Arc detection method and system based on ultraviolet photoelectricity

Abstract

The invention discloses an arc detection method and system based on ultraviolet photoelectricity, and the method comprises the steps: obtaining ultraviolet photoelectricity data, and carrying out the data preprocessing; performing spectrum feature extraction processing on the preprocessed data to obtain spectrum feature data; performing interception processing on the spectrum characteristic data according to a preset frequency interception rule to obtain spectrum characteristic data in a preset frequency band range; calculating an energy value of each frame in the spectrum feature data to obtain energy feature data of each frame; inputting the energy characteristic data of each frame into a pre-trained arc detection model to obtain an initial detection result of each frame; and performing optimization processing on the initial detection result of each frame according to an arc optimization rule to obtain an optimized arc detection result. According to the invention, arc detection processing can be carried out through the collected ultraviolet photoelectric data, and the arc detection accuracy and detection efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of arc detection, and in particular, to an arc detection method and system based on ultraviolet optoelectronics. Background Art

[0002] An arc refers to a spark formed by the free discharge of current in the air, which is characterized by strong light, high temperature and high energy. Arc faults are usually caused by insulation faults in electrical equipment, such as line breakdowns, equipment short circuits, etc. Once an arc fault occurs, it will generate very high voltages and currents, causing damage to the equipment. Therefore, corresponding measures need to be taken to detect and protect against abnormal arcs.

[0003] At present, most of the abnormal arc detections in the power system are based on electrical parameters, that is, current and voltage for identification; usually, special monitoring devices need to be added to the power system to collect voltage and current signals, and perform corresponding signal processing and analysis to judge the occurrence of arc faults. However, this single modality of electrical parameters is easily affected by external interference, resulting in misjudgment, and it is necessary to study arc detection methods based on other modalities to improve the accuracy of abnormal arc detection. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide an arc detection method and system based on ultraviolet optoelectronics, which can perform arc detection processing through the collected ultraviolet optoelectronic data, and improve the accuracy and efficiency of arc detection.

[0005] To solve the above technical problem, the present invention provides an arc detection method based on ultraviolet optoelectronics, including: acquiring ultraviolet optoelectronic data and performing data preprocessing; performing spectral feature extraction processing on the preprocessed data to obtain spectral feature data; performing interception processing on the spectral feature data according to a preset frequency interception rule to obtain spectral feature data within a preset frequency band range; calculating the energy value of each frame in the spectral feature data to obtain energy feature data for each frame; inputting the energy feature data of each frame into a pre-trained arc detection model to obtain an initial detection result for each frame; and performing optimization processing on the initial detection result of each frame according to an arc optimization rule to obtain an optimized arc detection result.

[0006] As an improvement of the above solution, the steps of the data preprocessing include: performing data caching processing on the collected ultraviolet optoelectronic data to obtain MXN cache data, where the MXN cache data is such that each of the M frames of data includes N data points, and the parameters M and N are both positive integers; and performing data conversion processing on the MXN stored data to obtain floating-point data.

[0007] As an improvement to the above solution, the step of extracting spectral features from the preprocessed data to obtain spectral feature data includes: performing FFT transformation on the preprocessed data to obtain spectral feature data.

[0008] As an improvement to the above solution, the method for calculating the energy value of each frame is: calculating the energy value of each frame through an energy calculation formula, and the formula is as follows:

[0009]

[0010] where freq_feature is the sum of the energy values of all frequency points in each frame and serves as the energy feature data, n is the total number of frequency points in the current frame data, and D cut(i) 2 is the energy value of the i-th frequency point in the current frame data, and both n and i are positive integers.

[0011] As an improvement to the above solution, the arc optimization rule includes: within a preset time, counting the number of detection frames with an arc in the arc detection result and the total number of frames of the arc detection result; determining whether the ratio of the number of detection frames to the total number of frames is greater than or equal to a preset optimization threshold. If the determination is yes, it means that the arc detection result within the preset time is that there is an arc. If the determination is no, it means that the arc detection result within the preset time is that there is no arc.

[0012] Correspondingly, the present invention also provides an arc detection system based on ultraviolet optoelectronics, including: an acquisition module for acquiring ultraviolet optoelectronic data and performing data preprocessing; a feature extraction module for extracting spectral features from the preprocessed data to obtain spectral feature data; a frequency truncation module for truncating the spectral feature data according to a preset frequency truncation rule to obtain spectral feature data within a preset frequency band range; an energy calculation module for calculating the energy value of each frame in the spectral feature data to obtain the energy feature data of each frame; a detection processing module for inputting the energy feature data of each frame into a pre-trained arc detection model to obtain an initial detection result for each frame; and an arc optimization module for optimizing the initial detection result of each frame according to the arc optimization rule to obtain an optimized arc detection result.

[0013] As an improvement to the above solution, the acquisition module includes: a data caching unit for caching the acquired ultraviolet optoelectronic data to obtain MXN cached data, where the MXN cached data is such that each of the M frames includes N data points, and the parameters M and N are both positive integers; and a data preprocessing unit for performing data conversion on the MXN stored data to obtain floating-point data.

[0014] As an improvement to the above solution, the feature extraction module includes: a feature extraction unit, configured to perform FFT transformation on the preprocessed data to obtain spectral feature data.

[0015] As an improvement to the above solution, the energy calculation module includes: an energy calculation unit, configured to calculate the energy value of each frame through an energy calculation formula, and the formula is as follows:

[0016]

[0017] where freq_feature is the sum of the energy values of all frequency points in each frame and serves as the energy feature data, n is the total number of frequency points in the current frame data, and D cut(i) 2 is the energy value of the i-th frequency point in the current frame data, and both n and i are positive integers.

[0018] As an improvement to the above solution, the arc optimization module includes: a statistics unit, configured to count the number of detected frames with an arc in the arc detection result and the total number of frames of the arc detection result within a preset time; a result optimization unit, configured to determine whether the ratio of the number of detected frames to the total number of frames is greater than or equal to a preset optimization threshold. If the determination is yes, it indicates that there is an arc in the arc detection result within the preset time. If the determination is no, it indicates that there is no arc in the arc detection result within the preset time.

[0019] Implementing the present invention has the following beneficial effects:

[0020] The present invention provides an arc detection method and system based on ultraviolet photoelectricity, which can perform arc detection processing through the collected ultraviolet photoelectric data, improving the accuracy and efficiency of arc detection.

[0021] Among them, the present invention can perform spectral feature extraction processing on the obtained ultraviolet photoelectric data to obtain spectral feature data; through a preset frequency truncation rule, the spectral feature data can be truncated to obtain useful spectral feature data, reducing the data processing volume and improving the data processing accuracy; by calculating the energy value of each frame in the spectral feature data, the energy feature data of each frame can be obtained; inputting the energy feature data of each frame into a pre-trained arc detection model to obtain the arc detection result of each frame, thereby determining whether there is an arc.

[0022] In addition, the present invention can also optimize the arc detection result according to the arc optimization rule to reduce the interference caused by external factors and further improve the accuracy of the detection result. Description of the Drawings

[0023] Figure 1It is the flow chart of the arc detection method based on ultraviolet optoelectronics of the present invention;

[0024] Figure 2 It is the schematic structural diagram of the training mechanism of the present invention;

[0025] Figure 3 It is the schematic structural diagram of the detection mechanism of the present invention;

[0026] Figure 4 It is the schematic structural diagram of the arc detection system based on ultraviolet optoelectronics of the present invention;

[0027] Figure 5 It is the schematic structural diagram of the acquisition module of the present invention;

[0028] Figure 6 It is the schematic structural diagram of the feature extraction module of the present invention;

[0029] Figure 7 It is the schematic structural diagram of the arc optimization module of the present invention. Detailed implementation manners

[0030] To make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings.

[0031] It should be noted that, as Figures 2 to 3 shown, the present invention has pre-built a training mechanism and a detection mechanism.

[0032] In the training stage, as Figure 2 shown, the training mechanism includes a host computer, an ultraviolet band sensor, a spectrometer and an arc generator. The arc generator is arranged on the power supply line of the electrical appliance, and the ultraviolet band sensor is close to the arc generator. The ultraviolet band sensor is used to collect the ultraviolet optoelectronic training data generated by the arc generator in real time. Since a large amount of light radiation, including a large amount of ultraviolet light, is generated during arc discharge, and almost no ultraviolet light radiation is generated in the circuit when there is no arc discharge, the ultraviolet band sensor can collect ultraviolet optoelectronic signals in real time, which are used as the basis for arc occurrence judgment.

[0033] The spectrometer can receive the ultraviolet optoelectronic training data input by the ultraviolet band sensor and perform spectrum processing and frequency band truncation processing to obtain spectrum data in the required frequency band and send it to the host computer; the host computer is used to calculate the energy value of each frame in the spectrum data and import the energy value of each frame into the training detection model, and batch train with the training data to obtain a trained arc detection model.

[0034] Among them, the set arc generator can automatically adjust the generation of the arc according to requirements, so as to facilitate the collection of useful ultraviolet photoelectric training data, and the ultraviolet photoelectric training data includes abnormal data with an arc or normal data without an arc.

[0035] In the detection stage, as Figure 3 shown, the detection mechanism includes a detection device, and the detection device includes a detection processor and an ultraviolet light band sensor. The ultraviolet light band sensor is arranged on one side of the power supply line of the electrical appliance. The set ultraviolet light band sensor can collect ultraviolet photoelectric signals in real time and send them to the detection processor; a method and system for arc detection based on ultraviolet photoelectricity are provided in the detection processor to quickly and accurately detect the arc. Among them, the detection processor is preferably an ARM processor, but not limited thereto.

[0036] Specifically, as Figure 1 shown, the present invention provides a method for detecting an arc based on ultraviolet photoelectricity, which includes:

[0037] S101. Obtain ultraviolet photoelectric data and perform data preprocessing;

[0038] Specifically, the steps of the data preprocessing include:

[0039] Step 1. Perform data caching processing on the collected ultraviolet photoelectric data to obtain MXN cache data, where the MXN cache data is M frames of data, and each frame includes N data points, and the parameters M and N are both positive integers;

[0040] It should be noted that the collected ultraviolet photoelectric data is cached to facilitate subsequent rapid processing of the MXN cache data and improve data processing efficiency.

[0041] Step 2. Perform data conversion processing on the MXN stored data to obtain floating-point data.

[0042] It should be noted that the cached ultraviolet photoelectric data is subjected to data conversion calibration to obtain high-precision floating-point data, reduce data storage space and calculation amount, and thus improve subsequent data processing efficiency and accuracy.

[0043] S102. Perform spectrum feature extraction processing on the preprocessed data to obtain spectrum feature data;

[0044] Specifically, the step of performing spectrum feature extraction processing on the preprocessed data to obtain spectrum feature data includes: performing FFT transform (fast Fourier transform) processing on the preprocessed data to obtain spectrum feature data.

[0045] It should be noted that by performing FFT transformation on the data, the time-domain data can be converted into frequency-domain data, which is convenient for analyzing the characteristics and frequency components of the signal, thereby obtaining the required spectral characteristic data, with high data processing efficiency and accuracy.

[0046] S103. Intercept the spectral characteristic data according to a preset frequency intercept rule to obtain spectral characteristic data within a preset frequency band range;

[0047] It should be noted that the preset frequency intercept rule is set with the required intercepted frequency band range. Through the preset frequency intercept rule, the useful frequency band range can be intercepted and retained in the spectral characteristic data, and the remaining useless frequency band ranges can be removed to reduce the data processing volume and improve the data processing efficiency and data processing accuracy.

[0048] Among them, the intercepted frequency band range is the frequency band range with a large distinction in the spectrum for comparing the presence or absence of arcs, and this frequency band range can be specifically set according to actual needs.

[0049] S104. Calculate the energy value of each frame in the spectral characteristic data to obtain the energy characteristic data of each frame;

[0050] It should be noted that the spectral characteristic data within the required frequency band range includes multiple frame data, and each frame data includes multiple frequency points. By calculating the sum of the energy values of all frequency points in each frame, the energy value of this frame can be obtained.

[0051] Specifically, the method for calculating the energy value of each frame is as follows:

[0052] Calculate the energy value of each frame through an energy calculation formula, and the formula is as follows:

[0053]

[0054] Among them, freq_feature is the sum of the energy values of all frequency points in each frame and serves as the energy characteristic data, n is the total number of frequency points in the frame data, D cut(i) 2 is the energy value of the i-th frequency point in the current frame data, and both n and i are positive integers.

[0055] It should be noted that through D cut(i) 2 the energy value of each frequency point after intercepting the data can be calculated. Through the energy calculation formula, the sum of the energy values of all frequency points in each frame data can be calculated to obtain the energy value of each frame data and serve as the energy characteristic data of each frame.

[0056] Secondly, the technical means of this step S104 is the same as the technical means of the host computer calculating the energy value of each frame during the training stage.

[0057] S105. Input the energy feature data of each frame into a pre-trained arc detection model to obtain an initial detection result for each frame.

[0058] It should be noted that there are differences or variations in the energy values of different frames, and correspondingly, there are feature differences or variations in the energy feature data of each frame. The arc detection model can identify and process the energy feature data of each frame to quickly obtain an initial detection result with high accuracy. Based on this initial detection result, it can be quickly determined whether there is an arc in that frame, thereby preliminarily determining whether an arc occurs in the circuit.

[0059] S106. Optimize the initial detection result of each frame according to the arc optimization rule to obtain an optimized arc detection result.

[0060] It should be noted that the arc detection result of a single frame is easily affected by external factors, resulting in a decrease in detection accuracy. Therefore, in this embodiment, the arc detection result is optimized according to the arc optimization rule to reduce the interference caused by external factors and further improve the accuracy of the final arc detection result.

[0061] Specifically, in order to further improve the accuracy of the detection result, the arc optimization rule includes:

[0062] Step 1. Within a preset time, count the number of detected frames with an arc in the arc detection result and the total number of frames of the arc detection result.

[0063] It should be noted that count the total number of all arc detection results within the preset time, and count the number of detected frames with an arc in the total number of frames.

[0064] Preferably, the preset time can be preferably 1S, but it is not limited thereto, and can be specifically adjusted according to actual needs.

[0065] Step 2. Determine whether the ratio of the number of detected frames to the total number of frames is greater than or equal to a preset optimization threshold. If the judgment is yes, it means that the arc detection result within the preset time is that there is an arc; if the judgment is no, it means that the arc detection result within the preset time is that there is no arc.

[0066] It should be noted that within the preset time, when the ratio of the number of detected frames with an arc to the total number of frames in the arc detection result is greater than or equal to the preset optimization threshold, it means that within the preset time, the arc detection result is that there is an arc, that is, an arc phenomenon actually occurs during this time; otherwise, it means that within the preset time, the arc detection result is that there is no arc, that is, no arc phenomenon occurs during this time.

[0067] Among them, the preset optimization threshold is set according to the comprehensive influence degree of actual external factors or other interference factors (such as one or more of the lighting environment, temperature, and humidity environment that can affect the acquisition accuracy of sensors). For example, the preset optimization threshold is set to 60% according to the interference environment of the application scenario. Within 1 second, when the total number of frames is 30 and the number of detected frames is 20, the ratio of the number of detected frames to the total number of frames is greater than the preset optimization threshold at this time, that is, the proportion of the number of frames with detected arcs in the total number of frames is greater than the preset optimization threshold, indicating that the arc detection result within this time period is that there is an arc, and the arc detection accuracy is high.

[0068] The data processing algorithm of the present invention is simple, which is conducive to implementing accurate abnormal arc detection on a relatively low-cost detection processor and improving the efficiency of abnormal arc detection.

[0069] As Figure 4 shown, the present invention provides an arc detection system based on ultraviolet photoelectricity, including:

[0070] An acquisition module 1, configured to acquire ultraviolet photoelectric data and perform data preprocessing;

[0071] Specifically, as Figure 5 shown, the acquisition module 1 includes:

[0072] A data cache unit 11, which performs data cache processing on the acquired ultraviolet photoelectric data to obtain MXN cache data, where the MXN cache data is M frames of data, and each frame includes N data points, and the parameters M and N are both positive integers;

[0073] It should be noted that caching the acquired ultraviolet photoelectric data is convenient for subsequent rapid processing of the MXN cache data and improves the data processing efficiency.

[0074] A data preprocessing unit 12, configured to perform data conversion processing on the MXN stored data to obtain floating-point data.

[0075] It should be noted that data conversion and calibration are performed on the cached ultraviolet photoelectric data to obtain high-precision floating-point data, which can reduce the data storage space and calculation amount, thereby improving the subsequent data processing efficiency and accuracy.

[0076] A feature extraction module 2, configured to perform spectrum feature extraction processing on the preprocessed data to obtain spectrum feature data;

[0077] Specifically, as Figure 6 shown, the feature extraction module 2 includes:

[0078] A feature extraction unit 21, configured to perform FFT transformation processing on the preprocessed data to obtain spectrum feature data.

[0079] It should be noted that by performing FFT transformation on the data, the time-domain data can be converted into frequency-domain data, which is convenient for analyzing the characteristics and frequency components of the signal, so as to obtain the required spectral characteristic data, with high data processing efficiency and accuracy.

[0080] The frequency intercept module 3 is used to intercept and process the spectral characteristic data according to a preset frequency intercept rule to obtain spectral characteristic data within a preset frequency band range;

[0081] It should be noted that the preset frequency intercept rule is set with the required intercepted frequency band range. Through the preset frequency intercept rule, the useful frequency band range can be intercepted and retained in the spectral characteristic data, and the remaining useless frequency band ranges can be removed to reduce the data processing volume and improve the data processing efficiency and data processing accuracy.

[0082] Among them, the intercepted frequency band range is the frequency band range with a large distinguishability in the spectrum for comparing the presence or absence of electric arcs, and this frequency band range can be specifically set according to actual needs.

[0083] The energy calculation module 4 is used to calculate the energy value of each frame in the spectral characteristic data to obtain the energy characteristic data of each frame;

[0084] It should be noted that the spectral characteristic data within the required frequency band range includes multiple frame data, and each frame data includes multiple frequency points. By calculating the sum of the energy values of all frequency points in each frame, the energy value of this frame can be obtained.

[0085] Specifically, the method for calculating the energy value of each frame is as follows:

[0086] The energy value of each frame is calculated through an energy calculation formula, and the formula is as follows:

[0087]

[0088] Among them, freq_feature is the sum of the energy values of all frequency points in each frame and serves as the energy characteristic data, n is the total number of frequency points in the frame data, and D cut(i) 2 is the energy value of the i-th frequency point in the current frame data, and both n and i are positive integers.

[0089] It should be noted that through D cut(i) 2 the energy value of each frequency point after intercepting the data can be calculated, and through the energy calculation formula, the sum of the energy values of all frequency points in each frame data can be calculated to obtain the energy value of each frame data and serve as the energy characteristic data of each frame.

[0090] The detection and processing module 5 is configured to input the energy feature data of each frame into a pre-trained arc detection model to obtain an initial detection result for each frame.

[0091] It should be noted that there are differences or variations in the energy values of different frames. Correspondingly, there are feature differences or variations in the energy feature data of each frame. The arc detection model can identify and process the energy feature data of each frame to quickly obtain an initial detection result with high accuracy. Based on this initial detection result, it can be quickly determined whether there is an arc in this frame, thereby preliminarily determining whether an arc occurs in the circuit.

[0092] The arc optimization module 6 is configured to optimize the initial detection result of each frame according to the arc optimization rule to obtain an optimized arc detection result.

[0093] It should be noted that the arc detection result of a single frame is easily affected by external factors, resulting in a decrease in detection accuracy. In this embodiment, the arc detection result is optimized according to the arc optimization rule to reduce the interference caused by external factors and further improve the accuracy of the final arc detection result.

[0094] Specifically, as Figure 7 shown, the arc optimization module 6 includes:

[0095] The statistics unit 61 is configured to count the number of detected frames with an arc in the arc detection result and the total number of frames of the arc detection result within a preset time.

[0096] It should be noted that the total number of frames of all arc detection results within a preset time is counted, and the number of detected frames with an arc in the total number of frames is counted.

[0097] Preferably, the preset time can be preferably 1S, but it is not limited thereto and can be adjusted according to actual needs.

[0098] The result optimization unit 62 is configured to determine whether the ratio of the number of detected frames to the total number of frames is greater than or equal to a preset optimization threshold. If the determination is yes, it indicates that there is an arc in the arc detection result within the preset time. If the determination is no, it indicates that there is no arc in the arc detection result within the preset time.

[0099] It should be noted that within a preset time, when the ratio of the number of detected frames with an arc to the total number of frames is greater than or equal to the preset optimization threshold, it indicates that there is an arc in the arc detection result within the preset time, that is, an arc phenomenon has indeed occurred during this time; otherwise, it indicates that there is no arc in the arc detection result within the preset time, that is, no arc phenomenon has occurred during this time.

[0100] Among them, the preset optimization threshold is set according to the comprehensive influence degree of actual external factors or other interference factors (such as one or more of the factors such as the lighting environment, temperature and humidity environment that can affect the acquisition accuracy of the sensor). For example, according to the interference environment of the application scenario, the preset optimization threshold is set to 60%. In 1 second, when the total number of frames is 30 and the number of detected frames is 20, at this time, the ratio of the number of detected frames to the total number of frames is greater than the preset optimization threshold, that is, the proportion of the number of frames with detected arcs in the total number of frames is greater than the preset optimization threshold, which means that the arc detection result in this time period is that there is an arc, and the arc detection accuracy is high.

[0101] In summary, the present invention can perform arc detection processing through the collected ultraviolet photoelectric data, improving the arc detection accuracy and detection efficiency.

[0102] Among them, the present invention can perform spectral feature extraction processing on the obtained ultraviolet photoelectric data to obtain spectral feature data; through a preset frequency truncation rule, the spectral feature data can be truncated to obtain useful spectral feature data, reducing the data processing amount and improving the data processing accuracy; by calculating the energy value of each frame in the spectral feature data, the energy feature data of each frame can be obtained; the energy feature data of each frame is input into a pre-trained arc detection model to obtain the arc detection result of each frame, so as to determine whether there is an arc.

[0103] In addition, the present invention can also optimize the arc detection result according to the arc optimization rule to reduce the interference brought by external factors and further improve the accuracy of the detection result.

[0104] Finally, it should be noted that the above embodiments are only specific implementation manners of the present invention, used to illustrate the technical solutions of the present invention, rather than limiting it. The protection scope of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: any person skilled in the art within the technical scope disclosed by the present invention can still modify the technical solutions recorded in the foregoing embodiments or easily think of changes, or perform equivalent replacement on some of the technical features; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. An arc detection method based on ultraviolet optoelectronics, characterized in that, Including: Obtain ultraviolet photoelectric data and perform data preprocessing; Perform spectrum feature extraction processing on the preprocessed data to obtain spectrum feature data; Perform truncation processing on the spectrum feature data according to a preset frequency truncation rule to obtain spectrum feature data within a preset frequency band range; Calculate the energy value of each frame in the spectrum feature data to obtain the energy feature data of each frame; Input the energy feature data of each frame into a pre-trained arc detection model to obtain the initial detection result of each frame; Optimize the initial detection result of each frame according to the arc optimization rule to obtain the optimized arc detection result.

2. The arc detection method according to claim 1, characterized in that, The steps of the data preprocessing include: Perform data caching processing on the collected ultraviolet photoelectric data to obtain MXN cache data, where the MXN cache data is that each frame in M frames of data includes N data points, and the parameters M and N are both positive integers; Perform data conversion processing on the MXN stored data to obtain floating-point data.

3. The arc detection method according to claim 1, wherein The steps of performing spectrum feature extraction processing on the preprocessed data to obtain spectrum feature data include: Perform FFT transformation processing on the preprocessed data to obtain spectrum feature data.

4. The arc detection method according to claim 1, characterized in that, The method for calculating the energy value of each frame is: Calculate the energy value of each frame through an energy calculation formula, and the formula is as follows: Among them, freq_feature is the sum of the energy values of all frequency points in each frame and serves as the energy feature data. n is the total number of frequency points in the current frame data, and Dcut(i ) 2 is the energy value of the i-th frequency point in the current frame data. Both n and i are positive integers.

5. The arc detection method according to claim 1, characterized in that, The arc optimization rule includes: Within a preset time, count the number of detected frames with arcs in the arc detection result and the total number of frames of the arc detection result; Judge whether the ratio of the number of detected frames to the total number of frames is greater than or equal to a preset optimization threshold. If the judgment is yes, it means that the arc detection result within the preset time is that there is an arc. If the judgment is no, it means that the arc detection result within the preset time is that there is no arc.

6. An arc detection system based on ultraviolet photoelectricity, characterized in that, Including: An acquisition module for acquiring ultraviolet photoelectric data and performing data preprocessing; A feature extraction module for performing spectrum feature extraction processing on the preprocessed data to obtain spectrum feature data; A frequency truncation module for performing truncation processing on the spectrum feature data according to a preset frequency truncation rule to obtain spectrum feature data within a preset frequency band range; An energy calculation module for calculating the energy value of each frame in the spectrum feature data to obtain the energy feature data of each frame; A detection processing module for inputting the energy feature data of each frame into a pre-trained arc detection model to obtain the initial detection result of each frame; An arc optimization module for optimizing the initial detection result of each frame according to the arc optimization rule to obtain the optimized arc detection result.

7. The arc detection system according to claim 6, wherein The acquisition module includes: A data caching unit for performing data caching processing on the collected ultraviolet photoelectric data to obtain MXN cache data, where the MXN cache data is that each frame in M frames of data includes N data points, and the parameters M and N are both positive integers; A data preprocessing unit for performing data conversion processing on the MXN stored data to obtain floating-point data.

8. The arc detection system according to claim 6, characterized in that, The feature extraction module includes: A feature extraction unit for performing FFT transformation processing on the preprocessed data to obtain spectrum feature data.

9. The arc detection method according to claim 6, wherein The energy calculation module includes: An energy calculation unit for calculating the energy value of each frame through an energy calculation formula, and the formula is as follows: Among them, freq_feature is the sum of the energy values of all frequency points in each frame and serves as the energy feature data. n is the total number of frequency points in the current frame data, and Dcut(i ) 2 is the energy value of the i-th frequency point in the current frame data. Both n and i are positive integers.

10. The arc detection system according to claim 6, wherein ,, the arc optimization module includes: A statistical unit for counting the number of detected frames with an arc in the arc detection result and the total number of frames of the arc detection result within a preset time; A result optimization unit for determining whether the ratio of the number of detected frames to the total number of frames is greater than or equal to a preset optimization threshold. If the determination is yes, it indicates that there is an arc in the arc detection result within the preset time. If the determination is no, it indicates that there is no arc in the arc detection result within the preset time.

Images