Early fault identification and positioning method based on recording data

Abstract

The invention discloses an early fault identification and positioning method based on recording data, and relates to the technical field of power equipment fault identification, and the method comprises the steps: collecting fault recording data corresponding to an early fault, carrying out the Fourier transform of the fault recording data, obtaining a fault frequency spectrum feature, and drawing a fault spectrogram; traversing the fault spectrogram and a fault spectrogram in a historical database, and selecting a similar spectrogram as a reference graph; according to the distribution interval of the fault spectrogram, the more concentrated the distribution interval of the fault spectrogram is, the more obvious the early fault is, the higher the probability of determining the early fault is, the sampling frequency can be improved along with the development of the early fault, and a high-performance oscillograph, a data processing device and a storage device do not need to be adopted in many fault low-incidence areas. Therefore, the sampling frequency is reasonably selected according to the specific requirements and hardware conditions of the power equipment so as to balance the precision and the system performance.

Description

Technical Field

[0001] The present invention relates to the technical field of power equipment fault identification, and particularly to an early fault identification and location method based on oscillographic data. Background Art

[0002] In a power system, the early identification and precise location of faults are crucial for ensuring the stable operation of the system. Traditional fault identification methods mainly rely on the action signals of relay protection devices and the indications of fault indicators. These methods can often only identify faults after the faults have occurred and caused certain impacts. Oscillographic data, as detailed data recording the operating state of the power system, contains rich information. When certain parameters in the power system (such as current and voltage) undergo mutations or exceed preset thresholds, the oscillograph starts data acquisition.

[0003] The higher the sampling frequency, the higher the accuracy of the oscillographic data. To avoid aliasing effects, the sampling rate needs to be at least twice the highest frequency component of the signal. In a power system, the sampling rate of the fault oscillograph system generally requires reaching an order of magnitude of several thousand Hz to several tens of thousand Hz. A higher sampling frequency can capture more details of the signal, thereby improving the accuracy of the data. In wide-band oscillography, the detection frequency resolution of ultra-low-frequency / low-frequency oscillations is not lower than 0.1 Hz, and the detection frequency resolution of high-order harmonics / inter-harmonics is not lower than 1 Hz. A high sampling frequency requires higher hardware performance, such as the conversion speed of the ADC (analog-to-digital converter) and the processing ability of the DSP (digital signal processor). For example, the ZH-3 device uses a 32-bit floating-point DSP and a 16-bit A / D converter, supporting a sampling rate of 10 kHz. However, a high sampling frequency will increase the data volume, posing higher requirements for data processing and storage. Moreover, the power grid has a large coverage area and long transmission lines. Due to cost reasons, it is impossible to fully adopt high-performance oscillographs, data processing, and storage devices. Therefore, an early fault identification and location method based on oscillographic data is needed to reasonably select the sampling frequency according to the specific requirements of power equipment and hardware conditions to balance accuracy and system performance. Summary of the Invention

[0004] The technical problem solved by the present invention is that in the related art, a high sampling frequency will increase the data volume, posing higher requirements for data processing and storage. Moreover, the power grid has a large coverage area and long transmission lines. Due to cost reasons, it is impossible to fully adopt high-performance oscillographs, data processing, and storage devices. Therefore, an early fault identification and location method based on oscillographic data is needed to reasonably select the sampling frequency according to the specific requirements of power equipment and hardware conditions to balance accuracy and system performance.

[0005] To solve the above technical problem, the present invention provides the following technical solution: An early fault identification and location method based on oscillographic data, comprising: Step S1: Collect the fault recording data corresponding to the early faults, perform Fourier transform on the fault recording data to obtain the fault frequency spectrum characteristics, and draw a fault frequency spectrum diagram; Step S2: Traverse the fault frequency spectrum diagram with the fault frequency spectrum diagrams in the historical database, select the similar frequency spectrum diagrams as the reference diagrams; Step S3: Obtain the fault frequency spectrum diagram corresponding to the reference diagram, divide the sampling frequency interval according to the distribution interval of the fault frequency spectrum diagram, establish the correlation relationship between the fault frequency spectrum diagram and the sampling frequency interval, and store the correlation relationship into the control database; Step S4: Collect the recording data in real time, perform Fourier transform on the recording data to obtain the real-time frequency spectrum characteristics, and adjust the sampling frequency interval according to the frequency spectrum characteristics and the correlation relationship.

[0006] As a preferred solution of the early fault identification and location method based on the recording data of the present invention, wherein: Step S1 specifically includes: using a filter to remove the noise and unnecessary frequency components in the fault recording data, extract the key features of the fault recording data, the key features include voltage amplitude, current amplitude, power factor, current frequency and voltage frequency, convert the time-domain signal into a frequency-domain signal according to the Fourier transform, generate a fault frequency spectrum diagram in combination with the key features, and the fault frequency spectrum diagram.

[0007] As a preferred solution of the early fault identification and location method based on the recording data of the present invention, wherein: the early faults include damaged insulation of distribution lines, aging of transformer insulation, overheating of transformers, poor contact of switches, refusal to operate of switches, misoperation of switches and contact with trees.

[0008] As a preferred solution of the early fault identification and location method based on the recording data of the present invention, wherein: the early faults are corresponded to the fault frequency spectrum characteristics and the fault frequency spectrum diagrams one by one as the first relationship.

[0009] As a preferred solution of the early fault identification and location method based on the recording data of the present invention, wherein: Step S2 specifically includes: the historical database includes the early faults of various types, the fault frequency spectrum characteristics and the corresponding fault frequency spectrum diagrams stored in advance; Traverse the fault frequency spectrum diagram with the fault frequency spectrum diagrams in the historical database, calculate the similarity between the fault frequency spectrum diagram and the fault frequency spectrum diagrams in the historical database, if the similarity is greater than the preset similarity threshold and is the highest similarity among all the fault frequency spectrum diagrams, it is determined as a similar frequency spectrum diagram as the reference diagram.

[0010] As a preferred solution of the early fault identification and location method based on oscillographic data according to the present invention, wherein: the step S3 specifically includes: obtaining the spectral characteristics corresponding to the reference diagram according to the first relationship, grading the distribution intervals of the spectral characteristics according to the distribution ratio, selecting the distribution interval corresponding to half of the spectral characteristics as the first level, re-selecting the distribution interval corresponding to half of the spectral characteristics in the remaining distribution intervals as the second level, repeating the operation, and selecting the third level, the fourth level and the fifth level.

[0011] As a preferred solution of the early fault identification and location method based on oscillographic data according to the present invention, wherein: the sampling frequency levels are divided according to the performance of the oscillograph, and the sampling frequency levels include the first sampling frequency level, the second sampling frequency level, the third sampling frequency level, the fourth sampling frequency level, the fifth sampling frequency level and the sixth sampling frequency level, wherein the first sampling frequency level corresponds to the first level, the second sampling frequency level corresponds to the second level, the third sampling frequency level corresponds to the third level, the fourth sampling frequency level corresponds to the fourth level, the fifth sampling frequency level corresponds to the fifth level, and the sixth sampling frequency is used to collect the oscillographic data in real time at the conventional sampling frequency.

[0012] As a preferred solution of the early fault identification and location method based on oscillographic data according to the present invention, wherein: for the distribution interval of the spectral characteristics of the fault spectrogram, an association relationship between the fault spectrogram and the sampling frequency interval is established, and the association relationship is stored in the control database. The control database is electrically connected to the wireless signal transmitter for sending a control instruction for frequency adjustment to the oscillograph.

[0013] As a preferred solution of the early fault identification and location method based on oscillographic data according to the present invention, wherein: the step S4 specifically includes: using the oscillograph to collect the oscillographic data of the distribution line in real time, performing Fourier transform on the oscillographic data to obtain the real-time spectral characteristics, and plotting them into a real-time spectrogram, and determining the sampling frequency level according to the content of the spectral characteristics.

[0014] As a preferred solution of the early fault identification and location method based on oscillographic data according to the present invention, wherein: according to the association relationship, obtain the sampling frequency level corresponding to the sampling frequency level, and the control database sends the control instruction for frequency adjustment of the sampling frequency level to the oscillograph through the wireless transmitter, and the oscillograph adjusts the sampling frequency according to the control instruction for frequency adjustment.

[0015] Advantages of the present invention: By traversing the fault spectrogram and the fault spectrograms in the historical database, similar spectrograms are selected as reference diagrams. Screening the fault spectrograms in the historical database helps reduce the amount of data processing. Through the correlation between the fault spectrogram and the sampling frequency interval, the sampling frequency interval is adjusted according to the spectral characteristics and the correlation. Many early fault characteristics are not obvious and may not necessarily be faults. Therefore, after dividing the sampling frequency interval according to the distribution interval of the fault spectrogram, according to the distribution interval of the fault spectrogram, the more concentrated the distribution interval of the fault spectrogram indicates the more obvious the early fault, and the greater the probability of being determined as an early fault. It can increase the sampling frequency as the early fault develops, without the need to use high-performance oscillographs, data processing, and storage devices in many low-fault-occurrence areas. Thus, according to the specific requirements and hardware conditions of the power equipment, the sampling frequency is reasonably selected to balance accuracy and system performance. Description of the Drawings

[0016] Figure 1 It is a schematic diagram of the basic process of an early fault identification and location method based on oscillographic data provided by an embodiment of the present invention. Detailed Embodiments

[0017] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following detailed description of the specific embodiments of the present invention is provided in conjunction with the drawings of the specification. Obviously, the described embodiments are some embodiments of the present invention, rather than all embodiments.

[0018] Embodiment 1, referring to Figure 1 , which is an embodiment of the present invention, provides an early fault identification and location method based on oscillographic data, including: Step S1: Collect the fault oscillographic data corresponding to the early fault, perform Fourier transform on the fault oscillographic data to obtain the fault spectral characteristics, and draw a fault spectrogram; Step S2: Traverse the fault spectrogram and the fault spectrograms in the historical database, and select similar spectrograms as reference diagrams; Step S3: Obtain the fault spectrogram corresponding to the reference diagram, divide the sampling frequency interval according to the distribution interval of the fault spectrogram, establish the correlation between the fault spectrogram and the sampling frequency interval, and store the correlation in the control database; Step S4: Collect the oscillographic data in real time, perform Fourier transform on the oscillographic data to obtain the real-time spectral characteristics, and adjust the sampling frequency interval according to the spectral characteristics and the correlation.

[0019] Preferably, in this embodiment, by traversing the fault spectrogram and the fault spectrograms in the historical database, similar spectrograms are selected as reference diagrams. Screening the fault spectrograms in the historical database helps reduce the amount of data processing. According to the correlation relationship between the fault spectrogram and the sampling frequency interval, the sampling frequency interval is adjusted based on the spectral characteristics and the correlation relationship. Many early fault characteristics are not obvious and may not necessarily be faults. Therefore, after dividing the sampling frequency interval according to the distribution interval of the fault spectrogram, according to the distribution interval of the fault spectrogram, the more concentrated the distribution interval of the fault spectrogram indicates the more obvious the early fault, and the greater the probability of being determined as an early fault. It can increase the sampling frequency as the early fault develops, without the need to use high-performance recorders, data processing, and storage devices in many low-fault-occurrence areas. Thus, according to the specific requirements and hardware conditions of the power equipment, the sampling frequency is reasonably selected to balance accuracy and system performance, so as to achieve more efficient identification of early faults, reduce the identification time, and improve the identification efficiency.

[0020] Embodiment 2 is another embodiment of the present invention. The difference between this embodiment and the first embodiment is that step S1 specifically includes: using a filter to remove noise and unwanted frequency components in the fault recording data, extracting the key features of the fault recording data. The key features include voltage amplitude, current amplitude, power factor, current frequency, and voltage frequency. According to the Fourier transform, the time-domain signal is converted into a frequency-domain signal, and a fault spectrogram is generated in combination with the key features, and the fault spectrogram.

[0021] Preferably, in this embodiment, a band-pass filter is used to remove noise and irrelevant frequency components in the fault recording data . When calculating the similarity between the fault spectrogram and the fault spectrograms in the historical database, the Euclidean distance is used as the measurement standard for similarity. The specific calculation formula is: Among them, represents the i-th spectral feature value of the current fault spectrogram, represents the i-th spectral feature value of the fault spectrogram in the historical database, and n is the total number of spectral features.

[0022] Early faults include damaged distribution line insulation, aged transformer insulation, overheated transformer, poor switch contact, switch refusal to operate, switch misoperation, and tree contact.

[0023] The early faults are corresponded one by one with the fault spectral features and the fault spectrogram as the first relationship.

[0024] Step S2 specifically includes: the historical database includes various types of pre-stored early faults and fault spectral features, as well as the corresponding fault spectrograms; Traverse the fault spectrogram and the fault spectrograms in the historical database, calculate the similarity between the fault spectrogram and the fault spectrograms in the historical database. If the similarity is greater than the preset similarity threshold and is the highest among all fault spectrograms, it is determined as a similar spectrogram and used as a reference diagram.

[0025] Step S3 specifically includes: obtaining the spectral features corresponding to the reference diagram according to the first relationship, grading the distribution intervals of the spectral features according to the distribution ratio, selecting the distribution interval corresponding to half of the spectral features as the first level, and re-selecting the distribution interval corresponding to half of the spectral features in the remaining distribution intervals as the second level, repeating the operation to select the third level, the fourth level and the fifth level.

[0026] Preferably in this embodiment, the initial sampling frequency interval is , where represents the lowest sampling frequency, is the highest sampling frequency.

[0027] Calculate the distribution ratio p of the spectral features of the fault spectrogram within the current sampling frequency interval.

[0028] If p≥0.5, divide the current sampling frequency interval into the first level; otherwise, divide the current sampling frequency interval into the second level.

[0029] In the remaining sampling frequency intervals, repeat steps 2 and 3 to divide the third level, the fourth level and the fifth level in turn.

[0030] Divide the sampling frequency levels according to the performance of the recorder. The sampling frequency levels include the first sampling frequency level, the second sampling frequency level, the third sampling frequency level, the fourth sampling frequency level, the fifth sampling frequency level and the sixth sampling frequency level. Among them, the first sampling frequency level corresponds to the first level, the second sampling frequency level corresponds to the second level, the third sampling frequency level corresponds to the third level, the fourth sampling frequency level corresponds to the fourth level, the fifth sampling frequency level corresponds to the fifth level, and the sixth sampling frequency is used as the conventional sampling frequency to collect the recording data in real time.

[0031] For the distribution interval of the spectral features of the fault spectrogram, establish the association relationship between the fault spectrogram and the sampling frequency interval, store the association relationship in the control database, and the control database is electrically connected to the wireless signal transmitter for sending the control instruction of frequency adjustment to the recorder.

[0032] Preferably, in this embodiment, by traversing the fault spectrogram and the fault spectrograms in the historical database, similar spectrograms are selected as reference diagrams. Screening the fault spectrograms in the historical database helps reduce the amount of data processing. Classification is performed according to the distribution ratio based on the distribution range of the spectral characteristics. For the distribution range of the spectral characteristics of the fault spectrogram, an association relationship between the fault spectrogram and the sampling frequency range is established, and the association relationship is stored in the control database. The control database is electrically connected to the wireless signal transmitter for sending a control instruction for frequency adjustment to the recorder, which helps further reduce the amount of data processing, thereby realizing efficient data processing. There is no need to use high-performance recorders, data processing, and storage devices in many low-fault areas. Therefore, according to the specific requirements and hardware conditions of the power equipment, the sampling frequency can be reasonably selected to balance accuracy and system performance. In low-fault areas, recorders, data processing, and storage devices with lower performance can also complete the early fault identification work.

[0033] Preferably, in this embodiment, after dividing the sampling frequency range according to the distribution range of the fault spectrogram, according to the distribution range of the fault spectrogram, the more concentrated the distribution range of the fault spectrogram indicates that the early fault is more obvious, and the greater the probability of being determined as an early fault. It can increase the sampling frequency with the development of the early fault. There is no need to use high-performance recorders, data processing, and storage devices in many low-fault areas. Therefore, according to the specific requirements and hardware conditions of the power equipment, the sampling frequency can be reasonably selected to balance accuracy and system performance.

[0034] Step S4 specifically includes: using a recorder to collect the recording data of the distribution line in real time, performing Fourier transform on the recording data to obtain real-time spectral characteristics, and plotting them into a real-time spectrogram, and determining the sampling frequency level according to the content of the spectral characteristics.

[0035] Preferably, in this embodiment, according to the recorder number information and location information, combined with the recording data collected by the recorder, the area of the distribution line identified as having an early fault is used as the fault area, and the fault is located by combining the recorder number information and location information in the fault area.

[0036] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. Among them, the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. These computer program instructions can also be stored in a computer-readable memory capable of guiding a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured article including an instruction device, and the instruction device implements the functions specified in one process Figure 1 one process or multiple processes and / or boxes Figure 1 functions specified in one box or multiple boxes.

[0037] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. An early fault identification and location method based on oscillogram data, characterized in that, Including: Step S1: Collect fault oscillogram data corresponding to early faults, perform Fourier transform on the fault oscillogram data to obtain fault spectrum characteristics, and draw a fault spectrum diagram; Step S2: Traverse the fault spectrum diagram with the fault spectrum diagrams in the historical database, select similar spectrum diagrams as the reference diagrams; Step S3: Obtain the fault spectrum diagram corresponding to the reference diagram, divide the sampling frequency interval according to the distribution interval of the fault spectrum diagram, establish the correlation relationship between the fault spectrum diagram and the sampling frequency interval, and store the correlation relationship in the control database; Step S4: Collect oscillogram data in real time, perform Fourier transform on the oscillogram data to obtain real-time spectrum characteristics, and adjust the sampling frequency interval according to the spectrum characteristics and the correlation relationship.

2. The early fault identification and location method based on oscillogram data according to claim 1, wherein: The specific content of step S1 includes: Using a filter to remove noise and unnecessary frequency components in the fault oscillogram data, extracting the key characteristics of the fault oscillogram data, the key characteristics include voltage amplitude, current amplitude, power factor, current frequency and voltage frequency, converting the time-domain signal into a frequency-domain signal according to Fourier transform, generating a fault spectrum diagram in combination with the key characteristics, and the fault spectrum diagram.

3. The early fault identification and location method based on oscillogram data according to claim 2, characterized in that: The early faults include distribution line insulation damage, transformer insulation aging, transformer overheating, switch poor contact, switch refusal to operate, switch misoperation and tree contact.

4. The early fault identification and location method based on oscillogram data according to claim 3, characterized in that: One-to-one correspondence is established between the early faults, the fault spectrum characteristics, and the fault spectrum diagram as the first relationship.

5. The early fault identification and location method based on oscillogram data according to claim 4, characterized in that: The specific content of step S2 includes: The historical database includes various types of the early faults, the fault spectrum characteristics, and the corresponding fault spectrum diagrams stored in advance; Traverse the fault spectrum diagram with the fault spectrum diagrams in the historical database, calculate the similarity between the fault spectrum diagram and the fault spectrum diagrams in the historical database. If the similarity is greater than the preset similarity threshold and is the highest similarity among all fault spectrum diagrams, it is determined as a similar spectrum diagram and used as the reference diagram.

6. The early fault identification and location method based on oscillogram data according to claim 5, characterized in that: The specific content of step S3 includes: Obtain the spectrum characteristics corresponding to the reference diagram according to the first relationship, classify the distribution intervals according to the spectrum characteristics according to the distribution ratio, select the distribution interval corresponding to half of the spectrum characteristics as the first level, re-select the distribution interval corresponding to half of the spectrum characteristics in the remaining distribution intervals as the second level, repeat the operation, and select the third level, the fourth level and the fifth level.

7. The early fault identification and location method based on oscillogram data according to claim 6, characterized in that: Divide the sampling frequency levels according to the performance of the oscillograph. The sampling frequency levels include the first sampling frequency level, the second sampling frequency level, the third sampling frequency level, the fourth sampling frequency level, the fifth sampling frequency level and the sixth sampling frequency level. Among them, the first sampling frequency level corresponds to the first level, the second sampling frequency level corresponds to the second level, the third sampling frequency level corresponds to the third level, the fourth sampling frequency level corresponds to the fourth level, the fifth sampling frequency level corresponds to the fifth level, and the sixth sampling frequency is used as the conventional sampling frequency to collect oscillogram data in real time.

8. The early fault identification and location method based on oscillogram data according to claim 7, wherein: The distribution range of the spectral characteristics of the fault spectrogram is used to establish the correlation between the fault spectrogram and the sampling frequency range, and the correlation is stored in the control database. The control database is electrically connected to the wireless signal transmitter for sending the control instruction for frequency adjustment to the oscillograph.

9. The early fault identification and location method based on oscillogram data according to claim 8, characterized in that: Step S4 specifically includes: using the oscillograph to collect the oscillogram data of the distribution line in real time, performing Fourier transform on the oscillogram data to obtain the real-time spectral characteristics, plotting them into a real-time spectrogram, and determining the sampling frequency level according to the content of the spectral characteristics.

10. The early fault identification and location method based on oscillogram data according to claim 9, characterized in that: According to the correlation, the sampling frequency corresponding to the sampling frequency level is obtained. The control database sends the control instruction for frequency adjustment of the sampling frequency level to the oscillograph through the wireless transmitter, and the oscillograph adjusts the sampling frequency according to the control instruction for frequency adjustment.

Images