A method and device for online monitoring of visible light and non-visible light

Abstract

The present invention relates to the field of multi-spectral detection technology, and specifically to an online monitoring method and device for visible light and non-visible light. The method analyzes the radiation change characteristics of spectral data in different bands in the multi-spectral data at a single time position in the monitoring time period to obtain the gas breakdown strength value; obtains the stage transition moment according to the change trend of the gas breakdown strength value in the continuous period and the visible light radiation situation; determines the temperature response section by the radiation intensity of the infrared band after the stage transition moment, and determines the effective response section by the change characteristics of the infrared spectrum data in the temperature response section; and trains the monitoring model with the effective response section. The present invention divides the initial response stage by analyzing the radiation situation in the visible light and non-visible light bands before the visible arc phenomenon, and performs more detailed monitoring and identification training to improve the online monitoring sensitivity of arc faults.

Description

Technical Field

[0001] The present invention relates to the technical field of multi-spectrum detection, and in particular to an online monitoring method and device for visible light and non-visible light. Background Art

[0002] In the power system, high and low voltage switchgear is a key power equipment used to distribute and control the transmission of power. During its operation, arc faults may occur, which is an abnormal current flow phenomenon caused by dielectric breakdown. Arcs will not only damage electrical equipment and reduce system efficiency, but may also cause serious safety accidents, such as fire or explosion. Therefore, the application of arc protection in high and low voltage switchgear is particularly important. Arc protection technology can respond in a very short time by quickly detecting and isolating arc faults, preventing fault expansion and equipment damage. In order to achieve this goal, visible light and non-visible light online monitoring methods and devices provide an efficient solution.

[0003] At present, common online monitoring methods and devices are mainly based on visible light detection technology, infrared detection technology, ultraviolet detection technology, multimodal data fusion, etc. Although the existing technology has improved the online monitoring capabilities of high and low voltage switchgear to a certain extent, due to the complex operating environment and changeable conditions of high and low voltage switchgear, the traditional single-mode monitoring method cannot meet the comprehensive analysis requirements of the equipment operating status. In a complex electromagnetic environment, false alarms may occur, resulting in unnecessary protection actions. Before the arc phenomenon occurs, some specific non-visible light phenomena will occur. The existing monitoring methods cannot accurately identify the data change behavior before the arc phenomenon occurs, which leads to a high delay in the system's response to the fault, reducing the overall performance of the visible light and non-visible light online monitoring system. Summary of the invention

[0004] In order to solve the technical problem that the prior art cannot accurately identify the data change behavior before the arc phenomenon occurs, which leads to a high delay in the system's response to the fault, the purpose of the present invention is to provide a visible light and non-visible light online monitoring method and device, and the technical solution adopted is as follows:

[0005] The present invention provides an online monitoring method for visible light and non-visible light, the method comprising:

[0006] In the monitoring period before the historical arc phenomenon, the radiation intensity of the multi-spectrum at each moment position is obtained; the multi-spectrum includes ultraviolet band, infrared band and visible light band;

[0007] According to the fluctuation characteristics of the radiation intensity of the multi-spectrum in different bands at each moment, the gas breakdown strength value at each moment is obtained; according to the change trend of the gas breakdown strength value at the moment during the monitoring period and the degree of change of the radiation intensity in the local visible light band, the stage transition moment is obtained;

[0008] The temperature response section is determined by the magnitude of the radiation intensity in the infrared band after the stage transition moment; the effective response section is determined according to the radiation intensity variation characteristics in the infrared band at different time positions in the temperature response section and the stage transition moment;

[0009] The monitoring model is obtained based on the multispectral data on the effective response segment.

[0010] Furthermore, the method for obtaining the gas breakdown strength value includes:

[0011] For any time position, the distribution size and distribution change frequency of the radiation intensity between the visible light band and the ultraviolet band at that time position are obtained to obtain the possible breakdown radiation index at that time position;

[0012] According to the continuous change degree of the radiation intensity in the infrared band at the position at that moment, a possible indicator of the breakdown temperature at that moment is obtained;

[0013] The gas breakdown strength value at the moment is obtained by combining the possible breakdown radiation index and the possible breakdown temperature index at the moment.

[0014] Furthermore, the method for obtaining the possible indicators of the breakdown radiation includes:

[0015] At each time position, the standard deviation of all radiation intensities in each band is taken as the radiation fluctuation of each band; the average value of the wavelength interval between each two adjacent radiation intensity maxima in each band is calculated as the average fluctuation length in each band; the ratio of the radiation fluctuation of each band to the average fluctuation length is taken as the high-frequency radiation of each band at each time position;

[0016] The product of the value after negative correlation mapping of the high-frequency radiance in the visible light band at the moment and the high-frequency radiance in the ultraviolet band is used as a possible indicator of breakdown radiation at the moment.

[0017] Furthermore, the method for obtaining the possible indicator of the breakdown temperature includes:

[0018] Perform curve fitting on the radiation intensity in the infrared band at the position at that moment to obtain a radiation intensity curve;

[0019] The average value of the corresponding slope at each radiation intensity on the radiation intensity curve is calculated to obtain a possible indicator of the breakdown temperature at that moment.

[0020] Furthermore, the method for obtaining the stage transition moment includes:

[0021] Arrange all gas breakdown strength values ​​in the monitoring time period in ascending order to obtain a gas breakdown sequence;

[0022] Calculate the difference between each two adjacent gas breakdown strength values ​​in the gas breakdown sequence to obtain the strength difference; take the time position corresponding to the two gas breakdown strength values ​​when the strength difference is the largest as the time to be divided; take the last time to be divided in the sequence as the maximum time to be divided, and take the first time to be divided in the sequence as the minimum time to be divided;

[0023] The average radiation intensity of the visible light band at each time position is taken as the visible light radiation uniformity at each time position; the product of the visible light radiation uniformity at each time position and the gas breakdown strength value is taken as the change significance at each time position;

[0024] The ratio of the change significance of the minimum time to be divided and the previous time in the time series is used as the possible division index of the minimum time to be divided;

[0025] The sum of the change significance of the minimum time to be divided and the previous time position in the time series is taken as the local change significance value; the ratio of twice the change significance index of the maximum time to be divided to the local change significance value is taken as the possible division index of the maximum time to be divided;

[0026] The time to be divided corresponding to the maximum possible division index is taken as the stage transition time.

[0027] Furthermore, the method for obtaining the temperature response segment includes:

[0028] The average value of all radiation intensities in the infrared band at each moment after the stage transition is taken as the temperature representation value at each moment;

[0029] Starting from the stage transition moment, traverse in chronological order. When there is a temperature characterization value at a moment position that is greater than a preset temperature threshold, the corresponding moment position is used as the temperature response moment and the traversal is stopped.

[0030] The period from the stage transition moment to the temperature response moment is the temperature response segment.

[0031] Furthermore, the method for obtaining the effective segment of the response includes:

[0032] For any time position in the temperature response segment, the temperature characteristic value at the time position is obtained according to the distribution change of all radiation intensities in the infrared fluctuation segment at the time position;

[0033] Calculate the mean of the temperature characteristic values ​​at all time positions except the time position in the temperature response section to obtain the residual characteristic value at the time position;

[0034] The difference between the temperature characteristic value at the time position and the residual characteristic value is used as a possible effective indicator at the time position;

[0035] The time position corresponding to the maximum possible effective index in the temperature response segment is taken as the effective cutoff time;

[0036] The period from the phase transition moment to the effective end moment is the response effective period.

[0037] Furthermore, the method for obtaining the temperature characteristic value includes:

[0038] The product of the standard deviation and the average value of all radiation intensities in the infrared band at the position at that moment is taken as the temperature characteristic value at that moment.

[0039] Furthermore, the acquisition of the monitoring model based on the multispectral data on the effective response segment includes:

[0040] All multispectral data in the effective response period are taken as input, and the response demand is taken as output. The monitoring model is obtained by training through the neural network model. The input of the monitoring model is the multispectral data in the time period, and the output is the monitoring response result.

[0041] The present invention also provides an online monitoring device for visible light and non-visible light, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of an online monitoring method for visible light and non-visible light as described in any one of the above items are implemented.

[0042] The present invention has the following beneficial effects:

[0043] The present invention obtains the gas breakdown strength value characterizing the initial characteristics of the arc phenomenon through the radiation change characteristics of the spectral data on different bands of visible light and invisible light in the multi-spectral data at a single moment position in the monitoring time period, and comprehensively considers the initial characteristics of invisible light and visible light that are not apparent, so as to provide data support for the subsequent stage division. Then, according to the change trend of the gas breakdown strength value in the continuous period and the visible light radiation situation, the time point that may transition from the gas breakdown stage to the arc formation stage is obtained to more accurately determine the initial transition point of the arc generation phenomenon. Since the temperature will change significantly in the initial stage of the arc, and the temperature response has a hysteresis, the temperature response section is determined by the radiation intensity of the infrared band, and the reference time period with more significant changes is first determined. Further, the change characteristics of the infrared spectrum data in the temperature response section reflect the sudden change deviation of the temperature, and the effective response time period is determined. Finally, the effective time period of data monitoring before the arc phenomenon occurs is used to train the monitoring model. The present invention divides the initial response stage by analyzing the radiation conditions in the visible light and non-visible light bands before the visible arc phenomenon, performs more precise monitoring and identification training, and improves the online monitoring sensitivity of arc faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0045] Figure 1 A flow chart of an online monitoring method for visible light and non-visible light provided by an embodiment of the present invention;

[0046] Figure 2 A flow chart of a method for obtaining a gas breakdown strength value provided by one embodiment of the present invention;

[0047] Figure 3 A schematic diagram of a phase transition moment provided by an embodiment of the present invention;

[0048] Figure 4 A schematic diagram of a response valid segment provided for an embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the online monitoring method and device for visible light and non-visible light proposed by the present invention, its specific implementation method, structure, characteristics and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0050] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0051] The specific scheme of the online monitoring method and device for visible light and non-visible light provided by the present invention is described in detail below with reference to the accompanying drawings.

[0052] See also Figure 1 , which shows a flow chart of an online monitoring method for visible light and non-visible light provided by an embodiment of the present invention, the method comprising the following steps:

[0053] S1: During the monitoring period before the historical arc phenomenon, the multi-spectrum radiation intensity at each moment position is obtained; the multi-spectrum includes ultraviolet band, infrared band and visible light band.

[0054] Generally, the main locations where arc phenomena occur are key locations such as circuit breaker contacts, contactor contacts, and cable connection ends. In the embodiments of the present invention, multispectral sensors are installed at key locations inside the power system, such as cable joints, bus ducts, or circuit breaker outlets, to ensure that the sensors have high sensitivity and anti-interference capabilities and can adapt to complex electrical environments. A high-performance central processing unit is also installed in the online device, and the data processing unit corrects the data set in real time to obtain new monitoring data. In order to complete the data analysis, the wavelength band of the multispectral sensor is selected from the wavelength band of 100-1200nm, covering multiple main wavelength bands, and the 100-400nm band is recorded as the ultraviolet band, the 400-700nm band is recorded as the visible light band, and the 700-1200nm band is recorded as the infrared light band, where the ultraviolet band and the infrared light band are both non-visible light bands.

[0055] Since the arc process is not completely presented in the form of visible light, usually before the arc phenomenon of visible light occurs, there will be arc phenomena that are invisible to the naked eye. The occurrence of arc phenomenon involves the complete process from gas breakdown, arc formation, arc maintenance to arc extinction. Therefore, multispectral sensors can use the detection data of visible light bands and non-visible light bands to conduct precise monitoring and identification of the entire stage of arc phenomenon, thereby providing reliable and accurate data support for the online monitoring system.

[0056] When a visible arc light phenomenon occurs in the monitoring system, the monitoring system will issue an early warning and make a protection mechanism. In order to reduce the lag of monitoring, the abnormal fault situation is mainly analyzed based on the abnormal situation before the arc phenomenon occurs. In an embodiment of the present invention, a period of 10 minutes before the historical arc phenomenon is used as the monitoring time period. The radiation intensity of the multi-spectrum can be collected at each time position in the monitoring time period. The multi-spectrum includes three different bands: ultraviolet, infrared and visible light.

[0057] S2: According to the fluctuation characteristics of the radiation intensity of the multi-spectrum in different bands at each moment, the gas breakdown intensity value at each moment is obtained; according to the change trend of the gas breakdown intensity value at the moment during the monitoring period and the degree of change of the radiation intensity in the local visible light band, the stage transition moment is obtained.

[0058] The arc phenomenon includes the gas breakdown and arc formation stages before the light phenomenon. First, the gas breakdown is characterized by combining the fluctuations in radiation intensity in the multi-spectrum to assist in the subsequent stage division and determination. During gas breakdown, electron excitation and molecular collisions lead to strong radiation peaks and large fluctuations. During the gas breakdown stage, the ultraviolet intensity has large and high-frequency radiation fluctuations, while the visible light band has small and low-frequency radiation fluctuations. As the gas breakdown proceeds, the radiation intensity gradually increases, but the amplitude of the fluctuation is relatively small. The radiation changes in the infrared band are often relatively gentle. As the temperature rises due to arc discharge, the infrared radiation gradually increases, but there will be no obvious pulse fluctuations like ultraviolet rays. The fluctuation frequency of infrared radiation is low, and the radiation intensity usually shows a slowly increasing trend.

[0059] Therefore, in the embodiment of the present invention, the method for obtaining the gas breakdown strength value can be found in Figure 2 , which shows a flow chart of a method for obtaining a gas breakdown strength value provided by an embodiment of the present invention, the method comprising the following steps:

[0060] S201: For any moment position, the distribution size and distribution change frequency of the radiation intensity at the moment position between the visible light band and the ultraviolet band are obtained to obtain a possible breakdown radiation index at the moment position.

[0061] From the perspective of the high frequency of radiation, combined with the radiation intensity in the ultraviolet and visible light bands, an analysis is conducted. The greater the radiation intensity in the ultraviolet band and the higher the frequency, the more significant the breakdown characteristics are. And because it occurs in the early stage, the radiation intensity in the visible light band is small and low-frequency.

[0062] In the embodiment of the present invention, at each time position, the standard deviation of all radiation intensities on each band is used as the radiation fluctuation degree of each band, and considering that there are radiation intensities corresponding to multiple wavelengths on the band, the standard deviation is used to reflect the degree of radiation fluctuation. The average value of the wavelength interval between each two adjacent radiation intensity maxima on each band is calculated as the average fluctuation length on each band. The radiation intensity can be curve-fitted in each band, and the maximum value and the wavelength interval between each two adjacent maxima are obtained by the derivative method. The mean of all wavelength intervals is used as the average fluctuation length to reflect the frequency of radiation fluctuation.

[0063] The ratio of the radiation fluctuation degree of each band to the average fluctuation length is further taken as the high-frequency radiation degree of each band at each time position. When the radiation fluctuation degree is larger and the average fluctuation length is smaller, the reflection intensity is greater and the frequency is higher, so the high-frequency radiation degree is greater.

[0064] Finally, the product of the negative correlation mapping of the high-frequency radiation in the visible light band at that moment and the high-frequency radiation in the ultraviolet band is used as a possible indicator of the breakdown radiation at that moment. When the high-frequency intensity in the visible light band is weaker and the high-frequency intensity in the ultraviolet band is stronger, the breakdown phenomenon is more severe and is moving towards the arc formation stage.

[0065] It should be noted that negative correlation mapping is a technical means well known to those skilled in the art, and may be in the form of inverse proportion or negative exponential power, etc., which will not be elaborated or limited here. As an example, the expression of the possible indicator of breakdown radiation is:

[0066] ;in, Expressed as The breakdown radiation at the position at a certain moment may be an indicator, Expressed as The ultraviolet band at the time position The radiation fluctuation, Expressed as The ultraviolet band at the time position The average fluctuation length of Expressed as The visible light band at the time position The radiation fluctuation, Expressed as The visible light band at the time position The average fluctuation length. Expressed as The ultraviolet band at the time position The high frequency radiation, Expressed as The visible light band at the time position The value after negative correlation mapping of the high-frequency radiance.

[0067] S202: Obtain a possible indicator of the breakdown temperature at the location at the moment according to the degree of continuous change of the radiation intensity in the infrared band at the location at the moment.

[0068] The intensity of the infrared band reflects the temperature conditions of key parts. As the radiation caused by gas breakdown becomes stronger and stronger, the temperature trend will become more and more obvious. When the gas breakdown reaches its peak, an arc phenomenon will occur. Therefore, the degree of breakdown can be analyzed from the significant trend of the continuous change of radiation intensity in the infrared band.

[0069] In the embodiment of the present invention, the radiation intensity in the infrared band at the moment position is curve fitted to obtain a radiation intensity curve. It should be noted that curve fitting is a technical means well known to those skilled in the art, and can be sampled by the least square method, etc., which is not limited or elaborated here. The average value of the corresponding slope size at each radiation intensity on the radiation intensity curve is calculated to obtain a possible indicator of the breakdown temperature at the moment position, and the trend of the temperature intensity at each moment is reflected by the size of the slope value.

[0070] S203: Combining the possible breakdown radiation index and the possible breakdown temperature index at the moment and position, obtain the gas breakdown strength value at the moment and position.

[0071] The combination of radiation intensity frequency and temperature reflects the degree of characterization of breakdown strength. In the embodiment of the present invention, the product of the breakdown radiation possible index and the breakdown temperature possible index at the time position is taken as the gas breakdown strength value at the time position. The larger the breakdown radiation possible index and the breakdown temperature possible index are, the more significant the breakdown phenomenon is, and the closer it is to the initial stage of arc generation, so the gas breakdown strength value is higher.

[0072] At this point, the preliminary analysis of the gas breakdown intensity corresponding to the moment in the monitoring period is completed. Further, the continuous change of the gas breakdown intensity value in the continuous period and the intensity change that tends to form in the visible light band can be combined to determine the stage change and obtain the stage transition moment.

[0073] Preferably, in an embodiment of the present invention, the method for obtaining the phase transition moment includes:

[0074] First, all gas breakdown strength values ​​in the monitoring time period are arranged in ascending order to obtain a gas breakdown sequence, and the sequence after sorting is used to determine and analyze the large deviation. Therefore, the difference between each two adjacent gas breakdown strength values ​​in the gas breakdown sequence is calculated to obtain the strength difference, and the node where the mutation occurs is determined by the degree of difference.

[0075] Further, the time position corresponding to the two gas breakdown strength values ​​when the intensity difference is the largest is taken as the time to be divided. Since the maximum deviation includes two moments, there may be two moments that belong to the initial stage of arc formation, but there is also only one moment that belongs to the initial stage of arc formation, which requires further analysis. Therefore, the last moment to be divided in the sequence is taken as the maximum moment to be divided, and the first moment to be divided in the sequence is taken as the minimum moment to be divided for subsequent analysis.

[0076] When the gas breaks down and triggers an arc, the temperature of the arc rises sharply, and the ionization degree of the gas increases, resulting in a large amount of thermal radiation and luminescence. During the arc formation stage, as the arc discharge continues, the temperature rises, and a large amount of visible light radiation is emitted from the arc area, so the visible light radiation intensity will show a significant increase.

[0077] Therefore, further, the average value of the radiation intensity in the visible light band at each moment is taken as the visible light radiation uniformity at each moment, and the product of the visible light radiation uniformity at each moment and the gas breakdown strength value is taken as the significance of the change at each moment, and the possibility is comprehensively analyzed in combination with the visible light radiation situation and the breakdown strength.

[0078] The ratio of the significance of change between the minimum time to be divided and the previous time in the time series is used as the possible division index of the minimum time to be divided. The larger the possible division index is, the more likely the minimum time to be divided is to be in the initial stage of the arc, and the more likely it is to be the division moment.

[0079] The sum of the change significance of the minimum moment to be divided and the previous moment in the time series is taken as the local change significance value, and the ratio of twice the change significance index of the maximum moment to be divided to the local change significance value is taken as the possible division index of the maximum moment to be divided. The degree of increased deviation of the change significance of the maximum moment to be divided relative to the overall increase of the previous sequence is measured. The higher the possible division index is, the better the division is when the maximum division moment is the initial stage of the arc phenomenon.

[0080] Finally, the time to be divided corresponding to the maximum possible division index is used as the stage transition time, and the period before the stage transition time can be used as the gas breakdown time period. Figure 3 , which shows a schematic diagram of a phase transition moment provided by an embodiment of the present invention, where Q represents the phase transition moment.

[0081] S3: Determine the temperature response section by the radiation intensity in the infrared band after the stage transition moment; determine the effective response section according to the radiation intensity variation characteristics in the infrared band at different time positions in the temperature response section and the stage transition moment.

[0082] When the arc phenomenon is in the invisible stage and is about to turn into a visible light phenomenon, the temperature data changes significantly. For the online monitoring system, because the significant change in temperature indicates that the arc phenomenon has occurred, that is, the temperature response has a certain hysteresis, that is, the monitoring data in the response delay period is very important for the early warning mechanism of the online monitoring system, because the data in this period contains the change trend of the visible light data before the invisible light changes significantly.

[0083] Therefore, it is more important to analyze the response stage before the temperature begins to change significantly. The analysis at this stage can improve the sensitivity of the response to arc anomalies, and provide early warning before the arc phenomenon occurs, which can protect the system safety to a greater extent. Therefore, the initial temperature response stage is determined first.

[0084] In an embodiment of the present invention, a method for obtaining the temperature response stage includes:

[0085] First, the average of all radiation intensities in the infrared band at each moment after the stage transition is taken as the temperature representation value at each moment, and the temperature intensity is reflected from the infrared radiation intensity.

[0086] Further, starting from the stage transition moment, the traversal is performed in chronological order. When the temperature characterization value at a time position is greater than the preset temperature threshold, the corresponding time position is used as the temperature response moment, and the traversal is stopped to determine the time point when the temperature exceeds the threshold due to the possible arc light, that is, the time when the visible light phenomenon occurs. In the embodiment of the present invention, the preset temperature threshold can be set to 125, and the specific value can be adjusted by the implementer according to the specific implementation scenario, which is not limited here.

[0087] The period from the stage transition moment to the temperature response moment is the temperature response section. This stage is from the initial stage of the arc phenomenon to the stage of arc light phenomenon. As the gas breakdown phenomenon continues and energy accumulates in the air, a temperature explosion critical point will appear in the temperature response section. The critical point will produce a large amount of heat radiation, providing a basis for the luminescence phenomenon. Therefore, the multi-spectral anomaly in the accumulation stage is the analysis stage that requires midpoint attention. The required response effective section is further analyzed through the sudden change in radiation intensity in the infrared band.

[0088] Preferably, in an embodiment of the present invention, the method for obtaining a valid response segment includes:

[0089] First, for any moment position in the temperature response segment, the temperature characteristic value of the moment position is obtained according to the distribution change of all radiation intensities in the infrared fluctuation segment at the moment position. The temperature characteristic value reflects the stability of the radiation intensity at each moment position. In an embodiment of the present invention, the product of the standard deviation and the average value of all radiation intensities in the infrared band at the moment position is used as the temperature characteristic value of the moment position. The larger the temperature characteristic value, the higher the radiation intensity and the worse the stability, which reflects the temperature accumulation characteristics of this moment position.

[0090] Further calculate the mean of the temperature characteristic values ​​of all the time positions except the time position in the temperature response section to obtain the residual characteristic value of the time position. Combined with the mean value outside each time position, the degree of mutation can be reflected by the difference later. Therefore, the difference between the temperature characteristic value of the time position and the residual characteristic value is used as an effective possible indicator of the time position. The higher the degree of deviation, the greater the possibility that the time position is a critical moment.

[0091] Finally, the time position corresponding to the maximum possible effective index in the temperature response segment is taken as the effective cutoff time, the accumulation outbreak time is determined, and the period from the stage transition time to the effective cutoff time is taken as the effective response segment. The multi-spectral data analysis of the effective response segment at this time can accelerate the analysis and early warning of arc anomalies. Figure 4 , which shows a schematic diagram of a response effective segment provided by an embodiment of the present invention, Q represents the stage transition moment, J represents the effective cut-off moment, R represents the temperature response moment, and the time period from Q to J is the response effective segment.

[0092] S4: Obtain monitoring model based on multispectral data on the effective response segment.

[0093] By obtaining multiple effective response segments in the monitoring time period before multiple historical arc phenomena, the richness of the data set can be improved to build an online monitoring model, record the data in the effective response segment, take all the multi-spectral data in the effective response segment as input, and take the response demand as output, and train the monitoring model through the neural network model. In an embodiment of the present invention, the features extracted by the features of the spectral data in the effective response segment include the shape of the spectral waveform, the rate of change, the mean and standard deviation in the time window, the spectrum characteristics and other data, and the features extracted from the multi-spectral data are used as input features, and the corresponding system response requirements are marked as output labels. The neural network model can be used to train a large amount of multi-spectral data in the effective response segment. After the model training is completed and its accuracy is verified, it can be deployed in the online monitoring system, and new spectral data can be collected in real time for prediction, and potential arc phenomena can be discovered in time. In addition, for the abnormal response of the online monitoring system, the system's response speed to abnormal situations can be effectively improved.

[0094] It should be noted that the model training process is a well-known technical means well known to those skilled in the art, and the training process is not limited here. By timely identifying and processing the signs before a fault occurs, the number of unplanned downtimes is reduced, the overall reliability of the system is improved, the monitoring model is further optimized, and the adaptive ability of the online monitoring system and the accuracy of monitoring are improved.

[0095] In summary, in the multi-spectral data at a single moment position in the monitoring time period, the present invention obtains the gas breakdown strength value that characterizes the initial characteristics of the arc phenomenon through the radiation change characteristics of the spectral data on different bands of visible light and invisible light, and comprehensively considers the initial characteristics of invisible light and visible light that are not apparent, so as to provide data support for the subsequent stage division. Then, according to the change trend of the gas breakdown strength value in the continuous period and the visible light radiation situation, the time point that may transition from the gas breakdown stage to the arc formation stage is obtained to more accurately determine the initial transition point of the arc generation phenomenon. Since the temperature will change significantly in the initial stage of the arc, and the temperature response has a hysteresis, the temperature response section is determined by the radiation intensity of the infrared band, and the reference time period with more significant changes is first determined. Further, the change characteristics of the infrared spectrum data in the temperature response section reflect the sudden change deviation of the temperature, and the effective response time period is determined. Finally, the effective time period of data monitoring before the arc phenomenon occurs is used to train the monitoring model. The present invention divides the initial response stage by analyzing the radiation conditions in the visible light and non-visible light bands before the visible arc phenomenon, performs more precise monitoring and identification training, and improves the online monitoring sensitivity of arc faults.

[0096] The present invention also provides an online monitoring device for visible light and non-visible light, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of an online monitoring method for visible light and non-visible light as described in any one of the above items are implemented.

[0097] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0098] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for online monitoring of visible light and non-visible light, characterized in that: The method comprises: In the monitoring period before the historical arc phenomenon, the radiation intensity of the multi-spectrum at each moment position is obtained; the multi-spectrum includes ultraviolet band, infrared band and visible light band; According to the fluctuation characteristics of the radiation intensity of the multi-spectrum in different bands at each moment, the gas breakdown strength value at each moment is obtained; according to the change trend of the gas breakdown strength value at the moment during the monitoring period and the degree of change of the radiation intensity in the local visible light band, the stage transition moment is obtained; The temperature response section is determined by the magnitude of the radiation intensity in the infrared band after the stage transition moment; the effective response section is determined according to the radiation intensity variation characteristics in the infrared band at different time positions in the temperature response section and the stage transition moment; Acquire monitoring model based on multispectral data on effective response segment; The method for obtaining the gas breakdown strength value comprises: For any time position, the distribution size and distribution change frequency of the radiation intensity between the visible light band and the ultraviolet band at that time position are obtained to obtain the possible breakdown radiation index at that time position; According to the continuous change degree of the radiation intensity in the infrared band at the position at that moment, a possible indicator of the breakdown temperature at that moment is obtained; The gas breakdown strength value at the moment is obtained by combining the possible breakdown radiation index and the possible breakdown temperature index at the moment.

2. The method for online monitoring of visible light and non-visible light according to claim 1, characterized in that: The method for obtaining the possible indicators of breakdown radiation includes: At each time position, the standard deviation of all radiation intensities in each band is taken as the radiation fluctuation of each band; the average value of the wavelength interval between each two adjacent radiation intensity maxima in each band is calculated as the average fluctuation length in each band; the ratio of the radiation fluctuation of each band to the average fluctuation length is taken as the high-frequency radiation of each band at each time position; The product of the value after negative correlation mapping of the high-frequency radiance in the visible light band at the moment and the high-frequency radiance in the ultraviolet band is used as a possible indicator of breakdown radiation at the moment.

3. The method for online monitoring of visible light and non-visible light according to claim 1, characterized in that: The method for obtaining the possible indicator of the breakdown temperature includes: Perform curve fitting on the radiation intensity in the infrared band at the position at that moment to obtain a radiation intensity curve; The average value of the corresponding slope at each radiation intensity on the radiation intensity curve is calculated to obtain a possible indicator of the breakdown temperature at that moment.

4. The method for online monitoring of visible light and non-visible light according to claim 1, characterized in that: The method for obtaining the phase transition moment includes: Arrange all gas breakdown strength values ​​in the monitoring time period in ascending order to obtain a gas breakdown sequence; Calculate the difference between each two adjacent gas breakdown strength values ​​in the gas breakdown sequence to obtain the strength difference; take the time position corresponding to the two gas breakdown strength values ​​when the strength difference is the largest as the time to be divided; take the last time to be divided in the sequence as the maximum time to be divided, and take the first time to be divided in the sequence as the minimum time to be divided; The average radiation intensity of the visible light band at each time position is taken as the visible light radiation uniformity at each time position; the product of the visible light radiation uniformity at each time position and the gas breakdown strength value is taken as the change significance at each time position; The ratio of the change significance of the minimum time to be divided and the previous time in the time series is used as the possible division index of the minimum time to be divided; The sum of the change significance of the minimum time to be divided and the previous time position in the time series is taken as the local change significance value; the ratio of twice the change significance index of the maximum time to be divided to the local change significance value is taken as the possible division index of the maximum time to be divided; The time to be divided corresponding to the maximum possible division index is taken as the stage transition time.

5. The method for online monitoring of visible light and non-visible light according to claim 1, characterized in that: The method for obtaining the temperature response segment includes: The average value of all radiation intensities in the infrared band at each moment after the stage transition is taken as the temperature representation value at each moment; Starting from the stage transition moment, traverse in chronological order. When there is a temperature characterization value at a moment position that is greater than a preset temperature threshold, the corresponding moment position is used as the temperature response moment and the traversal is stopped. The period from the stage transition moment to the temperature response moment is the temperature response segment.

6. The method for online monitoring of visible light and non-visible light according to claim 1, characterized in that: The method for obtaining the effective segment of the response includes: For any time position in the temperature response segment, the temperature characteristic value at the time position is obtained according to the distribution change of all radiation intensities in the infrared fluctuation segment at the time position; Calculate the mean of the temperature characteristic values ​​at all time positions except the time position in the temperature response section to obtain the residual characteristic value at the time position; The difference between the temperature characteristic value at the time position and the residual characteristic value is used as a possible effective indicator at the time position; The time position corresponding to the maximum possible effective index in the temperature response segment is taken as the effective cutoff time; The period from the phase transition moment to the effective end moment is the response effective period.

7. The method for online monitoring of visible light and non-visible light according to claim 6, characterized in that: The method for obtaining the temperature characteristic value comprises: The product of the standard deviation and the average value of all radiation intensities in the infrared band at the position at that moment is taken as the temperature characteristic value at that moment.

8. The method for online monitoring of visible light and non-visible light according to claim 1, characterized in that: The monitoring model is obtained based on multispectral data on the effective response segment, including: All multispectral data in the effective response period are taken as input, and the response demand is taken as output. The monitoring model is obtained by training through the neural network model. The input of the monitoring model is the multispectral data in the time period, and the output is the monitoring response result.

9. An online monitoring device for visible light and non-visible light, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the online monitoring method of visible light and non-visible light as described in any one of claims 1 to 8 are implemented.

Images