteristic influence of the waveform;
FIG. 6 is a diagram of the pulse waveform measured by the monitoring system when the generator bar applies an internal discharge source;
FIG. 7 is a graph of the pulse waveform measured by the monitoring system when the generator bar applies an end discharge source;
FIG. 8 is a graph of the pulse waveform detected by the monitoring system when the generator bar applies a slot discharge source;
FIG. 9 is a discharge spectrum measured by the monitored system when the generator bar applies an internal discharge source;
FIG. 10 is a discharge spectrum measured by the monitored system when the generator bar applies an end discharge source;
FIG. 11 is a discharge spectrum measured by the monitored system when a slot discharge source is applied to the generator bar.
Description of reference numerals:
a ground terminal 1; a generator set 2; a three-phase outlet bus 3; a generator outlet box 4; a VHF capacitive sensor 5; a high-frequency signal line 6; an amplifying and filtering module 7; a high-speed acquisition module 8; a threshold filtering module 9; a windowed pulse identification module 10; and a classification and recombination module 11.
Detailed Description
The present invention will be further described with reference to the following examples. The following examples are set forth merely to aid in the understanding of the invention. It should be noted that, for a person skilled in the art, several modifications can be made to the invention without departing from the principle of the invention, and these modifications and modifications also fall within the protection scope of the claims of the present invention.
The invention is based on analyzing the pulse characteristics of the partial discharge time domain, and judges the defect type of the partial discharge, wherein different discharge types have different pulse characteristics, and three major defect types of the generator partial discharge are as follows:
the reason for forming air gap discharge in the generator bar is mainly that the generator bar is mostly made of an insulation structure which takes mica as a base material and is impregnated and cured by epoxy resin, and in the manufacturing process, if the process control is not strict and the impregnation is not sufficient, tiny air bubbles or air gaps can be left in the insulation, and in addition, in the long-term operation process, the air bubbles can be generated in the insulation due to the combined action of electric stress, thermal stress and mechanical stress. Because the dielectric constant of the insulating material is much greater than that of air, under the action of an applied voltage, the bubbles break down first, forming an air gap discharge inside the insulation. It is known from the discharge mechanism of the insulating medium that the air gap discharge inside the generator bar is mainly generated by electron impact ionization, the air gap inside the bar is the place where the electric field is most concentrated, the accelerated free electrons collide with the neutral molecules of the air gap, when the kinetic energy of the electrons is high enough, the neutral molecules will excite the electrons to form new free electrons and positive ions, the new free electrons and the original free electrons will continue to accelerate under the action of the electric field and collide with other neutral molecules, and it is possible to excite new free electrons, so the number of the free electrons increases exponentially and electron avalanches are formed, because the mass of the electrons is much smaller than that of the ions, the moving speed of the electrons is much faster than that of the ions, the electrons always face the front of the positive electrode, and the heads forming the electron avalanches continuously expand, and finally form continuous discharge. The electron impact ionization discharge is generated in the bar, the discharge process is slow, the discharge quantity is small, the pulse width is generally between 150ns and 500ns, the rising edge is between 40ns and 100ns (when the amplitude of the rising edge is 10% to 90%), the overshoot ratio is between 5 and 8, and the number of oscillation pulses is between 5 and 12.
The reason for the formation of the end discharge of the generator bar is mainly that the insulation at the joint of the end part and the head sleeve of the bar needs to be manually processed on site, the installation quality of the insulation is related to the level of field workers and is generally fixed by adopting a binding or pressing plate structure, and when the relative humidity of cooling gas of the generator is too high and the breakdown voltage is greatly reduced, the overall insulation strength of the phases is possibly insufficient to bear the line voltage, so that the end discharge of the phases is caused. It is known from the insulating medium discharge mechanism that the end discharge is mainly due to the development action of photoionization, when the field intensity in the defect is increased, firstly, electrons collide and ionize to form electron avalanches, at the moment, the electrons are concentrated at the avalanche head to strengthen the field intensity near the anode, the positive ions are concentrated at the avalanche tail to strengthen the field intensity near the cathode, when the field intensity is further strengthened, the concentration of the electrons and the ions is very high, then compound emission photons are generated, the photons enter the high field intensity regions at the two ends of the electron avalanche, neutral molecules are excited to ionize, the electrons are emitted to generate secondary electron avalanches, the secondary electron avalanches and the initial electron avalanches are converged, the plasma region is enlarged and when the two electrodes are reached, the discharge is formed, each discharge requires about 10 discharges6The pulse width of the formed time domain pulse is narrower and generally within T40 ns, the rising edge of the pulse is less than T1 15ns (when the amplitude of the rising edge is 10% to 90%), the pulse is subjected to distributed parameter resistance-capacitance filtering, the overshoot ratio is less than 3, and the number of oscillation pulses is less than 3.
The generator bar slot discharge refers to discharge between the main insulation surface of the coil and the wall of the iron core slot, and is generated because the surface of the coil slot part can not be in complete contact with the wall of the iron core slot, a gap is always reserved between the surface of the coil slot part and the wall of the iron core slot, and the electric field at the ventilation slot is not uniformly distributed, so that discharge is generated when the local electric field strength reaches a certain amount, the discharge charge is only accumulated on one side of the bar, the normal field strength and the axial field strength of the surface of the coil are increased, when the low-resistance anti-corona layer is unevenly coated or the anti-corona layer is poorly or unstably contacted with the wall of the iron core slot, a contact point is separated under the action of electromagnetic vibration, and then the high-temperature clearance spark discharge can be caused besides the electron impact discharge, and a pockmark with the depth of more than 1mm can be generated in a short time, so that the generator bar slot discharge is represented as clearance spark type discharge. The discharge pulse has obvious characteristics, the pulse width of the formed time domain pulse is wider than 500ns, the rising edge of the pulse is more than 100ns (when the amplitude of the rising edge is 10-90%), the pulse is subjected to distributed parameter resistance-capacitance filtering, the overshoot ratio is more than 8, and the number of oscillation pulses is more than 8.
The method for identifying the partial discharge type of the generator based on the time domain pulse characteristics specifically comprises the following steps:
step 1) installing a VHF capacitive sensor 5 on a three-phase outlet bus 3, wherein the frequency band of the VHF capacitive sensor 5 is 5 MHz-100 MHz; the signal is received by the VHF capacitive sensor 5 on the three-phase outlet bus 3 of the generator;
step 2) connecting the VHF capacitive sensor 5 to an amplifying and filtering module 7 through a high-frequency signal wire 6; the high-speed acquisition module 8 is connected after the amplifying and filtering module 7; the signal is amplified and filtered by the amplifying and filtering module 7 and then transmitted into the high-speed acquisition module 8;
step 3) a threshold filtering module 9 is accessed after the high-speed acquisition module 8; threshold filtering is carried out on the data acquired by the high-speed acquisition module 8 by using a threshold filtering module 9, and background noise is filtered; judging the basic discharge characteristics, if the basic discharge characteristics do not accord with the basic discharge characteristics, the discharge pulse is an interference pulse, and directly filtering the interference pulse; if the discharge pulse is consistent with the basic characteristics of the discharge pulse, the discharge pulse is a defect discharge pulse, and characteristic parameters of the discharge pulse are extracted; then combining to form a time domain pulse sequence;
step 4), after the windowed pulse identification module 10 is connected to the threshold filtering module 9; inputting the time domain pulse sequences formed by combining in the step 3) into a windowed pulse identification module 10, and classifying the pulse sequences through the characteristic parameters and the discharge characteristics of the defect pulses which are already set in the windowed pulse identification module 10; if the discharge characteristics of the defect pulse are not consistent with those of the defect pulse, judging to be other discharge types;
step 5) a classification recombination module 11 is connected behind the windowed pulse identification module 10; the determined discharge pulse sequences are classified and recombined by a classification and recombination module 11, and different types of discharge spectrograms are formed respectively.
From the above, the discharge characteristics of the defect pulse of step 4) include the following three types: time domain pulse characteristics of generator bar internal air gap discharge: the pulse width is between 150ns and 500ns, the rising edge is between 40ns and 100ns, the overshoot ratio is between 5 and 8, and the number of oscillation pulses is between 5 and 12; time domain pulse characteristics of generator bar end discharge: the pulse width is within T40 ns, the rising edge is less than T1 ns 15ns, the overshoot ratio is less than 3, and the number of oscillation pulses is less than 3; time domain pulse characteristics of generator wire rod groove discharge: the pulse width is more than 500ns, the rising edge is more than 100ns, the overshoot ratio is more than 8, and the number of oscillation pulses is more than 8.
Example (b):
the partial discharge signal is transmitted to a generator bus bar outlet along a generator bar, the VHF capacitive sensor 5 installed at the partial discharge signal is coupled to enter a test system, the pulse is influenced by distribution parameters in the transmission process and changes certain characteristics of the pulse, and the factor must be considered, so a laboratory prepares a 9m long generator bar, simulates the distance from a discharge source inside a generator to the VHF capacitive sensor 5 on the site, inputs different types of signals on one side of the bar, receives and outputs the signals by the VHF capacitive sensor 5 on the other side of the bar, as shown in FIG. 4, the waveform No. 1 of the graphs A and B is the waveform output to one side of the bar by a signal source, the waveform No. 2 is the waveform coupled to the signals by the VHF capacitive sensor 5 and output, as can be known by the graph A, a square wave with the width of T1 ns is input to one end of the bar, the output waveform width T2 of the VHF capacitive sensor 5 transmitted by the bar is 50ns, the time delay between the two waveforms is caused by the transmission of signal waves in the bar, and the amplitude is attenuated to a certain extent; as can be seen from fig. B, a square wave with a width T1-10 ns is input at one end of the bar, and the output waveform of the VHF capacitive sensor 5 transmitted through the bar has a width T2-10 ns; it can thus be concluded that: whether the pulse is a narrow pulse or a wide pulse, the pulse wave only affects the amplitude of the pulse and does not affect the width of the pulse in the transmission process of the bar. As shown in fig. 5, the waveform No. 1 in fig. a and the waveform No. 1 in fig. B are pulse waves input at one end of the line bar, the waveform No. 2 in both the diagrams is a waveform that is coupled and output by the VHF capacitive sensor 5, the rising edge time T1 of the waveform No. 1 in fig. a is 25ns, the rising edge time T2 of the waveform No. 2 output by the VHF capacitive sensor 5 is 25ns, the amplitude is attenuated, the rising edge time T1 of the waveform No. 1 in fig. B is 10ns, the rising edge time T2 of the waveform No. 2 output by the VHF capacitive sensor 5 is 10ns, and the amplitude attenuation degree is slightly larger than that in fig. a.
From the above tests, the selected judging characteristic parameters are not influenced in the transmission process of the signal in the wire rod, and the parameters can directly reflect the self characteristics of the discharge, so that the type of the partial discharge defect can be judged from the pulse characteristics.
The process of the method was verified by the laboratory: according to the laboratory verification, as shown in the connection of the figure 1, an experimental platform is built in a laboratory, a VHF capacitive sensor 5 and a signal processing system are connected according to the requirements shown in the figure 1, and one generator with three types of partial discharge simulation pulse generators is provided.
In the test verification process, all parts of the monitoring system are connected according to the diagram shown in fig. 1, the correctness of various connecting lines is confirmed, so that the system can be normally tested, three types of discharge pulse sources are respectively applied to a stator bar of the generator, and the classified discharge pulses and discharge spectrograms are tested by the observation system.
The analog pulse generator is adjusted to output analog bar internal discharge pulses, a pulse diagram tested by a system is observed, as shown in fig. 6, the pulse diagram is a typical generator bar internal air gap discharge waveform, basic characteristic parameters can be seen from the pulse diagram, the whole pulse width is T248 ns, the rising edge is T1 127ns (when the rising edge is 10% to 90% of the amplitude), the overshoot ratio is 6.23, the number of oscillation pulses is 5, and therefore the discharge pulse is judged to be generator bar internal air gap discharge.
The analog pulse generator is adjusted to output the analog bar end discharge pulse, a pulse diagram tested by the system is observed, as shown in fig. 7, the typical generator bar end discharge waveform is shown, the pulse width is T35 ns, the pulse rising edge is T1 ns 13ns (when the rising edge amplitude is 10% to 90%), the overshoot ratio is 2.35, and the number of oscillation pulses is 1. This discharge pulse is therefore determined as a generator bar end discharge.
The analog pulse generator is adjusted to output analog bar slot discharge pulses, a pulse pattern tested by the system is observed, as shown in fig. 8, the pulse width is wider than T559 ns, a pulse rising edge T1 is 210ns (when the rising edge amplitude is 10% to 90%), an overshoot ratio is 8, and the number of oscillation pulses is 8, which is typical of a generator bar slot discharge waveform.
Fig. 9, fig. 10, and fig. 11 show three types of discharge spectrograms that are finally formed, and then discharge spectrogram feature recognition based on a neural network algorithm is performed, and the result after re-recognition is consistent with the final result of this time.
Therefore, the method for judging the type of the partial discharge defect based on the analysis of the partial discharge time domain pulse waveform characteristics can effectively identify three types of the discharge defects of the starting motor, fills the technical defect of identifying the discharge type by only using the discharge spectrogram characteristics, promotes the development of the generator partial discharge monitoring technology, and provides a reliable judgment basis for preventing the occurrence of the generator insulation breakdown accident and the accurate maintenance of the generator.
Claims (2)
1. A generator partial discharge type identification method based on time domain pulse characteristics is characterized by comprising the following steps:
step 1), a VHF capacitive sensor (5) is installed on a three-phase outlet bus (3), and the frequency band of the VHF capacitive sensor (5) is 5 MHz-100 MHz; the signals are received by a VHF capacitive sensor (5) on a three-phase outlet bus (3) of the generator;
step 2), the VHF capacitive sensor (5) is connected into an amplifying and filtering module (7) through a high-frequency signal line (6); the high-speed acquisition module (8) is connected behind the amplification filtering module (7); the signal is amplified and filtered by the amplifying and filtering module (7) and then transmitted into the high-speed acquisition module (8);
step 3) a threshold filtering module (9) is accessed after the high-speed acquisition module (8); threshold filtering is carried out on the data acquired by the high-speed acquisition module (8) by using a threshold filtering module (9) to filter background noise; judging the basic discharge characteristics, if the basic discharge characteristics do not accord with the basic discharge characteristics, the discharge pulse is an interference pulse, and directly filtering the interference pulse; if the discharge pulse is consistent with the basic characteristics of the discharge pulse, the discharge pulse is a defect discharge pulse, and characteristic parameters of the discharge pulse are extracted; then combining to form a time domain pulse sequence;
step 4), after the windowed pulse identification module (10) is connected to the threshold filtering module (9); inputting the time domain pulse sequences formed by combination in the step 3) into a windowed pulse recognition module (10), and recognizing the pulse sequences by the windowed pulse recognition module (10) through the set characteristic parameters and the discharge characteristics of the defect pulses in the windowed pulse recognition module (10) by adopting a neural network algorithm to classify the pulse sequences; if the discharge characteristics of the defect pulse are not consistent with those of the defect pulse, judging to be other discharge types;
step 5), a classification recombination module (11) is connected behind the windowed pulse identification module (10); the determined discharge pulse sequences are classified and recombined by a classification and recombination module (11) and form different types of discharge spectrograms respectively.
2. The method for identifying the partial discharge type of the generator based on the time-domain pulse characteristics as claimed in claim 1, wherein the discharge characteristics of the defect pulse in the step 4) include the following three types: time domain pulse characteristics of generator bar internal air gap discharge: the pulse width is between 150ns and 500ns, the rising edge is between 40ns and 100ns, the overshoot ratio is between 5 and 8, and the number of oscillation pulses is between 5 and 12; time domain pulse characteristics of generator bar end discharge: the pulse width is within T40 ns, the rising edge is less than T1 ns 15ns, the overshoot ratio is less than 3, and the number of oscillation pulses is less than 3; time domain pulse characteristics of generator wire rod groove discharge: the pulse width is more than 500ns, the rising edge is more than 100ns, the overshoot ratio is more than 8, and the number of oscillation pulses is more than 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910973832.4A CN110794264B (en) | 2019-10-14 | 2019-10-14 | Generator partial discharge type identification method based on time domain pulse characteristics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910973832.4A CN110794264B (en) | 2019-10-14 | 2019-10-14 | Generator partial discharge type identification method based on time domain pulse characteristics |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110794264A CN110794264A (en) | 2020-02-14 |
| CN110794264B true CN110794264B (en) | 2021-11-19 |
Family
ID=69440233
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910973832.4A Active CN110794264B (en) | 2019-10-14 | 2019-10-14 | Generator partial discharge type identification method based on time domain pulse characteristics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110794264B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112130037A (en) * | 2020-09-16 | 2020-12-25 | 杭州西湖电子研究所 | Method for identifying local discharge and pulse interference based on pulse form |
| CN113189459B (en) * | 2021-04-29 | 2021-11-02 | 四川大学 | Method for detecting insulation weak point of driving motor based on three-curve method |
| CN115187527B (en) * | 2022-06-27 | 2023-04-07 | 上海格鲁布科技有限公司 | Separation and identification method for multi-source mixed ultrahigh frequency partial discharge spectrum |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101086515A (en) * | 2007-07-03 | 2007-12-12 | 西安交通大学 | Dynamotor local discharge on-line monitoring method based on double sensor directional coupling |
| CN101655538A (en) * | 2009-09-27 | 2010-02-24 | 西安博源电气有限公司 | Generator local discharge on-line monitoring device and monitoring method thereof |
| CN102565645A (en) * | 2012-01-13 | 2012-07-11 | 广东电网公司电力科学研究院 | Anti-interference on-line monitoring method for partial discharging of generator |
| CN104198898A (en) * | 2014-08-04 | 2014-12-10 | 西安交通大学 | Local discharge development process diagnosis method based on pulse-train analysis |
| CN104502821A (en) * | 2014-12-29 | 2015-04-08 | 河海大学常州校区 | Capacitive sensor based on-line switch cabinet partial discharge monitoring system and method |
| CN107167716A (en) * | 2017-07-11 | 2017-09-15 | 国网福建省电力有限公司泉州供电公司 | A kind of shelf depreciation default kind identification method and device |
| CN110244199A (en) * | 2019-05-16 | 2019-09-17 | 上海金艺检测技术有限公司 | Partial discharge pulse's statistical method based on time domain reconstruction |
-
2019
- 2019-10-14 CN CN201910973832.4A patent/CN110794264B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101086515A (en) * | 2007-07-03 | 2007-12-12 | 西安交通大学 | Dynamotor local discharge on-line monitoring method based on double sensor directional coupling |
| CN101655538A (en) * | 2009-09-27 | 2010-02-24 | 西安博源电气有限公司 | Generator local discharge on-line monitoring device and monitoring method thereof |
| CN102565645A (en) * | 2012-01-13 | 2012-07-11 | 广东电网公司电力科学研究院 | Anti-interference on-line monitoring method for partial discharging of generator |
| CN104198898A (en) * | 2014-08-04 | 2014-12-10 | 西安交通大学 | Local discharge development process diagnosis method based on pulse-train analysis |
| CN104502821A (en) * | 2014-12-29 | 2015-04-08 | 河海大学常州校区 | Capacitive sensor based on-line switch cabinet partial discharge monitoring system and method |
| CN107167716A (en) * | 2017-07-11 | 2017-09-15 | 国网福建省电力有限公司泉州供电公司 | A kind of shelf depreciation default kind identification method and device |
| CN110244199A (en) * | 2019-05-16 | 2019-09-17 | 上海金艺检测技术有限公司 | Partial discharge pulse's statistical method based on time domain reconstruction |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110794264A (en) | 2020-02-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110794264B (en) | Generator partial discharge type identification method based on time domain pulse characteristics | |
| Pirker et al. | Partial discharge measurement at DC voltage—Evaluation and characterization by NoDi∗ pattern | |
| Gao et al. | Analysis of the intrinsic characteristics of the partial discharge induced by typical defects in GIS | |
| JP2017111134A (en) | Systems and methods for monitoring electrical machines | |
| Contin et al. | Classification and separation of partial discharge signals by means of their auto-correlation function evaluation | |
| CN105548832A (en) | High-voltage power cable fault recognition method | |
| Yang et al. | Recognising multiple partial discharge sources in power transformers by wavelet analysis of UHF signals | |
| Chou et al. | Measurement and analysis of partial discharge of high and medium voltage power equipment | |
| Huecker et al. | UHF partial discharge monitoring and expert system diagnosis | |
| Nyamupangedengu et al. | Partial discharge spectral response to variations in the supply voltage frequency | |
| Nasr Esfahani et al. | Partial discharge detection and identification at low air pressure in noisy environment | |
| Zeng et al. | Typical internal defects of gas-insulated switchgear and partial discharge characteristics | |
| EP3588108A1 (en) | Method and device for use of multiple frequencies prpd pattern to reduce noise or select suitable frequency ranges for partial discharge measurements | |
| Hassan et al. | Investigating the progression of insulation degradation in power cable based on partial discharge measurements | |
| Uckol et al. | Characterization of DC corona discharge current pulses using high-frequency measurement techniques | |
| Shahsavarian et al. | Surface discharge studies of insulation materials in aviation power system under DC voltage | |
| Florkowski et al. | Partial discharge forms for DC insulating systems at higher air pressure | |
| Wang et al. | Enhanced distinction of surface and cavity discharges by trapezoid-based arbitrary voltage waveforms | |
| CN116010892A (en) | Sample database identification diagnosis method of cable insulation degradation equivalent parameter model | |
| Reid et al. | Frequency distribution of RF energy from PD sources and its application in combined RF and IEC60270 measurements | |
| Yao et al. | Statistical operators calculation of partial discharge on floating electrode defect in GIS | |
| CN110703080B (en) | GIS spike discharge diagnosis method, discharge degree identification method and device | |
| Choudhary et al. | Partial discharge pulse shape analysis for progression of insulation degradation in medium voltage cable | |
| CN113848442A (en) | Method and device for identifying particle size defects in gas insulation environment | |
| Hussain et al. | Separation of partial discharge faults in Metal-clad switchgear based on pulse shape analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |