Abnormal Detection Method, Device, Medium and System for Power Equipment

Abstract

The present application relates to the technical field of power equipment, and discloses an abnormal detection method, device, medium and system for power equipment. The method includes: acquiring ultrasonic data and non-ultrasonic data collected at the power equipment site, where the non-ultrasonic data is audio data with a sound frequency band different from that of the ultrasonic data; respectively performing feature extraction on the ultrasonic data and the non-ultrasonic data to extract at least one ultrasonic feature from the ultrasonic data and at least one non-ultrasonic feature from the non-ultrasonic data; mapping each ultrasonic feature to an ultrasonic feature score and mapping each non-ultrasonic feature to a non-ultrasonic feature score; obtaining a total feature score based on each ultrasonic feature score and each non-ultrasonic feature score; and obtaining an abnormal detection result for the power equipment according to the total feature score. This method can improve the accuracy of abnormal detection while maintaining high economy.

Description

Technical Field

[0001] This application relates to the technical field of power equipment, and particularly to an abnormal detection method, device, medium and system for power equipment. Background Art

[0002] In a distribution network, there are a large number of power equipment such as switch cabinets. Whether the power equipment such as switch cabinets can operate stably and normally is an important factor related to whether the power grid can operate safely and reliably.

[0003] Abnormalities such as mechanical vibration and partial discharge will occur in power equipment, and these abnormalities are important factors that cause power equipment to malfunction.

[0004] Currently, some abnormal detection methods require a large number of detection devices to be deployed on site. The devices are numerous, large in size, and high in cost, and the accuracy cannot be guaranteed; if the deployment of detection devices is reduced, the detection accuracy will be further reduced. Summary of the Invention

[0005] In the technical field of power equipment, in order to solve the technical problems of high detection cost and low detection accuracy of existing abnormal detection methods, the purpose of this application is to provide an abnormal detection method, device, medium and system for power equipment.

[0006] According to one aspect of this application, an abnormal detection method for power equipment is provided. The method includes:

[0007] Obtain ultrasonic data and non-ultrasonic data collected at the power equipment site, where the non-ultrasonic data is audio data with a sound frequency band different from that of the ultrasonic data;

[0008] Extract features from the ultrasonic data and the non-ultrasonic data respectively, so as to extract at least one ultrasonic feature from the ultrasonic data and at least one non-ultrasonic feature from the non-ultrasonic data;

[0009] Map each ultrasonic feature to an ultrasonic feature score, and map each non-ultrasonic feature to a non-ultrasonic feature score;

[0010] Obtain a total feature score according to each ultrasonic feature score and each non-ultrasonic feature score;

[0011] Obtain an abnormal detection result for the power equipment according to the total feature score.

[0012] According to another aspect of this application, an abnormal detection device for power equipment is provided. The device includes:

[0013] An acquisition module, configured to acquire ultrasonic data and non-ultrasonic data collected at the site of a power device, where the non-ultrasonic data is audio data with a sound frequency band different from that of the ultrasonic data;

[0014] A feature extraction module, configured to respectively perform feature extraction on the ultrasonic data and the non-ultrasonic data, so as to extract at least one ultrasonic feature from the ultrasonic data and at least one non-ultrasonic feature from the non-ultrasonic data;

[0015] A mapping module, configured to map each ultrasonic feature to an ultrasonic feature score and map each non-ultrasonic feature to a non-ultrasonic feature score;

[0016] A calculation module, configured to obtain a total feature score according to each ultrasonic feature score and each non-ultrasonic feature score;

[0017] A judgment module, configured to obtain an abnormal detection result of the power device according to the total feature score.

[0018] According to another aspect of the present application, there is provided a computer-readable medium, on which a computer program is stored, and when the computer program is executed by a processor, the method described in the above embodiment is implemented.

[0019] According to another aspect of the present application, there is provided an abnormal detection system for a power device, including:

[0020] An acquisition module, configured to acquire ultrasonic data and non-ultrasonic data at the site of a power device, where the non-ultrasonic data is audio data with a sound frequency band different from that of the ultrasonic data;

[0021] A data processing module, configured to process the ultrasonic data and the non-ultrasonic data to obtain at least one ultrasonic feature and at least one non-ultrasonic feature;

[0022] A discrimination and early warning module, configured to analyze the at least one ultrasonic feature and the at least one non-ultrasonic feature to obtain an abnormal detection result, and issue an alarm signal corresponding to the abnormal detection result.

[0023] The technical solution provided by the embodiment of the present application may include the following beneficial effects:

[0024] For the abnormal detection method, device, medium and system of the power equipment provided by this application, the method includes the following steps: obtaining ultrasonic data and non-ultrasonic data collected at the power equipment site, where the non-ultrasonic data is audio data with a sound frequency band different from that of the ultrasonic data; respectively extracting features from the ultrasonic data and the non-ultrasonic data to extract at least one ultrasonic feature from the ultrasonic data and at least one non-ultrasonic feature from the non-ultrasonic data; mapping each ultrasonic feature to an ultrasonic feature score and mapping each non-ultrasonic feature to a non-ultrasonic feature score; obtaining a total feature score according to each ultrasonic feature score and each non-ultrasonic feature score; obtaining an abnormal detection result for the power equipment according to the total feature score.

[0025] Under this method, by first obtaining the ultrasonic data and non-ultrasonic data collected at the power equipment site, then respectively extracting at least one ultrasonic feature and at least one non-ultrasonic feature from the ultrasonic data and the non-ultrasonic data, then mapping each ultrasonic feature and each non-ultrasonic feature to the corresponding feature scores respectively, and determining the total feature score, and finally obtaining the abnormal detection result according to the total feature score. Therefore, in the solution of the embodiment of this application, the ultrasonic features and non-ultrasonic features extracted from the ultrasonic data are used to perform abnormal detection on the power equipment. The sound frequency band of the non-ultrasonic data is different from that of the ultrasonic data. Therefore, the non-ultrasonic features can additionally provide the information required for abnormal detection, thus greatly improving the accuracy of abnormal detection; at the same time, since this solution only needs to deploy detection equipment for collecting ultrasonic data and non-ultrasonic data, without additionally deploying a large number of detection equipment, the overall solution is simple to implement, has a low cost, good economy, and is easy to popularize.

[0026] It should be understood that the above general description and the following detailed description are only exemplary and do not limit the present invention. Brief Description of the Drawings

[0027] The drawings here are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with the present invention and used together with the specification to explain the principles of the present invention.

[0028] Figure 1 is a schematic diagram of the system architecture of an abnormal detection method for a power equipment shown according to an exemplary embodiment;

[0029] Figure 2 is a flowchart of an abnormal detection method for a power equipment shown according to an exemplary embodiment;

[0030] Figure 3 is a flowchart of extracting at least one ultrasonic feature from ultrasonic data shown according to an exemplary embodiment;

[0031] Figure 4A flowchart for extracting at least one non-ultrasonic feature from non-ultrasonic data according to an exemplary embodiment;

[0032] Figure 5 A schematic diagram of the overall process according to an exemplary embodiment;

[0033] Figure 6 A block diagram of an abnormal detection device for a power device according to an exemplary embodiment;

[0034] Figure 7 A schematic diagram of the structure of a computer system of an electronic device suitable for implementing the embodiments of the present application is shown. Detailed implementation manners

[0035] Here, the exemplary embodiments will be described in detail, and the examples are shown in the accompanying drawings. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with the present invention. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0036] In addition, the accompanying drawings are only schematic diagrams of the present application and are not necessarily drawn to scale. The same reference numerals in the drawings represent the same or similar parts, and thus repeated descriptions thereof will be omitted. Some of the block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities.

[0037] Although power devices such as switchgear have a certain tolerance, a series of problems such as component loosening and insulation material aging will inevitably occur during long-term operation. In particular, the occurrence of partial discharge will have a huge impact on the insulation condition and operation condition of power devices, and in severe cases, it will cause accidents, endanger personal safety, and cause large-scale power outages of the power grid.

[0038] Therefore, it is necessary to detect the partial discharge of power devices.

[0039] Traditional partial discharge detection methods are costly and often require combined detection of multiple methods such as sound, electricity, and temperature. The equipment is numerous and large in size, which is not conducive to on-site real-time monitoring and is not an effective and economical choice under complex on-site operating conditions.

[0040] In the related art, there is also a partial discharge detection method called the ultrasonic method. Specifically, the main detection frequency band of the ultrasonic method is 20 kHz - 200 kHz. The detection method of the ultrasonic method is generally to place several ultrasonic sensors outside the cavity of the GIS (Gas Insulated Substation) equipment, and detect the discharge source by analyzing the ultrasonic audio.

[0041] However, since the ultrasonic method detects high-frequency signals such as ultrasonic waves, the natural wind at the detection site will generate high-frequency sound waves, which will also cause interference. The air and the power equipment itself will also generate electromagnetic interference. Therefore, using only the ultrasonic method has high environmental requirements, is easily affected by high-frequency signal interference, has poor anti-mechanical vibration interference ability, and low detection accuracy.

[0042] The inventor of the present application found that in power equipment, under the action of electrodynamic force, component loosening will cause mechanical vibration and then generate sound; due to the presence of tiny bubbles and air gaps in the insulating material, audible sound and ultrasound will be generated during partial discharge.

[0043] Based on this, the present application first provides an abnormal detection method for power equipment. Through this method, the above defects can be overcome. By extracting ultrasonic features and non-ultrasonic features from ultrasonic data to detect abnormalities in power equipment, it can not only efficiently and accurately monitor abnormal phenomena such as partial discharge in power equipment, but also has low cost, good economy, and is easy to promote.

[0044] The implementation terminal of the present application can be any device with computing, processing, and communication functions. This device can be connected to external devices for receiving or sending data. Specifically, it can be a portable mobile device, such as a smart phone, a tablet computer, a laptop computer, a PDA (Personal Digital Assistant), etc., or a fixed device, such as a computer device, a field terminal, a desktop computer, a server, a workstation, etc. It can also be a collection of multiple devices, such as the physical infrastructure of cloud computing or a server cluster.

[0045] Optionally, the implementation terminal of the present application can be a server or the physical infrastructure of cloud computing.

[0046] Figure 1 It is a schematic diagram of the system architecture of an abnormal detection method for power equipment shown according to an exemplary embodiment. As Figure 1As shown, the system architecture 100 includes: a server 110, an alarm light 120, a collection-side module 130, a smart phone 140, and a switchgear 150. Among them, the alarm light 120, the collection-side module 130, and the smart phone 140 are all communicatively connected to the server 110. The collection-side module 130 is located at the collection site near the switchgear 150. The collection-side module 130 includes an ultrasonic sensor 131 and an audible sound sensor 132. The server 110 is the execution terminal of the embodiment of the present application. When an abnormal detection method for a power device provided by the embodiment of the present application is applied to Figure 1 the system architecture shown, a process can be as follows: First, the collection-side module 130 collects ultrasonic signals near the switchgear 150 through the ultrasonic sensor 131, and collects audible sound signals near the switchgear 150 through the audible sound sensor 132. The collection-side module 130 converts the ultrasonic signals into ultrasonic data and converts the audible sound signals into audible sound data. Then, the collection-side module 130 sends the ultrasonic data and the audible sound data to the server 110 through a communication link. The server 110 extracts at least one ultrasonic feature and at least one non-ultrasonic feature from the ultrasonic data and the audible sound data respectively, and generates a feature score corresponding to each feature. Finally, a total feature score is generated based on the feature scores corresponding to each feature. Finally, the server 110 obtains an abnormal detection result according to the total feature score, and sends a corresponding alarm signal to the alarm light 120 based on the abnormal detection result, so that the alarm light 120 displays a color matching the abnormal detection result. At the same time, the server 110 also generates a corresponding warning message based on the abnormal detection result and sends the warning message to the smart phone 140 to remind the on-site staff using the smart phone 140 to perform maintenance immediately.

[0047] In an embodiment of the present application, if the abnormal detection result is abnormal, the alarm light 120 will display red.

[0048] Specifically, the abnormal detection method for the power device is used to detect whether the switchgear has local leakage. When the switchgear has local leakage, the abnormal detection result is abnormal.

[0049] In an embodiment of the present application, after receiving the audible sound data, the server 110 will also send a corresponding alarm signal to the alarm light 120 according to the signal amplitude of the audible sound data to instruct the alarm light 120 to display a corresponding color.

[0050] Specifically, if the signal amplitude of the audible sound data exceeds a predetermined signal amplitude threshold, the server 110 instructs the alarm light 120 to display a yellow light; if the signal amplitude of the audible sound data does not exceed the predetermined signal amplitude threshold, the server 110 instructs the alarm light 120 to display a green light.

[0051] In one embodiment of the present application, the acquisition side module 130 further includes an amplifier and an analog-to-digital converter. The ultrasonic signal and the audible signal are amplified by the amplifier and then input into the analog-to-digital converter, and the analog quantity is converted into a digital quantity through the analog-to-digital converter to obtain ultrasonic data and audible data.

[0052] It is worth mentioning that Figure 1 This is only one embodiment of the present application. Although in Figure 1 the embodiment, the solution is used for abnormal detection of switch cabinets, in other embodiments of the present application, abnormal detection can also be performed on various other power equipment; although in Figure 1 the embodiment, the alarm method is to control the color of the light and send an alarm message to the user's smartphone, but in other embodiments of the present application, various other types of alarm methods can also be adopted. For example, an alarm can be issued by controlling the buzzer to emit a sound or by controlling the flashing of the light; although in Figure 1 the embodiment, the execution terminal is a server, and the execution terminal and the acquisition side module are separately deployed, but in other embodiments of the present application, the execution terminal can also be various types of terminal devices such as a desktop computer or a laptop computer, and the execution terminal can also be deployed together with the acquisition side module, that is, the execution terminal can also be deployed at the acquisition site close to the switch cabinet 150; although in Figure 1 the embodiment, the acquisition side module 130 collects signals from the outside of the switch cabinet 150, but in other embodiments of the present application, the acquisition side module 130 can also be located inside the switch cabinet 150 to improve the accuracy of signal acquisition; the present application makes no limitation in this regard, and the protection scope of the present application should not be limited thereby.

[0053] Figure 2 It is a flowchart of a method for abnormal detection of a power equipment according to an exemplary embodiment. Logically, the method for abnormal detection of a power equipment provided by the embodiments of the present application can be executed by an abnormal detection system of the power equipment; physically, the method can be executed by an electronic device such as a server, as Figure 2 shown, including the following steps:

[0054] Step 210, obtaining ultrasonic data and non-ultrasonic data collected at the power equipment site, where the non-ultrasonic data is audio data with a sound frequency band different from that of the ultrasonic data.

[0055] The power equipment can be a switch cabinet. The power equipment site can be any location where the sound data of the power equipment can be collected, such as a location close to the power equipment or the inside of the power equipment.

[0056] The ultrasonic data is audio data in the frequency band of 20 kHz - 200 kHz collected by an ultrasonic sensor; the non-ultrasonic data is audio data with a frequency lower than 20 kHz, such as audible sound data that can be heard by the human ear and belongs to the frequency range of 20 Hz - 20,000 Hz. Therefore, the non-ultrasonic data can be collected by an audible sound sensor.

[0057] In an embodiment of the present application, the non-ultrasonic data and the ultrasonic data are data collected within the same time period.

[0058] By collecting non-ultrasonic data and ultrasonic data within the same time period in the present application, the abnormality of the power equipment can be detected in a timely and accurate manner, avoiding the problem of inaccurate detection caused by the change of the abnormal state of the power equipment.

[0059] The ultrasonic data and the non-ultrasonic data can be respectively collected in real time by an ultrasonic sensor and an audible sound sensor, or can be obtained by extracting the audio files collected by the ultrasonic sensor and the audible sound sensor by using professional software.

[0060] In an embodiment of the present application, the obtaining of the ultrasonic data and the non-ultrasonic data collected at the power equipment site includes:

[0061] Obtain the ultrasonic raw data and the non-ultrasonic raw data collected at the power equipment site;

[0062] Perform normalization processing on the ultrasonic raw data and the non-ultrasonic raw data respectively to obtain the normalized ultrasonic raw data and the normalized non-ultrasonic raw data;

[0063] Perform segmentation processing on the normalized ultrasonic raw data and the normalized non-ultrasonic raw data respectively to obtain multiple segments of ultrasonic data and multiple segments of non-ultrasonic data;

[0064] Obtain data pairs from the multiple segments of ultrasonic data and the multiple segments of non-ultrasonic data, where the data pair includes a segment of ultrasonic data obtained from the multiple segments of ultrasonic data and a segment of non-ultrasonic data obtained from the multiple segments of non-ultrasonic data.

[0065] Specifically, both the ultrasonic raw data and the non-ultrasonic raw data include multiple data points. Normalizing the ultrasonic raw data and the non-ultrasonic raw data respectively means normalizing the values of the data points of the ultrasonic raw data and the non-ultrasonic raw data to between -1 and 1. The segmentation process means dividing the data according to time periods. For example, the normalized ultrasonic raw data or the normalized non-ultrasonic raw data can be segmented into multiple segments of data according to 500 ms. Of course, in other embodiments of the present application, the lengths of the time periods corresponding to each segment of ultrasonic data can be different, and the lengths of the time periods corresponding to each segment of non-ultrasonic data can also be different.

[0066] The ultrasonic data and the non-ultrasonic data included in each data pair can be data collected within the same time period.

[0067] If the sampling resolution is 16 bits, then the range of the energy values of the sampled data points is between -32767 and 32767. By dividing the energy value of the sampled data point by 32767, normalization can be achieved.

[0068] The ultrasonic sensor and the audible sound sensor can be controlled in a dual-channel manner to sample at a certain sampling frequency to obtain data points.

[0069] Step 220: Extract features from the ultrasonic data and the non-ultrasonic data respectively to extract at least one ultrasonic feature from the ultrasonic data and at least one non-ultrasonic feature from the non-ultrasonic data.

[0070] One or more ultrasonic features can be extracted from the ultrasonic data, and one or more non-ultrasonic features can also be extracted from the non-ultrasonic data.

[0071] In an embodiment of the present application, the at least one ultrasonic feature includes a power frequency feature. The ultrasonic data includes multiple data points. The specific steps for extracting the power frequency feature from the ultrasonic data can be as Figure 3 shown.

[0072] Figure 3 is a flowchart showing the extraction of at least one ultrasonic feature from ultrasonic data according to an exemplary embodiment. Please refer to Figure 3 shown. Specifically, the extraction of at least one ultrasonic feature from the ultrasonic data can include the following steps:

[0073] Step 310: Frame the ultrasonic data according to a predetermined duration, and take each frame as an energy point.

[0074] The ultrasonic data can be the data points within a time period of 500 ms. The predetermined duration can be 1 ms, 2 ms, etc. Usually, multiple data points are included within the predetermined duration, and each data point has a corresponding energy value.

[0075] Taking the predetermined duration of 1 ms as an example, 500 ms can be divided into 500 frames, and each frame with a length of 1 ms is used as an energy point.

[0076] Step 320: Determine the amplitude area of each frame and the average value of the amplitude areas of all frames, and determine the strong energy points among all energy points according to the amplitude area of each frame and the average value of the amplitude areas of all frames.

[0077] An energy point is a frame. Therefore, the strong energy points determined among all energy points are the frames with relatively strong energy selected from all frames.

[0078] In an embodiment of the present application, the determining the strong energy points among all energy points according to the amplitude area of each frame and the average value of the amplitude areas of all frames includes: determining the ratio of the amplitude area of each frame to the average value of the amplitude areas of all frames; for each frame, if the ratio corresponding to this frame is greater than a predetermined ratio threshold, then use this frame as a strong energy point.

[0079] Specifically, first determine the amplitude area S of the ultrasonic data, that is, the area under the curve drawn by all data points in the ultrasonic data; then, divide the amplitude area S by the total number of frames into which the ultrasonic data is divided to obtain the average amplitude area Sa; then, calculate the ratio of the amplitude area s of each frame to the average amplitude area Sa. If this ratio is greater than 1.5, then use the corresponding frame as a strong energy point and represent it with the value 1. Otherwise, do not use the corresponding frame as a strong energy point and represent this frame with the value 0.

[0080] Step 330: Determine the coordinates of each strong energy point according to the predetermined duration and the relationship between the power frequency period and the phase, and distribute each strong energy point on the unit circle according to the corresponding coordinates.

[0081] The power frequency unit circle is a circle with a radius of unit length (such as 1).

[0082] Power frequency refers to the rated frequency adopted by power generation, transmission, transformation, and distribution equipment in the power system, as well as industrial and civil electrical equipment, with the unit of Hz (Hertz). The power frequency adopted in China is 50 Hz. Correspondingly, the power frequency period is 1000 / 50 = 20 ms. The relationship between the power frequency period and the phase can be that 20 ms corresponds to 360 degrees. Each time, the energy points corresponding to 20 ms are taken; when the predetermined duration is 1 ms, 20 energy points are taken each time and numbered. According to the numbers, the phase θ is generated. For example, the phase of the first energy point is 0°, and the phase of the second energy point is 18°. Then, the coordinates of each energy strong point are calculated according to (cosθ, sinθ), and the energy strong points are marked on the circumference of the unit circle. By analogy, with 20 energy points as a cycle, the marking of energy strong points is carried out until all the energy strong points corresponding to the ultrasonic data are marked on the circumference of the unit circle. Abnormalities such as partial discharge usually occur near the peaks and troughs of the waveform of alternating current. Then, if partial discharge occurs, the phase corresponding to the moment of partial discharge is around 90 degrees and 270 degrees, so that partial discharge can be better identified.

[0083] Step 340: Cluster the energy strong points on the unit circle according to the density to obtain a clustering result, and use the clustering result as the power frequency feature.

[0084] Algorithms that can achieve clustering, such as shallow neural networks and k-means, can be used to cluster the energy strong points on the unit circle. The clustering result is one or more clusters formed by dividing the energy strong points on the unit circle, and each cluster includes one or more energy strong points.

[0085] Although factors such as electromagnetic interference and natural wind will generate ultrasonic waves, these ultrasonic waves basically do not conform to the power frequency characteristics. In the embodiments of the present application, by extracting the power frequency characteristics from the ultrasonic data and performing anomaly detection based on the power frequency characteristics, the interference caused by other factors such as electromagnetic interference and natural wind is effectively avoided, and the accuracy of anomaly detection is improved.

[0086] In an embodiment of the present application, the at least one non-ultrasonic feature includes a frequency component feature. The specific steps for extracting the frequency component feature from the non-ultrasonic data can be as Figure 4 shown.

[0087] Figure 4 is a flowchart of extracting at least one non-ultrasonic feature from non-ultrasonic data shown according to an exemplary embodiment. As Figure 4 shown, extracting at least one non-ultrasonic feature from non-ultrasonic data may specifically include the following steps:

[0088] Step 410: Convert the non-ultrasonic data from time-domain data to frequency-domain data.

[0089] Specifically, the non-ultrasonic data exists in the form of time-domain data. By performing a Fourier transform on the time-domain data, frequency-domain data can be obtained.

[0090] The frequency-domain data includes spectra corresponding to multiple frequencies.

[0091] Step 420: Determine the total spectral area at all frequencies in the frequency-domain data and the spectral area of the target frequency band in the frequency-domain data.

[0092] In an embodiment of the present application, the target frequency band belongs to the audible frequency band and there is a frequency band lower than the target frequency band in the audible frequency band.

[0093] In an embodiment of the present application, the target frequency band is 2KHz - 10KHz.

[0094] Therefore, the target frequency band is a frequency band with a relatively high frequency in the audible frequency band.

[0095] In an embodiment of the present application, the areas of the spectra corresponding to all frequencies are summed to obtain the total spectral area F at all frequencies; the areas of the spectra corresponding to the frequencies belonging to the target frequency band of 2KHz - 10KHz are summed to obtain the spectral area f of the target frequency band.

[0096] Step 430: Determine the frequency component feature according to the spectral area and the total spectral area.

[0097] In an embodiment of the present application, determining the frequency component feature according to the spectral area and the total spectral area includes:

[0098] Taking the ratio of the spectral area to the total spectral area as the frequency component feature.

[0099] The frequency component feature can be calculated by the following formula:

[0100]

[0101] where f is the spectral area of the target frequency band, F is the total spectral area, and f k is the frequency component feature.

[0102] The sound caused by mechanical vibration due to loose components in electrical equipment usually has only low-frequency audible sound, while partial discharge generates high-frequency audible sound. In an embodiment of the present application, by taking the ratio of the spectral area of the target frequency band with a relatively high frequency in the audible frequency band to the total spectral area as the frequency component feature, the mechanical vibration and partial discharge can be better distinguished using this frequency component feature, thereby improving the accuracy of anomaly detection.

[0103] In an embodiment of the present application, the at least one non-ultrasonic feature further includes a steepness feature. The non-ultrasonic data includes multiple data points. Extracting at least one non-ultrasonic feature from the non-ultrasonic data further includes:

[0104] Taking the absolute value of the energy difference between two adjacent data points in the non-ultrasonic data as the steepness, and determining the average value of all the steepnesses corresponding to the non-ultrasonic data as the steepness feature.

[0105] Specifically, if the non-ultrasonic data includes n data points and each data point has an energy value, calculating the absolute value of the difference between the energy values of every two adjacent data points as the steepness, then n - 1 steepnesses can be calculated, and the average value of all the steepnesses corresponding to the non-ultrasonic data = the sum of the n - 1 steepnesses / (n - 1).

[0106] The steepness of the audible sound data generated by mechanical vibration is usually not high, while partial discharge occurs suddenly, so the steepness of the audible sound data generated by partial discharge is usually very high. In the embodiment of the present application, by determining the steepness feature, mechanical vibration and partial discharge can be better distinguished, thereby improving the accuracy of anomaly detection.

[0107] In an embodiment of the present application, before respectively extracting features from the ultrasonic data and the non-ultrasonic data, the method further includes: determining whether the signal amplitude of the non-ultrasonic data exceeds a predetermined signal amplitude threshold, where extracting features from the ultrasonic data and the non-ultrasonic data respectively is performed when the signal amplitude of the non-ultrasonic data exceeds the predetermined signal amplitude threshold.

[0108] When the signal amplitude of the non-ultrasonic data (i.e., audible sound data) is small enough, it is basically impossible for an anomaly to occur. In the embodiment of the present application, by only extracting features from the ultrasonic data and the non-ultrasonic data when the signal amplitude of the non-ultrasonic data exceeds the predetermined signal amplitude threshold, and not performing subsequent steps when the signal amplitude of the non-ultrasonic data is small enough and not extracting features from the ultrasonic data and the non-ultrasonic data, the consumption of resources can be effectively saved.

[0109] Please continue to refer to Figure 2 , step 230, mapping each ultrasonic feature to an ultrasonic feature score and mapping each non-ultrasonic feature to a non-ultrasonic feature score.

[0110] Specifically, a machine learning model such as a neural network model can be used to map the ultrasonic feature to an ultrasonic feature score and map the non-ultrasonic feature to a non-ultrasonic feature score. Of course, other methods can also be used to map the feature to a feature score.

[0111] In one embodiment of the present application, each ultrasonic feature is mapped to an ultrasonic feature score, and each non-ultrasonic feature is mapped to a non-ultrasonic feature score, including:

[0112] According to the eigenvalue interval to which each ultrasonic feature belongs, each ultrasonic feature is mapped to the ultrasonic feature score corresponding to the eigenvalue interval;

[0113] According to the eigenvalue interval to which each non-ultrasonic feature belongs, each non-ultrasonic feature is mapped to the non-ultrasonic feature score corresponding to the eigenvalue interval.

[0114] Specifically, for the power frequency feature, if the clustering result is 1-2 clusters, then the feature score b1 mapped from the power frequency feature can be 1 point; if the clustering result is more than 2 clusters, then the feature score b1 mapped from the power frequency feature can be 0 point.

[0115] For the frequency component feature, if f k is greater than 0.5, it can be considered that the high-frequency component is high, then the feature score b2 mapped from the frequency component feature can be 1 point; if f k does not exceed 0.5, then the feature score b2 mapped from the frequency component feature can be 0 point.

[0116] For the steepness feature, if the steepness feature is greater than 0.7, it is considered that the average steepness is high, then the feature score b3 mapped from the steepness feature can be 2 points; if the steepness feature belongs to [0.3, 0.7], then the feature score b3 mapped from the steepness feature can be 1 point; if the steepness feature is less than 0.3, then the feature score b3 mapped from the steepness feature can be 0 point.

[0117] Step 240, obtain the total feature score according to each ultrasonic feature score and each non-ultrasonic feature score.

[0118] Specifically, the total feature score can be obtained by summing each ultrasonic feature score and each non-ultrasonic feature score, that is, the total feature score = b1 + b2 + b3.

[0119] Of course, in other embodiments of the present application, the total feature score can also be obtained by other means. For example, the total feature score can be obtained by inputting each ultrasonic feature score and each non-ultrasonic feature score into a machine learning model such as a logistic regression model, or the total feature score can be obtained using the following formula: total feature score = A * b1 + B * b2 + C * b3, where A, B, and C are the weight coefficients of each feature score.

[0120] Step 250, obtain the abnormal detection result of the power equipment according to the total feature score.

[0121] In an embodiment of the present application, the abnormal detection result is any one of the following: possible partial discharge, partial discharge, mechanical vibration.

[0122] When the total feature score is determined by the formula total feature score = b1 + b2 + b3, the abnormal detection result of the power equipment can be obtained by using the following rules: if the total feature score is 4 points, the abnormal detection result is partial discharge; if the total feature score is 1 point, the abnormal detection result is mechanical vibration, that is, there is no partial discharge; if the total feature score is 2 - 3 points, the abnormal detection result is possible partial discharge, that is, partial discharge may have occurred during the acquisition period of ultrasonic data and non - ultrasonic data.

[0123] In an embodiment of the present application, after obtaining the abnormal detection result of the power equipment according to the total feature score, the method further includes: sending an alarm signal according to the alarm method corresponding to the abnormal detection result.

[0124] Specifically, a control signal is sent to the alarm lamp to instruct the alarm lamp to switch to the corresponding color.

[0125] In an embodiment of the present application, after obtaining the abnormal detection result of the power equipment according to the total feature score, the method further includes: sending an alarm message according to the type of the abnormal detection result.

[0126] Specifically, an alarm message can be generated according to the type of the abnormal detection result and sent to a communication device such as a smart phone of the relevant staff to prompt the relevant personnel to repair the power equipment in time.

[0127] In an embodiment of the present application, after obtaining the abnormal detection result of the power equipment according to the total feature score, the method further includes: obtaining the final abnormal detection result of the power equipment according to the abnormal detection results corresponding to each data pair.

[0128] Specifically, a corresponding abnormal detection result can be determined for each data pair; when all abnormal detection results are partial discharge or the number of abnormal detection results that are partial discharge accounts for more than half of all abnormal detection results, the final abnormal detection result can be determined as partial discharge.

[0129] Therefore, a corresponding alarm message can be sent after obtaining the final abnormal detection result.

[0130] In an embodiment of the present application, by generating corresponding abnormal detection results for each group of data pairs and comprehensively analyzing all abnormal detection results to obtain the final abnormal detection result, the accuracy of abnormal detection is improved.

[0131] Figure 5It is a schematic overall process diagram shown according to an exemplary embodiment. As Figure 5 shown, it includes the following steps:

[0132] Step 510, obtaining audible sound data and ultrasonic data.

[0133] Step 520, performing normalization processing and segmentation processing.

[0134] Step 530, extracting power frequency characteristics of the ultrasonic time-domain data and scoring.

[0135] Step 540, extracting frequency-domain component characteristics of the audible sound data and scoring.

[0136] Step 550, extracting steepness characteristics of the audible sound time-domain data and scoring.

[0137] Step 560, calculating the total score and discriminating the abnormal sound type of this segment.

[0138] Step 570, determining whether the data processing is completed.

[0139] If so, execute step 580; otherwise, re-execute step 520 and the subsequent steps.

[0140] Step 580, end.

[0141] In summary, according to the abnormal detection method of the power equipment provided by the embodiments of the present application, by adopting the combination of audible sound and ultrasonic waves to discriminate partial discharge, compared with other detection methods, the cost is lower, more economical, and easier to promote; moreover, this method is simple and easy to understand and conforms to the occurrence mechanism of abnormalities such as partial discharge, basically eliminating the interference influence of high-frequency signals, and significantly improving the accuracy and reliability of abnormal discrimination.

[0142] The present application also provides an abnormal detection device for power equipment. The following is the device embodiment of the present application.

[0143] Figure 6 It is a block diagram of an abnormal detection device for power equipment shown according to an exemplary embodiment.

[0144] As Figure 6 shown, the device 600 includes:

[0145] An acquisition module 610, configured to acquire ultrasonic data and non-ultrasonic data collected at the power equipment site, and the non-ultrasonic data is audio data with a sound frequency band different from that of the ultrasonic data;

[0146] A feature extraction module 620, configured to perform feature extraction on the ultrasonic data and the non-ultrasonic data respectively, so as to extract at least one ultrasonic feature from the ultrasonic data and at least one non-ultrasonic feature from the non-ultrasonic data;

[0147] A mapping module 630, configured to map each ultrasonic feature to an ultrasonic feature score and map each non-ultrasonic feature to a non-ultrasonic feature score;

[0148] A calculation module 640, configured to obtain a total feature score according to each ultrasonic feature score and each non-ultrasonic feature score;

[0149] A judgment module 650, configured to obtain an abnormal detection result of the power equipment according to the total feature score.

[0150] According to another aspect of the present application, there is provided an abnormal detection system for a power equipment, including:

[0151] An acquisition module, configured to acquire ultrasonic data and non-ultrasonic data at the site of the power equipment, where the non-ultrasonic data is audio data with a sound frequency band different from that of the ultrasonic data;

[0152] A data processing module, configured to process the ultrasonic data and the non-ultrasonic data to obtain at least one ultrasonic feature and at least one non-ultrasonic feature;

[0153] A discrimination and warning module, configured to analyze the at least one ultrasonic feature and the at least one non-ultrasonic feature to obtain an abnormal detection result, and issue an alarm signal corresponding to the abnormal detection result.

[0154] Specifically, the acquisition module may include an audible sound sensor, an ultrasonic sensor, an amplifier, and an analog-to-digital converter. The ultrasonic signal and the audible sound signal collected by the audible sound sensor and the ultrasonic sensor are amplified by the amplifier and then input into the analog-to-digital converter. The analog quantity is converted into a digital quantity by the analog-to-digital converter to obtain digital ultrasonic data and audible sound data.

[0155] The discrimination and warning module may include a remote reminder terminal and an alarm light. When the abnormal detection result is partial discharge, the discrimination and warning module may control the alarm light to emit red light. When the abnormal detection result is possible partial discharge, the discrimination and warning module may control the alarm light to emit yellow light. The discrimination and warning module may also remind on-site staff to perform maintenance through the remote reminder terminal.

[0156] Figure 7 The structure diagram of a computer system of an electronic device suitable for implementing the embodiments of the present application is shown.

[0157] It should be noted thatFigure 7 The computer system 700 of the illustrated electronic device is merely an example and should not impose any limitation on the functions and scope of use of the embodiments of the present application.

[0158] As Figure 7 shown, the computer system 700 includes a central processing unit (CPU) 701, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 702 or the program loaded from the storage section 708 into the random access memory (RAM) 703, such as executing the method described in the above embodiments. In the RAM 703, various programs and data required for system operation are also stored. The CPU 701, ROM 702, and RAM 703 are connected to each other via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0159] The following components are connected to the I / O interface 705: an input section 706 including a keyboard, a mouse, etc.; an output section 707 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc. and a speaker, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the I / O interface 705 as required. A removable medium 711, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 710 as required so that a computer program read from it can be installed into the storage section 708 as required.

[0160] Specifically, according to the embodiments of the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments of the present application include a computer program product that includes a computer program carried on a computer-readable medium, and the computer program includes program codes for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication section 709, and / or installed from the removable medium 711. When the computer program is executed by the central processing unit (CPU) 701, various functions defined in the system of the present application are executed.

[0161] It should be noted that the computer-readable medium shown in the embodiments of the present application can be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. The 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 of the above. More specific examples of the computer-readable storage medium can include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, the computer-readable storage medium can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, apparatus, or device. In the present application, the computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable storage medium, and this computer-readable medium can send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any appropriate medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0162] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present application. Among them, each block in the flowchart or block diagram can represent a module, a program segment, or a part of the code, and the above module, program segment, or part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order from that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram or flowchart, as well as the combination of blocks in the block diagram or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0163] The units involved in the embodiments described in this application can be implemented in software or in hardware, and the described units can also be provided in a processor. Among them, the names of these units do not, in some cases, constitute a limitation on the unit itself.

[0164] As one aspect, this application also provides a computer-readable medium, which may be included in the electronic device described in the above embodiments; or it may exist alone without being assembled into the electronic device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by a processor of the electronic device, the electronic device implements the methods described in the above embodiments.

[0165] It should be noted that although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of this application, the features and functions of the two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0166] From the description of the above embodiments, those skilled in the art can easily understand that the example embodiments described herein can be implemented by software or by a combination of software and necessary hardware. Therefore, the technical solutions according to the embodiments of this application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (such as a personal computer, a server, a touch terminal, or a network device, etc.) to execute the methods according to the embodiments of this application.

[0167] After considering the specification and practicing the embodiments disclosed herein, those skilled in the art will readily conceive of other embodiments of this application. This application is intended to cover any variations, uses, or adaptations of this application, which follow the general principles of this application and include known common knowledge or conventional technical means in the technical field not disclosed in this application.

[0168] It should be understood that this application is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is only limited by the appended claims.

Claims

1. An abnormal detection method for a power device, characterized in that, The method includes: Obtaining ultrasonic data and non-ultrasonic data collected at the site of the power equipment, where the non-ultrasonic data is audio data with a different sound frequency band from the ultrasonic data; Framing the ultrasonic data according to a predetermined duration, and taking each divided frame as an energy point; Determining the amplitude area of each frame and the average value of the amplitude areas of all frames; wherein, the amplitude area of each frame is the area under the curve drawn by all data points in each frame of the ultrasonic data; the average value of the amplitude areas of all frames is the value obtained by dividing the sum of the amplitude areas of all frames by the total number of frames into which the ultrasonic data is divided; Calculating the ratio of the amplitude area of each frame to the average value of the amplitude areas of all frames, and if the ratio is greater than a predetermined ratio threshold, taking the corresponding frame as an energy strong point; Determining the coordinates of each energy strong point according to the predetermined duration and the relationship between the power frequency period and the phase, and distributing each energy strong point on the unit circle according to the corresponding coordinates; Clustering the energy strong points on the unit circle according to the density to obtain a clustering result, and taking the clustering result as the power frequency feature; Performing feature extraction on the non-ultrasonic data to extract at least one non-ultrasonic feature from the non-ultrasonic data; Mapping each power frequency feature to an ultrasonic feature score, and mapping each non-ultrasonic feature to a non-ultrasonic feature score; Obtaining a total feature score according to each ultrasonic feature score and each non-ultrasonic feature score; Obtaining an abnormal detection result for the power equipment according to the total feature score.

2. The method according to claim 1, wherein The at least one non-ultrasonic feature includes a frequency component feature, and extracting at least one non-ultrasonic feature from the non-ultrasonic data includes: Converting the non-ultrasonic data from time-domain data to frequency-domain data; Determining the total spectral area at all frequencies in the frequency-domain data and the spectral area of the target frequency band in the frequency-domain data; Determining the frequency component feature according to the spectral area and the total spectral area.

3. The method according to claim 2, wherein The at least one non-ultrasonic feature further includes a steepness feature, the non-ultrasonic data includes a plurality of data points, and extracting at least one non-ultrasonic feature from the non-ultrasonic data further includes: Taking the absolute value of the energy difference between two adjacent data points in the non-ultrasonic data as the steepness, and determining the average value of all the steepnesses corresponding to the non-ultrasonic data as the steepness feature.

4. The method according to any one of claims 1 to 3, characterized in that The obtaining the ultrasonic data and the non-ultrasonic data collected at the site of the power equipment includes: Obtaining the ultrasonic raw data and the non-ultrasonic raw data collected at the site of the power equipment; Respectively performing normalization processing on the ultrasonic raw data and the non-ultrasonic raw data to obtain the normalized ultrasonic raw data and the normalized non-ultrasonic raw data; Respectively performing segmentation processing on the normalized ultrasonic raw data and the normalized non-ultrasonic raw data to obtain multiple segments of ultrasonic data and multiple segments of non-ultrasonic data; Obtaining data pairs from the multiple segments of ultrasonic data and the multiple segments of non-ultrasonic data, where the data pair includes a segment of ultrasonic data obtained from the multiple segments of ultrasonic data and a segment of non-ultrasonic data obtained from the multiple segments of non-ultrasonic data; After obtaining the abnormal detection result of the power equipment according to the total feature score, the method further includes: Obtaining the final abnormal detection result of the power equipment according to the abnormal detection results corresponding to the respective data pairs.

5. The method according to any one of claims 1-3, characterized in that, Before respectively performing feature extraction on the ultrasonic data and the non-ultrasonic data, the method further includes: Judging whether the signal amplitude of the non-ultrasonic data exceeds a predetermined signal amplitude threshold, wherein the feature extraction of the ultrasonic data and the non-ultrasonic data is performed when the signal amplitude of the non-ultrasonic data exceeds the predetermined signal amplitude threshold.

6. The method according to any one of claims 1-3, characterized in that, After obtaining the abnormal detection result of the power equipment according to the total feature score, the method further includes: Sending an alarm signal according to the alarm mode corresponding to the abnormal detection result.

7. An abnormal detection device for a power device, characterized in that, The device includes: An acquisition module configured to acquire ultrasonic data and non-ultrasonic data collected at the power equipment site, where the non-ultrasonic data is audio data with a sound frequency band different from that of the ultrasonic data; A feature extraction module configured to frame the ultrasonic data according to a predetermined duration, and take each divided frame as an energy point; determine the amplitude area of each frame and the average value of the amplitude areas of each frame; wherein, the amplitude area of each frame is the area under the curve drawn by all data points in each frame of the ultrasonic data; the average value of the amplitude areas of each frame is the value obtained by dividing the sum of the amplitude areas of each frame by the total number of frames into which the ultrasonic data is divided; calculate the ratio of the amplitude area of each frame to the average value of the amplitude areas of each frame, and if the ratio is greater than a predetermined ratio threshold, take the corresponding frame as an energy strong point; determine the coordinates of each energy strong point according to the predetermined duration and the relationship between the power frequency period and the phase, and distribute each energy strong point on the unit circle according to the corresponding coordinates; cluster the energy strong points of the unit circle according to the density to obtain a clustering result, and use the clustering result as the power frequency feature; perform feature extraction on the non-ultrasonic data to extract at least one non-ultrasonic feature from the non-ultrasonic data; A mapping module configured to map each power frequency feature to an ultrasonic feature score and map each non-ultrasonic feature to a non-ultrasonic feature score; A calculation module configured to obtain a total feature score according to each ultrasonic feature score and each non-ultrasonic feature score; A judgment module configured to obtain the abnormal detection result of the power equipment according to the total feature score.

8. A computer-readable medium having a computer program stored thereon, characterized in that, The computer program, when executed by a processor, implements the method according to any one of claims 1 to 6.

9. An abnormal detection system for a power device, characterized in that, Including: An acquisition module for acquiring ultrasonic data and non-ultrasonic data at the power equipment site, where the non-ultrasonic data is audio data with a sound frequency band different from that of the ultrasonic data; A data processing module for framing the ultrasonic data according to a predetermined duration and taking each divided frame as an energy point; Determine the amplitude area of each frame and the average value of the amplitude areas of each frame; wherein, the amplitude area of each frame is the area under the curve drawn by all data points in the ultrasonic data of each frame; the average value of the amplitude areas of each frame is the value obtained by dividing the sum of the amplitude areas of each frame by the total number of frames into which the ultrasonic data is divided; calculate the ratio of the amplitude area of each frame to the average value of the amplitude areas of each frame, and if this ratio is greater than a predetermined ratio threshold, then take the corresponding frame as an energy strong point; according to the predetermined duration and the relationship between the power frequency period and the phase, determine the coordinates of each energy strong point, and distribute each energy strong point on the unit circle according to the corresponding coordinates; cluster the energy strong points on the unit circle according to the degree of density to obtain a clustering result, and take the clustering result as the power frequency feature; process the non-ultrasonic data to obtain at least one non-ultrasonic feature; A discrimination and early warning module, which is used to analyze the power frequency feature and the at least one non-ultrasonic feature to obtain an abnormal detection result, and issue an alarm signal corresponding to the abnormal detection result.

Images