A power transformation equipment fault detection method and device and electronic equipment
By combining the analysis of audio and vibration signals with a pre-trained model, the high cost and low efficiency of fault detection in power equipment have been solved, achieving efficient and accurate continuous detection.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GUANGDONG POWER GRID CO LTD
- Filing Date
- 2024-10-14
- Publication Date
- 2026-06-02
AI Technical Summary
Existing methods for detecting faults in power equipment rely on manual inspections, resulting in high costs, low efficiency, and low accuracy. They also fail to achieve continuous monitoring and timely fault detection.
By acquiring the operating audio and vibration data of the substation, dividing the audio data into preset time intervals, determining the stationary properties, and inputting them into a pre-trained fault detection model under certain conditions, the fault category is identified by combining audio and vibration signal analysis.
It improves the comprehensiveness and accuracy of fault identification in power equipment, reduces the amount of data processing, enables uninterrupted detection, reduces the recording of signals under stable operating conditions, and improves detection efficiency.
Smart Images
Figure CN119323974B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of information technology, and in particular to a method, apparatus, and electronic device for detecting faults in power equipment. Background Technology
[0002] The operating status of substation equipment directly affects social economic production and people's daily lives. Therefore, monitoring the operating condition of substation equipment is an important task.
[0003] Currently, manual inspection is the conventional method for detecting power equipment faults in order to ensure the safe and stable operation of substations. However, this existing method has two main drawbacks: first, it heavily relies on manual supervision, requiring significant manpower and material resources; second, it cannot achieve continuous monitoring and is highly subjective, heavily dependent on the experience of the implementers. This results in low efficiency, poor comprehensiveness, and low accuracy in detecting power equipment faults, making it impossible to locate and warn of faults in a timely manner, which can easily lead to power equipment safety issues. Summary of the Invention
[0004] This invention provides a method, apparatus, and electronic device for detecting faults in power equipment, which discards most of the signals from power equipment in a stable operating state and captures more dimensions of fault characteristics through joint analysis of audio and vibration signals, thereby improving the comprehensiveness, detection efficiency, and accuracy of power equipment fault identification.
[0005] In a first aspect, embodiments of the present invention provide a method for detecting faults in power equipment, the method comprising:
[0006] Acquire the operating audio and vibration data of the substation under test within the target time period;
[0007] The running audio data is divided into multiple audio sub-data according to a preset time interval, and the stationary properties of the running audio data are determined based on the similarity between the multiple audio sub-data.
[0008] If the stability attribute of the operating audio data meets the preset conditions, the operating audio data and the operating vibration data are input into a pre-trained fault detection model to determine the fault category of the substation under test.
[0009] Secondly, embodiments of the present invention also provide a fault detection device for power equipment, the device comprising:
[0010] The operation data acquisition module is used to acquire the operating audio data and operating vibration data of the substation under test within the target time period;
[0011] The stationary attribute segmentation module is used to divide the running audio data into multiple audio sub-data according to a preset time interval, so as to determine the stationary attribute of the running audio data based on the similarity between the multiple audio sub-data.
[0012] The fault category determination module is used to input the operating audio data and the operating vibration data into a pre-trained fault detection model to determine the fault category of the substation under test, provided that the stability attribute of the operating audio data meets preset conditions.
[0013] Thirdly, embodiments of the present invention also provide an electronic device, the electronic device comprising:
[0014] One or more processors;
[0015] Storage device for storing one or more programs.
[0016] When one or more programs are executed by one or more processors, the one or more processors implement a power equipment fault detection method as described in any of the embodiments of the present invention.
[0017] Fourthly, embodiments of the present invention also provide a storage medium containing computer-executable instructions, which, when executed by a computer processor, are used to perform a power equipment fault detection method as described in any of the embodiments of the present invention.
[0018] The technical solution of this invention allows for the acquisition of operating audio and vibration data of the substation under test during operation within a target time period. Through joint analysis of audio and vibration signals, more multi-dimensional fault characteristics can be captured, revealing some hidden faults that are difficult to detect solely through sound or vibration. This improves the comprehensiveness of substation fault identification and enhances the efficiency and accuracy of fault detection. Furthermore, the operating audio data is divided into multiple audio sub-data based on a preset time interval. The stationary properties of the operating audio data are determined based on the similarity between these sub-data. Then, when the stationary properties of the operating audio data meet preset conditions, the operating audio data and operating vibration data are input into a pre-trained fault detection model to determine the fault category of the substation under test. This embodiment determines whether to begin fault detection of the substation by using the stationary properties of the operating audio data. This allows for the discarding of most signals from the substation under stable operating conditions without overlooking complete monitoring of the equipment's operating status. This significantly reduces data processing volume and achieves uninterrupted detection, further improving the comprehensiveness, detection efficiency, and accuracy of substation fault identification. Attached Figure Description
[0019] To more clearly illustrate the technical solutions of exemplary embodiments of the present invention, the accompanying drawings used in describing the embodiments are briefly introduced below. Obviously, the accompanying drawings described are only a portion of the drawings of the embodiments to be described in this invention, and not all of the drawings. For those skilled in the art, other drawings can be obtained from these drawings without any creative effort.
[0020] Figure 1 This is a flowchart illustrating a method for detecting faults in power equipment according to an embodiment of the present invention.
[0021] Figure 2 This is a flowchart illustrating another method for detecting faults in power equipment provided in an embodiment of the present invention.
[0022] Figure 3 This is a schematic diagram of the power equipment operation data processing device involved in the embodiments of the present invention;
[0023] Figure 4 This is a schematic diagram of the casing of the power equipment operation data processing device involved in the embodiments of the present invention;
[0024] Figure 5 This is a schematic diagram of the structure of a power equipment fault detection device provided in an embodiment of the present invention;
[0025] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation
[0026] Figure 1 This is a flowchart illustrating a method for detecting faults in power equipment according to an embodiment of the present invention. This embodiment is applicable to situations where fault detection of power equipment is required. The method can be executed by a power equipment fault detection device, which can be implemented in the form of software and / or hardware. The hardware can be an electronic device, such as a mobile terminal, a PC, or a server.
[0027] like Figure 1 As shown, the fault detection method for this power equipment includes:
[0028] S110. Acquire the operating audio data and operating vibration data of the substation under test within the target time period.
[0029] Transformer equipment refers to electrical facilities used for voltage transformation, receiving, and distributing electrical energy. Transformer equipment emits sound during normal operation, and this sound changes with changes in its operating state (such as during faults or overload operation). The transformer equipment under test is the one about to undergo fault detection.
[0030] The target time period can be a time period corresponding to a preset duration. The operating audio signal is the audio signal generated by the substation under test during operation. The operating vibration data includes various parameters and indicators of the vibration of the substation under test; for example, the operating audio data includes multiple discrete operating sound signals within the target time period. Specifically, the operating vibration data may include vibration velocity, vibration acceleration, and / or vibration displacement parameters.
[0031] For example, the running audio data can be the running audio values corresponding to each sampling point within a certain time period. For instance, the running audio data can be represented as {x1, x2, x3, ..., xn}, where x1 is the running audio value at time t1, x2 is the running audio value at time t2, x3 is the running audio value at time t3, and xn is the running audio value at time tn. The running vibration data can be the vibration velocity, vibration acceleration, and vibration displacement corresponding to each sampling point within a certain time period. For instance, the running vibration data can be represented as {(v1, a1, s1), (v2, a2, s2), (v3, a3, s3)...(vn, an, sn)}, where v1 is the vibration velocity at time t1, a1 is the vibration acceleration at time t1, s1 is the vibration displacement at time t1, v2 is the vibration velocity at time t2, a2 is the vibration acceleration at time t2, s2 is the vibration displacement at time t2, and so on.
[0032] Specifically, audio data of the substation under test can be collected using an audio sensor configured within its detection range, and vibration data can be collected using a vibration sensor configured within its detection range. This allows for the periodic acquisition of audio and vibration data of the substation under test within a target time period, based on a pre-set timed task.
[0033] For example, the audio sensor and vibration sensor can be configured at associated locations on the substation under test, such as at designated locations on the side wall of the substation under test, to collect real-time audio and vibration data of the substation under test during operation. A pre-set timer task is configured to acquire audio and vibration data from the past 5 minutes (i.e., the target time period) every 5 minutes.
[0034] S120. Divide the running audio data into multiple audio sub-data according to a preset time interval, and determine the stationary properties of the running audio data based on the similarity between the multiple audio sub-data.
[0035] The preset time interval is a pre-defined time length. Audio sub-data is a portion of the running audio data, and multiple audio sub-data do not overlap.
[0036] In this context, audio data similarity refers to the degree of similarity in the features of audio sub-data. For example, the higher the similarity value of two audio sub-data, the higher the degree of similarity in their features. The specific representation of similarity can be frequency domain similarity, such as spectral similarity or amplitude similarity. Stationarity attributes characterize the stationarity of the entire segment of operating audio data. Stationarity attributes include signal stationarity attributes and signal non-stationarity attributes. In this embodiment, the purpose of determining the stationarity attribute of the operating audio data is that, in practical applications, the operating conditions of transformer equipment need to be continuously monitored, and the operating condition data during operation needs to be saved. If all data were recorded, the data volume would be enormous, requiring a large amount of storage space. Based on the fact that the operating conditions of transformer equipment are basically stable in a short period of time, when the equipment operating conditions change significantly, the corresponding audio and vibration signals will also change significantly. Therefore, for unstable operating conditions, i.e., when the stationarity attribute is signal non-stationarity, the operating audio and vibration data can be further analyzed and processed to determine the possible fault categories of the transformer equipment under test. For stable operating conditions, i.e. when the stable attribute is the signal stable attribute, there is no need to determine the fault category; it is only necessary to store the data generated during operation.
[0037] Specifically, after dividing the running audio data into multiple audio sub-data, a reference audio sub-data can be determined from these sub-data (for example, the last audio sub-data can be designated as the reference audio sub-data). All other audio sub-data are designated as comparison audio sub-data. Then, the similarity between each comparison audio sub-data and the reference audio sub-data is calculated, thus obtaining the similarity score corresponding to each comparison audio sub-data. Furthermore, if all similarities are less than or equal to a similarity threshold, the stationary attribute of the running audio data is considered a stationary signal attribute; otherwise, the stationary attribute is considered a non-stationary signal attribute.
[0038] For example, with a preset time interval of 30 seconds, and the running audio data including 5 minutes of running audio values, the running audio data can be divided into 10 non-overlapping audio sub-data segments according to the chronological order, specifically including: audio sub-data A1, audio sub-data A2, audio sub-data A3, ..., audio sub-data A10. Audio sub-data A10 can be determined as the reference audio sub-data, and the remaining audio sub-data segments are determined as comparison audio sub-data. Then, the similarity value between audio sub-data A10 and audio sub-data A1 is calculated, for example, the spectral similarity value Y1; the spectral similarity values Y2, ..., between audio sub-data A10 and audio sub-data A2 are calculated, as well as the spectral similarity value Y9 between audio sub-data A10 and audio sub-data A9. A preset similarity threshold is δ. If Y1, Y2, ..., Y9 are all greater than Y9, then the stationary attribute of the running audio data is a signal stationary attribute; otherwise, the stationary attribute of the running audio data is a signal non-stationary attribute.
[0039] S130. If the stability attribute of the running audio data meets the preset conditions, the running audio data and running vibration data are input into the pre-trained fault detection model to determine the fault category of the substation under test.
[0040] In this embodiment, the preset conditions include the stable property of the running audio data being a non-stable property.
[0041] The fault detection model is a pre-trained neural network model. Generally, for a given transformer, a louder, sharper, and more uniform buzzing sound usually indicates an excessively high power supply voltage; a buzzing sound that fluctuates in pitch but is free of other noises usually indicates a large load change; a loud and heavy buzzing sound without other noises usually indicates an overload; a loud and noisy buzzing sound, sometimes accompanied by clanging sounds, usually indicates a loose internal structure; a hissing sound usually indicates discharge caused by a dirty high-voltage bushing; and a gurgling sound may indicate a short circuit between transformer turns. Based on this, historical operating audio and vibration data, including these fault types, can be pre-acquired as training samples. The fault detection model is then trained using a large number of training samples to obtain a fully trained fault detection model. In this embodiment, corresponding to the training samples determined above, the fault attributes may include at least one of the following: the fault categories include at least one of the following: overvoltage fault category, overload fault category, excessive load change fault category, loose accessories fault category, partial discharge fault category, and short circuit fault category between conductor turns.
[0042] Specifically, if the operating audio data of the substation under test corresponds to the signal instability attribute, the operating audio data and operating vibration data can be input into a pre-trained fault detection model, and the fault detection model can output the fault category of the substation under test.
[0043] The technical solution of this invention allows for the acquisition of operating audio and vibration data of the substation under test during operation within a target time period. Through joint analysis of audio and vibration signals, more multi-dimensional fault characteristics can be captured, revealing some hidden faults that are difficult to detect solely through sound or vibration. This improves the comprehensiveness of substation fault identification and enhances the efficiency and accuracy of fault detection. Furthermore, the operating audio data is divided into multiple audio sub-data based on a preset time interval. The stationary properties of the operating audio data are determined based on the similarity between these sub-data. Then, when the stationary properties of the operating audio data meet preset conditions, the operating audio data and operating vibration data are input into a pre-trained fault detection model to determine the fault category of the substation under test. This embodiment determines whether to begin fault detection of the substation by using the stationary properties of the operating audio data. This allows for the discarding of most signals from the substation under stable operating conditions without overlooking complete monitoring of the equipment's operating status. This significantly reduces data processing volume and achieves uninterrupted detection, further improving the comprehensiveness, detection efficiency, and accuracy of substation fault identification.
[0044] Example 2
[0045] Figure 2 This is a schematic diagram of a fault detection method for power equipment provided in an embodiment of the present invention. Based on the foregoing embodiments, S120 is further refined, and the specific implementation method can be found in the technical solution of this embodiment. Technical terms that are the same as or corresponding to those in the above embodiments will not be repeated here.
[0046] like Figure 2 As shown, the method specifically includes the following steps:
[0047] S210. Acquire the operating audio data and operating vibration data of the substation under test within the target time period.
[0048] S220: Divide the running audio data into multiple audio sub-data according to a preset time interval.
[0049] S230. Determine the Mel-frequency cepstral coefficient vector corresponding to each audio sub-data.
[0050] In this embodiment, the audio sub-data includes multiple discrete running audio signals. Mel-frequency cepstral coefficients are a characteristic parameter that can simulate the structure of the human cochlea. Therefore, by calculating the Mel-frequency cepstral coefficient vector corresponding to the audio sub-data, data preparation can be provided for subsequent calculation of the spectral similarity between the audio sub-data.
[0051] Specifically, the steps for determining the Mel-spectral coefficient vector corresponding to each audio sub-data may include:
[0052] S2301. For each audio sub-data, perform modal decomposition processing on the current audio sub-data to obtain the frequency value of at least one modal component contained in the current audio sub-data.
[0053] In this embodiment, the processing procedure for each audio sub-data is the same. Here, we take any one of the audio sub-data as the current audio sub-data as an example for illustrative explanation.
[0054] Specifically, a target mode decomposition algorithm can be used to perform mode decomposition processing on the current audio sub-data to obtain at least one modal component corresponding to the current audio sub-data. Each modal component corresponds to a different frequency value, thus, the frequency values of at least one modal component corresponding to the current audio sub-data can be obtained. For example, an empirical mode decomposition algorithm can be used to perform mode decomposition processing on the current audio sub-data to obtain 8 modal components, namely the first modal component, the second modal component, ..., the eighth modal component, where the frequency value corresponding to the first modal component is H1, the frequency value corresponding to the second modal component is H2, ..., the frequency value corresponding to the eighth modal component is H8.
[0055] S2302. For each frequency value, determine the Mel frequency value corresponding to the current frequency value based on the preset frequency threshold conversion function, and determine the cepstral coefficient value corresponding to the Mel frequency value to obtain the Mel cepstral coefficient value corresponding to the current frequency value.
[0056] In this embodiment, the processing procedure for each frequency value is the same. Here, we take any one of the frequency values as the current frequency value and use the current frequency value as an example for illustrative explanation.
[0057] The expression corresponding to the preset frequency threshold conversion function is:
[0058]
[0059] Where m is the Mel frequency and f is the frequency value in Hz.
[0060] Specifically, for each frequency value, the corresponding Mel frequency value can be calculated by substituting the frequency value into formula (1), and then cepstral analysis is performed in the Mel frequency domain, including operations such as taking the logarithm and inverse transformation, and finally the Mel frequency cepstral value is obtained.
[0061] S2303. Based on the Mel-Cepstral Coefficient values corresponding to each frequency value, determine the Mel-Cepstral Coefficient Vector corresponding to the current audio sub-data.
[0062] In this embodiment, the set of all Mel-Cepstral coefficient values corresponding to each frequency value can be used as the Mel-Cepstral coefficient vector corresponding to the current audio sub-data.
[0063] S240. The last audio sub-data among multiple audio sub-data is determined as the first audio sub-data, and each audio sub-data other than the first audio sub-data is determined as the second audio sub-data.
[0064] In this embodiment, the running audio data can be divided into multiple non-overlapping audio sub-data according to the chronological order. For example, 5 minutes of running audio data can be divided into 10 audio sub-data, including 30 seconds of data. The 10th audio sub-data is the first audio sub-data, and the remaining audio sub-data are the second audio sub-data.
[0065] S250. Based on the first Mel-Cepstral Coefficient Vector of the first audio sub-data and the second Mel-Cepstral Coefficient Vector of each second audio sub-data, determine the spectral similarity between the first audio sub-data and each second audio sub-data.
[0066] In this embodiment, the specific implementation of determining spectral similarity can be as follows: for each second Mel-Cepstral Coefficient Vector, calculate the distance information between the first Mel-Cepstral Coefficient Vector and the current second Mel-Cepstral Coefficient Vector; based on the distance information, determine the spectral similarity between the first audio sub-data and the current second audio sub-data.
[0067] In this embodiment, the processing procedure for each second Mel-Cepstral coefficient vector is the same. Here, we take any one of the second Mel-Cepstral coefficient vectors as the current second Mel-Cepstral coefficient vector and use the current second Mel-Cepstral coefficient vector as an example for illustrative explanation.
[0068] For example, the current second audio sub-data X2 m The corresponding current second Mel-frequency cepstral coefficient vector can be represented as C2(n), and the first audio sub-data X1 S The corresponding first Mel-Cepstral coefficient vector can be represented as C1(n), and the distance information between the first Mel-Cepstral coefficient vector and the current second Mel-Cepstral coefficient vector can be represented as... Based on this, since the smaller the value of the distance information, the closer the two vectors are in the spatial dimension, the distance information can be numerically converted to obtain the spectral similarity between the first audio sub-data and the current second audio sub-data.
[0069] S260. Based on the similarity of each spectrum and the preset similarity threshold, determine the stationary properties of the running audio data.
[0070] In this embodiment, the stationary attribute includes: signal stationary attribute and signal non-stationary attribute. Specifically, if all spectral similarities are greater than a preset similarity threshold, the running audio data is determined to have a signal stationary attribute; if at least one spectral similarity is less than or equal to the preset similarity threshold, the running audio data is determined to have a signal non-stationary attribute.
[0071] S270. If the stability attribute of the running audio data meets the preset conditions, the running audio data and running vibration data are input into the pre-trained fault detection model to determine the fault category of the substation under test.
[0072] Furthermore, based on the above technical solutions, equipment fault warnings can be issued to target users when the fault type is determined. For example, the target user can be notified of a potential fault in the substation under test via instant messaging methods such as SMS, email, or telephone. This allows the target user to further confirm the issue on-site and take corresponding measures in advance to avoid negative consequences.
[0073] S280. If the stability attribute of the running audio data does not meet the preset conditions, at least one audio sub-data and the running vibration data segment corresponding to the audio sub-data time information are stored in the preset storage unit for the substation under test to perform operating status verification.
[0074] In this embodiment, if the operating audio data has a stable signal attribute, one audio sub-data can be randomly selected from all audio sub-data as the audio sub-data to be stored, and the audio sub-data to be stored and the corresponding operating vibration data segment based on the time information of the audio sub-data to be stored are stored in a preset storage unit. This is because, in actual production environments, the operating conditions of transformer equipment need to be continuously monitored, and the operating condition data of the transformer equipment needs to be comprehensively recorded for verification of the operating status of the transformer under test. However, if all data is recorded, the data volume is enormous, requiring a large amount of storage resources. In this embodiment, when the operating audio data has a stable signal attribute, a certain audio sub-data can represent an entire segment of operating audio data. Therefore, this segment of operating condition data can be stored, while most redundant signals are not saved. By focusing on recording signals during the operating condition change process, discarding most signals under stable operating conditions, and ensuring complete monitoring of the equipment's operating condition, not only can the audio data storage volume be greatly reduced, but the effect of uninterrupted monitoring is also achieved.
[0075] The technical solution of this invention, when determining the stationary attribute of running audio data based on the similarity between multiple audio sub-data, can determine the Mel-frequency cepstral coefficient vector corresponding to each audio sub-data. Then, the last audio sub-data among the multiple audio sub-data is determined as the first audio sub-data, and each audio sub-data other than the first audio sub-data is determined as the second audio sub-data. Further, based on the first Mel-frequency cepstral coefficient vector of the first audio sub-data and the second Mel-frequency cepstral coefficient vectors of each second audio sub-data, the spectral similarity between the first audio sub-data and each second audio sub-data is determined. Thus, based on each spectral similarity and a preset similarity threshold, the stationary attribute of the running audio data is determined. The technical solution provided in this embodiment, by determining the Mel-frequency cepstral coefficient vector of the audio sub-data, and thus determining the stationary attribute of the running audio data, improves the accuracy of the stationary attribute, and further improves the detection efficiency and accuracy of substation equipment fault identification. In addition, when the running audio data has a stationary signal attribute, most redundant signals are not saved. By focusing on recording signals during the process of changing operating conditions and discarding most signals under stable operating conditions, not only is the audio data storage size greatly reduced, but the effect of uninterrupted detection is also achieved.
[0076] Example 3
[0077] The following is a specific example illustrating the implementation of this substation fault detection method, which may include:
[0078] Operating audio and vibration data can be collected using substation operating data processing equipment, and the stationarity properties of the operating audio data can be determined. See the schematic diagram of the substation operating data processing equipment. Figure 3.like Figure 3 As shown, the power equipment operation data processing equipment is based on a data processing unit (e.g., a microcontroller) and is equipped with peripheral circuits such as buttons, vibration information acquisition modules, audio processing modules, power supply modules (this equipment can use a 5000mA / h power supply module), LCD display, LED array, and SD memory card.
[0079] The data processing unit can utilize STM series microcontrollers, which offer high pin-to-pin compatibility, high peripheral and software compatibility, and flexibility. Applications can be upgraded to require more storage space, streamlined to use less storage space, or switched to different package specifications without modifying the original framework and software.
[0080] The audio processing module is used to acquire and process audio data. For example, the sound acquisition module can employ an audio decoding chip, which includes a corresponding audio decoder and encoder. This could be a single-chip Ogg Vorbis / MP3 / AAC / WMA / MIDI audio decoder, an IMA ADPCM encoder, or a user-loaded Ogg Vorbis encoder. It includes a high-performance, patented low-power DSP processor core VS_DSP4, working data memory, serial control and input data interfaces, up to eight available general-purpose I / O pins, a UART, a high-quality variable sampling rate stereo ADC ("microphone", "line", "line + microphone", or "line * 2") and stereo DAC, a headphone amplifier, and a common voltage buffer. This module supports decoding of MP3 / WMA / OGG / WAV / FLAC / MIDI / AAC audio formats and recording of OGG / WAV audio formats. It also supports bass and treble adjustment and EarSpeaker spatial effects settings.
[0081] The vibration information acquisition module is used to collect operational vibration data. For example, the vibration information acquisition module can use a triaxial accelerometer, which can capture the acceleration changes of an object in three-dimensional space, providing a complete detection signal for detecting the vibration information of transformer equipment, and providing a reliable basis for equipment fault diagnosis, performance optimization and reliability assessment.
[0082] To display the audio recording status and set the recording time, the substation's data processing equipment is also equipped with a display screen. For example, a liquid crystal display (LCD) can be used, specifically a thin-film transistor (TFT) LCD. A TFT LCD has a semiconductor switch for each pixel, allowing each pixel to be directly controlled by a pulse. Therefore, each pixel is relatively independent and can be continuously controlled, improving the display's response speed and enabling precise control of color levels, resulting in more realistic colors. TFT LCDs are characterized by high brightness, clear display, high contrast, strong sense of depth, vibrant colors, and moderate power consumption.
[0083] In addition to the hardware mentioned above, the substation operation data processing equipment includes a housing. See the schematic diagram of the housing for the substation operation data processing equipment. Figure 4 .like Figure 4 As shown, the outer casing adopts a design concept of bevels and right angles. The bevel design on the front and top prevents rainwater from flowing into the core components inside the casing, while the right angle design on the rear ensures that the power equipment operation data processing equipment can be placed conveniently. There is also a groove on the rear, which can be used to place magnets or clips to fix the power equipment operation data processing equipment.
[0084] In practical applications, during the operation of the substation under test, audio and vibration data can be acquired through the audio processing and vibration information acquisition modules in the substation operation data processing equipment. After performing a stability analysis on the audio data, the data processing unit determines which audio and vibration data need to be stored on the SD memory card. If the stability attribute does not meet the preset conditions, the server (i.e., the computer) can read the audio and vibration data from the substation under test from the SD memory card and apply a fault detection model deployed on the computer to process the audio and vibration data to determine the fault category of the substation under test.
[0085] The power equipment operation data processing device provided in this example uses inexpensive sensors and corresponding data processing algorithms to remove a large amount of redundant data, reducing the manufacturing and maintenance costs of the detection equipment. Furthermore, audio processing can be effectively deployed within the data processing unit, minimizing computational resources.
[0086] Example 4
[0087] Figure 5 This is a schematic diagram of a fault detection device for power equipment provided in an embodiment of the present invention. The device includes: an operation data acquisition module 410, a stable attribute classification module 420, and a fault category determination module 430.
[0088] Among them, the operation data acquisition module 410 is used to acquire the operation audio data and operation vibration data of the substation under test within the target time period;
[0089] The stationary attribute segmentation module 420 is used to divide the running audio data into multiple audio sub-data according to a preset time interval, so as to determine the stationary attribute of the running audio data based on the similarity between the multiple audio sub-data.
[0090] The fault category determination module 430 is used to input the operating audio data and the operating vibration data into a pre-trained fault detection model to determine the fault category of the substation under test when the stability attribute of the operating audio data meets the preset conditions.
[0091] Based on the above-mentioned device, the optional stable attribute partitioning module 420 includes:
[0092] The cepstral vector determination unit is used to determine the Mel cepstral coefficient vector corresponding to each of the audio sub-data.
[0093] An audio data partitioning unit is used to determine the last audio sub-data in the plurality of audio sub-data as the first audio sub-data, and to determine each of the audio sub-data other than the first audio sub-data as the second audio sub-data;
[0094] The similarity determination unit is used to determine the spectral similarity between the first audio sub-data and each of the second audio sub-data based on the first Mel-cepstral coefficient vector of the first audio sub-data and the second Mel-cepstral coefficient vector of each of the second audio sub-data.
[0095] The stationary attribute determination unit is used to determine the stationary attribute of the running audio data based on the spectral similarity and a preset similarity threshold.
[0096] Based on the above-mentioned device, the optional cepstral vector determination unit includes:
[0097] The modal decomposition subunit is used to perform modal decomposition processing on the current audio subdata to obtain the frequency value of at least one modal component contained in the current audio subdata.
[0098] The coefficient value determination subunit is used to determine the Mel frequency value corresponding to the current frequency value based on a preset frequency threshold conversion function for each of the frequency values, and to determine the cepstral coefficient value corresponding to the Mel frequency value, so as to obtain the Mel cepstral coefficient value corresponding to the current frequency value.
[0099] The cepstral vector determination subunit is used to determine the Mel cepstral coefficient vector corresponding to the current audio sub-data based on the Mel cepstral coefficient values corresponding to each frequency value.
[0100] Based on the above-mentioned device, optionally, a similarity determination unit is used to calculate the distance information between the first Mel-Cepstral Coefficient Vector and the current second Mel-Cepstral Coefficient Vector for each of the second Mel-Cepstral Coefficient Vectors, so as to determine the spectral similarity between the first audio sub-data and the current second audio sub-data based on the distance information.
[0101] Based on the above-mentioned device, optionally, the stability attribute includes: signal stability attribute and signal non-stability attribute. The stability attribute determining unit is specifically used to determine that the running audio data has a signal stability attribute if all the spectral similarities are greater than a preset similarity threshold; and to determine that the running audio data has a signal non-stability attribute if at least one of the spectral similarities is less than or equal to the preset similarity threshold.
[0102] Based on the above-mentioned device, optionally, the preset condition includes the stable attribute of the running audio data being a non-stable attribute.
[0103] Based on the above-mentioned device, optionally, the fault category includes at least one of the following: overvoltage fault category, overload fault category, excessive load change fault category, loose component fault category, partial discharge fault category, and short circuit fault category between conductor turns.
[0104] Optionally, based on the above-mentioned device, the power equipment fault detection device may further include: a data storage module;
[0105] The data storage module is used to store at least one audio sub-data and the corresponding operating vibration data segment in a preset storage unit if the stability attribute of the operating audio data does not meet the preset conditions, so as to allow the substation under test to perform an operating status verification.
[0106] The technical solution of this invention allows for the acquisition of operating audio and vibration data of the substation under test during operation within a target time period. Through joint analysis of audio and vibration signals, more multi-dimensional fault characteristics can be captured, revealing some hidden faults that are difficult to detect solely through sound or vibration. This improves the comprehensiveness of substation fault identification and enhances the efficiency and accuracy of fault detection. Furthermore, the operating audio data is divided into multiple audio sub-data based on a preset time interval. The stationary properties of the operating audio data are determined based on the similarity between these sub-data. Then, when the stationary properties of the operating audio data meet preset conditions, the operating audio data and operating vibration data are input into a pre-trained fault detection model to determine the fault category of the substation under test. This embodiment determines whether to begin fault detection of the substation by using the stationary properties of the operating audio data. This allows for the discarding of most signals from the substation under stable operating conditions without overlooking complete monitoring of the equipment's operating status. This significantly reduces data processing volume and achieves uninterrupted detection, further improving the comprehensiveness, detection efficiency, and accuracy of substation fault identification.
[0107] The power equipment fault detection device provided in the embodiments of the present invention can execute the power equipment fault detection method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the method.
[0108] It is worth noting that the various units and modules included in the above system are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of each functional unit are only for easy differentiation and are not used to limit the protection scope of the embodiments of the present invention.
[0109] Example 5
[0110] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Figure 6 A block diagram is shown of an exemplary electronic device 50 suitable for implementing embodiments of the present invention. Figure 6 The electronic device 50 shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of the present invention.
[0111] like Figure 6 As shown, the electronic device 50 is represented in the form of a general-purpose computing device. The components of the electronic device 50 may include, but are not limited to: one or more processors or processing units 501, system memory 502, and bus 503 connecting different system components (including system memory 502 and processing unit 501).
[0112] Bus 503 represents one or more of several bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the various bus architectures. Examples of these architectures include, but are not limited to, the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MAC) bus, the Enhanced ISA bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnect (PCI) bus.
[0113] Electronic device 50 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by electronic device 50, including volatile and non-volatile media, removable and non-removable media.
[0114] System memory 502 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 504 and / or cache memory 505. Electronic device 50 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 506 may be used to read and write non-removable, non-volatile magnetic media (… Figure 6 Not shown; usually referred to as a "hard drive"). Although Figure 6 Not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk") and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 503 via one or more data media interfaces. Memory 502 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the embodiments of the present invention.
[0115] A program / utility 508 having a set (at least one) of program modules 507 may be stored, for example, in memory 502. Such program modules 507 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. Program modules 507 typically perform the functions and / or methods described in the embodiments of the present invention.
[0116] Electronic device 50 can also communicate with one or more external devices 509 (e.g., keyboard, pointing device, display 810, etc.), and with one or more devices that enable a user to interact with electronic device 50, and / or with any device that enables electronic device 50 to communicate with one or more other computing devices (e.g., network card, modem, etc.). This communication can be performed via input / output (I / O) interface 511. Furthermore, electronic device 50 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 512. As shown, network adapter 512 communicates with other modules of electronic device 50 via bus 503. It should be understood that, although... Figure 6 As not shown, other hardware and / or software modules may be used in conjunction with electronic device 50, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.
[0117] The processing unit 501 executes various functional applications and page processing by running programs stored in the system memory 502, such as implementing the power equipment fault detection method provided in the embodiments of the present invention.
[0118] Example 6
[0119] This invention also provides a storage medium containing computer-executable instructions, which, when executed by a computer processor, are used to perform a method for detecting faults in power equipment. The method includes:
[0120] Acquire the operating audio and vibration data of the substation under test within the target time period;
[0121] The running audio data is divided into multiple audio sub-data according to a preset time interval, and the stationary properties of the running audio data are determined based on the similarity between the multiple audio sub-data.
[0122] If the stability attribute of the operating audio data meets the preset conditions, the operating audio data and the operating vibration data are input into a pre-trained fault detection model to determine the fault category of the substation under test.
[0123] The computer storage medium of this invention can be any combination of one or more computer-readable media. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.
[0124] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.
[0125] Program code contained on a computer-readable medium may be transmitted using any suitable medium, including—but not limited to—wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.
[0126] Computer program code for performing the operations of embodiments of the present invention can be written in one or more programming languages or a combination thereof. Programming languages include object-oriented programming languages—such as Java, Smalltalk, and C++—as well as conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0127] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.
Claims
1. A method for detecting faults in power equipment, characterized in that, include: Acquire the operating audio and vibration data of the substation under test within the target time period; The running audio data is divided into multiple audio sub-data according to a preset time interval, and the stationarity attribute of the running audio data is determined based on the similarity between the multiple audio sub-data. The stationarity attribute includes signal stationarity attribute and signal non-stationarity attribute. When the stationarity attribute of the operating audio data meets a preset condition, the operating audio data and the operating vibration data are input into a pre-trained fault detection model to determine the fault category of the substation under test. The preset condition includes the stationarity attribute of the operating audio data being a signal non-stationarity attribute. The fault detection model is trained using historical operating audio data and historical operating vibration data of substation faults as training samples, and is used to determine the fault category of the substation under test based on the operating audio data and the operating vibration data. If the stability attribute of the operating audio data does not meet the preset conditions, at least one audio sub-data and the corresponding operating vibration data segment of the audio sub-data time information are stored in a preset storage unit for the substation under test to perform operating status verification. The stability attribute of the operating audio data not meeting the preset conditions includes the stability attribute of the operating audio data being a signal stability attribute.
2. The method according to claim 1, characterized in that, Similarity includes spectral similarity. The determination of the stationary properties of the running audio data based on the similarity between the multiple audio sub-data includes: Determine the Mel-spectral coefficient vector corresponding to each of the aforementioned audio sub-data; The last audio sub-data among the plurality of audio sub-data is determined as the first audio sub-data, and each of the audio sub-data except the first audio sub-data is determined as the second audio sub-data; Based on the first Mel-Cepstral Coefficient Vector of the first audio sub-data and the second Mel-Cepstral Coefficient Vector of each of the second audio sub-data, the spectral similarity between the first audio sub-data and each of the second audio sub-data is determined. Based on the aforementioned spectral similarity and preset similarity threshold, the stationary properties of the running audio data are determined.
3. The method according to claim 2, characterized in that, The audio sub-data includes multiple discrete running audio signals, and determining the Mel-spectral coefficient vector corresponding to each audio sub-data includes: For each of the aforementioned audio sub-data, modal decomposition processing is performed on the current audio sub-data to obtain the frequency value of at least one modal component contained in the current audio sub-data; For each of the frequency values, the Mel frequency value corresponding to the current frequency value is determined based on a preset frequency threshold conversion function, and the cepstral coefficient value corresponding to the Mel frequency value is determined to obtain the Mel cepstral coefficient value corresponding to the current frequency value; Based on the Mel-Cepstral Coefficient values corresponding to each frequency value, the Mel-Cepstral Coefficient Vector corresponding to the current audio sub-data is determined.
4. The method according to claim 2, characterized in that, The determination of the spectral similarity between the first audio sub-data and each of the second audio sub-data based on the first Mel-Cepstral Coefficient Vector of the first audio sub-data and the second Mel-Cepstral Coefficient Vector of each of the second audio sub-data includes: For each second Mel-Cepstral Coefficient Vector, calculate the distance information between the first Mel-Cepstral Coefficient Vector and the current second Mel-Cepstral Coefficient Vector; Based on the distance information, the spectral similarity between the first audio sub-data and the current second audio sub-data is determined.
5. The method according to claim 2, characterized in that, The determination of the stationary properties of the running audio data based on the spectral similarity and a preset similarity threshold includes: If all the spectral similarities are greater than a preset similarity threshold, then the running audio data is determined to have a stable signal attribute. If at least one of the spectral similarities is less than or equal to the preset similarity threshold, then the running audio data is determined to have a signal instability attribute.
6. The method according to claim 1, characterized in that, The fault categories include at least one of the following: overvoltage fault category, overload fault category, excessive load change fault category, loose component fault category, partial discharge fault category, and short circuit fault category between conductor turns.
7. A fault detection device for power equipment, characterized in that, include: The operation data acquisition module is used to acquire the operating audio data and operating vibration data of the substation under test within the target time period; The stationary attribute segmentation module is used to divide the running audio data into multiple audio sub-data according to a preset time interval, so as to determine the stationary attribute of the running audio data based on the similarity between the multiple audio sub-data. The stationary attribute includes signal stationary attribute and signal non-stationary attribute. A fault category determination module is used to input the operating audio data and the operating vibration data into a pre-trained fault detection model to determine the fault category of the substation under test when the stationarity attribute of the operating audio data meets preset conditions. The preset conditions include that the stationarity attribute of the operating audio data is a signal non-stationarity attribute. The fault detection model is trained using historical operating audio data and historical operating vibration data of substation faults as training samples. It is a model used to determine the fault category of the substation under test based on the operating audio data and the operating vibration data. The data storage module is used to store at least one audio sub-data and the corresponding operating vibration data segment in a preset storage unit if the stability attribute of the operating audio data does not meet the preset conditions, so as to allow the substation under test to perform an operating status verification. The stability attribute of the operating audio data not meeting the preset conditions includes the stability attribute of the operating audio data being a signal stability attribute.
8. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the power equipment fault detection method according to any one of claims 1-6.