Oil filling equipment intelligent detection method and system based on fluorescence reflection

Abstract

The invention relates to the technical field of oil leakage detection, and provides an oil-filled equipment intelligent detection method and system based on fluorescence reflection, and the method comprises the steps: obtaining a background image collected by each image collection unit when a fluorescence excitation unit is not started; acquiring a fluorescence image acquired by each image acquisition unit when the fluorescence excitation unit is started; determining a common difference image; determining a to-be-monitored area of the oil filling equipment according to the common differential image, controlling a pulse thermal excitation module to send a thermal excitation pulse to the to-be-monitored area, and controlling a set image acquisition unit to acquire fluorescence signal response characteristics of the to-be-monitored area; and the intelligent analysis module traces the leakage area according to the fluorescent signal response characteristics of the insulating oil and the leakage characteristic area. According to the invention, the technical problem of high false alarm rate in the existing insulating oil leakage detection technology based on the ultraviolet fluorescence characteristic of the insulating oil can be solved.

Description

Technical Field

[0001] This invention relates to the field of oil leakage detection technology, and in particular to an intelligent detection method and system for oil-filled equipment based on fluorescence reflection. Background Technology

[0002] Oil-filled equipment is a core component of industrial systems (such as power systems). Leakage of the insulating oil inside the equipment can not only lead to a decrease in insulation performance but also potentially cause serious accidents such as fires and explosions. Traditional methods for detecting oil leaks mainly rely on manual inspections (such as visual observation and test paper testing), which suffer from problems such as long detection cycles, high false negative rates, and inability to respond in real time. Existing online monitoring technologies are mostly based on infrared sensing or acoustic detection, which are easily affected by ambient temperature, equipment vibration, and other factors, and their detection accuracy is difficult to meet industrial requirements.

[0003] Detection technology based on the ultraviolet fluorescence characteristics of insulating oil offers a new approach to solving the aforementioned problems. This technology utilizes ultraviolet light of a specific wavelength to illuminate the transformer surface, exciting fluorescent substances in the leaking oil to emit light, and identifying leaks by detecting the fluorescence signal. However, during use, it was found that existing fluorescence imaging-based detection schemes still suffer from severe detection interference. The reasons are as follows: First, severe ambient light interference: Ambient light such as sunlight and artificial light contains abundant ultraviolet and visible light components, which create strong background noise during imaging, easily drowning out the weak fluorescence signal of the leaking oil, causing the system's signal-to-noise ratio to drop sharply or even fail in daylight or strong light environments; Second, numerous static interference factors: Existing schemes mostly rely on the fluorescence intensity threshold of a single image for judgment. However, reflections on the equipment surface, existing oil stains, paint of specific colors, or other contaminants may also produce similar bright spots under ultraviolet light, leading to frequent false alarms and low accuracy.

[0004] Therefore, existing insulating oil leakage detection technologies based on the ultraviolet fluorescence characteristics of insulating oil have a high false alarm rate. Summary of the Invention

[0005] The main objective of this invention is to provide an intelligent detection method and system for oil-filled equipment based on fluorescence reflection, aiming to solve the technical problem of high false alarm rate in existing insulating oil leakage detection technologies based on the ultraviolet fluorescence characteristics of insulating oil.

[0006] To achieve the above objectives, this invention provides an intelligent detection method for oil-filled equipment based on fluorescence reflection. The method utilizes an intelligent detection system for oil-filled equipment, which includes a control module, an optical path detection module, an intelligent analysis module, and a pulsed thermal excitation module. The optical path detection module includes a fluorescence excitation unit, a first filter unit, a second filter unit, and at least two image acquisition units positioned at different acquisition angles. The method comprises the following steps: Acquire the background images of the oil-filled equipment collected by each image acquisition unit when the fluorescence excitation unit is not turned on; The fluorescence images acquired by each image acquisition unit on the oil-filled equipment when the fluorescence excitation unit is turned on are obtained. The fluorescence image is the equipment image acquired by the image acquisition unit after the ultraviolet light output by the fluorescence excitation unit is irradiated onto the oil-filled equipment by the first filter unit and the stray light is filtered out by the second filter unit. Calculate the difference image between the fluorescence image and the background image corresponding to each image acquisition unit, and register all difference images to the same image reference system to determine the common difference image; The monitoring area of ​​the oil-filled equipment is determined based on the common differential image. The pulse thermal excitation module is controlled to send thermal excitation pulses to the monitoring area, and the set image acquisition unit is controlled to collect the fluorescence signal response characteristics of the monitoring area. The intelligent analysis module traces the leakage area based on the fluorescent signal response characteristics and leakage feature areas of the insulating oil.

[0007] Optionally, the intelligent analysis module's step of tracing the leakage area based on the fluorescent signal response characteristics and leakage characteristic areas of the insulating oil includes: The intelligent analysis module extracts the insulating oil region based on fluorescence signal response characteristics; Obtain the leakage feature area based on the key component area of ​​the oil-filled equipment surface in the image reference frame; The insulating oil area is mapped to the corresponding leakage characteristic area to trace the leakage area.

[0008] Optionally, the image acquisition unit includes a first image acquisition unit disposed at a first acquisition angle and a second image acquisition unit disposed at a second acquisition angle; the first acquisition angle and the second acquisition angle are not equal; the step of acquiring the background images of the oil-filled equipment acquired by each image acquisition unit when the fluorescence excitation unit is not turned on includes: Acquire a first background image of the oil-filled equipment acquired by the first image acquisition unit when the fluorescence excitation unit is not turned on, and acquire a second background image of the oil-filled equipment acquired by the second image acquisition unit when the fluorescence excitation unit is not turned on; The step of acquiring fluorescence images of the oil-filled equipment acquired by each image acquisition unit when the fluorescence excitation unit is turned on includes: Acquire a first fluorescence image of the oil-filled equipment acquired by the first image acquisition unit when the fluorescence excitation unit is turned on, and acquire a second fluorescence image of the oil-filled equipment acquired by the second image acquisition unit when the fluorescence excitation unit is turned on.

[0009] Optionally, the step of calculating the difference image between the fluorescence image and the background image corresponding to each image acquisition unit, and registering all difference images to the same image reference frame to determine the common difference image includes: Calculate the first difference image between the first fluorescence image and the first background image; Calculate the second difference image between the second fluorescence image and the second background image; Obtain the common difference image after the first difference image and the second difference image are registered to the same image reference frame.

[0010] Optionally, the step of obtaining the leakage feature region based on the key component area of ​​the oil-filled equipment surface in the image reference frame includes: The key component areas on the surface of the oil-filled equipment are identified in the image reference system; wherein, the key component areas include: weld area, flange connection area, bolt connection area, flat box wall area, and sleeve base area; Based on the areas of each key component, the leakage characteristic areas are divided.

[0011] Optionally, the standard leakage characteristics of the weld area are defined as: a linear or banded fluorescent distribution that is continuous or discontinuous in space and extends in the same direction as the weld, with a width within a preset pixel range. For flange connection areas, the standard leakage characteristic is defined as: fluorescent areas distributed in a ring or arc shape along the flange sealing ring trajectory; For bolted connection areas, the standard leakage characteristic is defined as: fluorescent areas that are dotted around the bolt holes; For flat box wall areas, the standard leakage characteristic is defined as: there is a high-risk component upstream of the gravity direction that has been identified as a leakage point, and the high-risk component has a leakage path that coincides with the fluorescent area.

[0012] Optionally, the step of the intelligent analysis module extracting the insulating oil region based on fluorescence signal response features includes: In the sequence of fluorescence images of the area to be monitored acquired by the set image acquisition unit, the selected fluorescence images are globally segmented based on the fluorescence signal response to divide them into multiple grids; Within each grid, the pixel with the highest gray value is selected as the candidate seed point; For each candidate seed point, determine whether the gray value exceeds the seed point threshold; Candidate seed points that exceed the seed point threshold are taken as starting points. Using a set growth threshold, region growth is performed in the neighborhood of the starting point to extract complete candidate regions for insulating oil.

[0013] Optionally, the step of the intelligent analysis module extracting the insulating oil region based on fluorescence signal response features further includes: The insulating oil region is determined by whether the fluorescence intensity change of each pixel in the candidate region of insulating oil in the fluorescence image sequence conforms to the fluorescence intensity-time change law of insulating oil.

[0014] Optionally, the step of mapping the insulating oil area to the corresponding leakage feature area to trace the leakage area includes: Obtain the spatial distribution characteristics of pixels in each insulating oil region; Obtain the correlation between the fluorescence intensity distribution of pixels in each insulating oil region and the physical structure of the leakage feature region; Based on the spatial distribution characteristics and the correlation with the physical structure, the leakage area is traced.

[0015] To achieve the above objectives, the present invention also proposes an intelligent detection system for oil-filled equipment based on fluorescence reflection, which is used to detect oil leakage. The intelligent detection system includes a control module, an optical path detection module, an intelligent analysis module, and a pulsed thermal excitation module. The optical path detection module includes a fluorescence excitation unit, a first filter unit, a second filter unit, and at least two image acquisition units set at different acquisition angles.

[0016] The technical solution of this invention helps to solve the problem of high false alarm rate in existing insulating oil leakage detection technologies based on the ultraviolet fluorescence characteristics of insulating oil. Specifically, for the surface to be monitored, background images are acquired at different angles when the fluorescence excitation unit is not turned on, and fluorescence images are acquired at different angles when the fluorescence excitation unit is turned on. For the same image acquisition angle, the difference image between the fluorescence image and the background image can remove static false alarm interference. Furthermore, by registering the difference images from different image acquisition angles to the same image reference system and obtaining a common difference image, the interference of ambient light can be removed. In the monitoring area of ​​the oil-filled equipment determined by the common differential image, the fluorescence signal response characteristics of the monitoring area are collected by sending thermal excitation pulses. Since the fluorescence signal response characteristics of the new oil film of the leaked insulating oil are significantly different from those of old oil stains, other static fluorescent luminescent substances and other interference items, static interference such as paint spots and old oil films can be eliminated based on the fluorescence signal response characteristics that match the insulating oil and combined with the various leakage characteristic areas of the oil-filled equipment. Therefore, the technical solution of the present invention can accurately determine the area where insulating oil exists and trace back to the leakage area, thereby reducing the false alarm rate. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating one embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection in this invention. Figure 2 This is a schematic diagram of the functional modules of the intelligent detection system in this invention.

[0018] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0020] In the following description, the use of suffixes such as "unit," "component," or "element" to denote elements is solely for the purpose of illustrative purposes and has no specific meaning in itself. Therefore, "unit," "component," or "element" may be used interchangeably.

[0021] Please see Figures 1 to 2 This invention provides an intelligent detection method for oil-filled equipment based on fluorescence reflection. The method utilizes an intelligent detection system for oil-filled equipment, which includes a control module, an optical path detection module, an intelligent analysis module, and a pulsed thermal excitation module. The optical path detection module includes a fluorescence excitation unit, a first filtering unit, a second filtering unit, and at least two image acquisition units positioned at different acquisition angles. The method comprises the following steps: Step S10: Obtain the background images of the oil-filled equipment collected by each image acquisition unit when the fluorescence excitation unit is not turned on; Step S20: Obtain the fluorescence images of the oil-filled equipment acquired by each image acquisition unit when the fluorescence excitation unit is turned on. The fluorescence image is the equipment image acquired by the image acquisition unit after the ultraviolet light output by the fluorescence excitation unit is irradiated onto the oil-filled equipment by the first filter unit and the stray light is filtered out by the second filter unit. Step S30: Calculate the difference image between the fluorescence image and the background image corresponding to each image acquisition unit, and register all difference images to the same image reference system to determine the common difference image; Step S40: Determine the monitoring area of ​​the oil-filled equipment based on the common differential image, control the pulse thermal excitation module to send thermal excitation pulses to the monitoring area, and control the set image acquisition unit to acquire the fluorescence signal response characteristics of the monitoring area. In step S50, the intelligent analysis module traces the leakage area based on the fluorescent signal response characteristics and leakage characteristic areas of the insulating oil.

[0022] The technical solution of this invention helps to solve the problem of high false alarm rate in existing insulating oil leakage detection technologies based on the ultraviolet fluorescence characteristics of insulating oil. Specifically, for the surface to be monitored, background images are acquired at different angles when the fluorescence excitation unit is not turned on, and fluorescence images are acquired at different angles when the fluorescence excitation unit is turned on. For the same image acquisition angle, the difference image between the fluorescence image and the background image can remove static false alarm interference. Furthermore, by registering the difference images from different image acquisition angles to the same image reference system and obtaining a common difference image, the interference of ambient light can be removed. In the monitoring area of ​​the oil-filled equipment determined by the common differential image, the fluorescence signal response characteristics of the monitoring area are collected by sending thermal excitation pulses. Since the fluorescence signal response characteristics of the new oil film of the leaked insulating oil are significantly different from those of old oil stains, other static fluorescent luminescent substances and other interference items, static interference such as paint spots and old oil films can be eliminated based on the fluorescence signal response characteristics that match the insulating oil and combined with the various leakage characteristic areas of the oil-filled equipment. Therefore, the technical solution of the present invention can accurately determine the area where insulating oil exists and trace back to the leakage area, thereby reducing the false alarm rate.

[0023] In one specific embodiment, the fluorescence excitation unit uses a pulsed xenon lamp as the excitation light source, the first filtering unit is a first bandpass filter, and the second filtering unit is a second bandpass filter; the image acquisition unit is a high-speed imaging camera (e.g., a high-speed scientific-grade CMOS (Complementary Metal-Oxide Semiconductor) camera with external triggering function, and the control module can be an STM microcontroller (STMicroelectronics microcontroller).

[0024] Specifically, in step S20, the laser generated by the pulsed xenon lamp is filtered by the first bandpass filter and outputs ultraviolet light in the 225~300nm band, which irradiates the oil-filled equipment. If there is oil leakage in the oil-filled equipment, substances such as polycyclic aromatic hydrocarbons in the oil will emit fluorescence in the 320~400nm band after being excited by ultraviolet light. After the stray light is filtered out by the second bandpass filter, the fluorescence image is acquired by the image acquisition unit and the weak fluorescence signal is amplified by the photomultiplier tube (PMT). The control module sequentially performs Gaussian noise reduction and grayscale conversion on the acquired fluorescence images, calculates the threshold using the maximum inter-class variance method and generates a binary image, and quantizes the mean grayscale value of the bright spot region as the fluorescence intensity index; at the same time, it performs analog-to-digital conversion and feature extraction on the electrical signal output by the PMT.

[0025] Furthermore, the pulsed xenon lamp has a wavelength range of 190~2000nm, features short pulse and high peak power characteristics, and consumes ≤5W, making it suitable for complex industrial environments such as substations and oil pipelines.

[0026] Preferably, the center wavelength of the first filter unit is 254nm and the bandwidth is 75nm; the center wavelength of the second filter unit is 360nm and the bandwidth is 80nm, which are used to filter excitation light and ambient light interference.

[0027] The control module can be an STM32 microcontroller, specifically the STM32F407ZGT6 model. It transmits acquired data through DMA (Direct Memory Access) technology, and the ADC (Analog-to-Digital Converter) has a sampling accuracy of ≥12 bits and a sampling rate of ≥1MHz to ensure real-time signal processing.

[0028] Furthermore, the control module can communicate and connect with an IoT cloud platform, which supports data visualization, can automatically generate oil leakage trend curves, and has an automatic abnormal data marking function, thereby realizing intelligent assistance for operation and maintenance decisions. Furthermore, the control module communicates with the IoT cloud platform, uploading the collected background images, fluorescence images, and leakage detection results to the IoT cloud platform to achieve data storage, trend analysis, and real-time monitoring and historical data query of remote terminals.

[0029] Specifically, there may be some static interference items on the surface of oil-filled equipment, such as nameplates and paint spots. These static interference items may also exhibit fluorescence. By taking a difference image between the fluorescence image and the background image at the same angle, obvious static interference items can be removed.

[0030] Furthermore, it is easy to understand that ambient light can cause reflections on the surface of the device. The characteristic of this reflection is that the reflection changes or disappears as the viewing angle is adjusted. Therefore, this invention uses different image acquisition angles to obtain a common difference image, which can eliminate the interference of ambient light.

[0031] Therefore, the detection area determined based on the common difference image includes the insulating oil region, although some interfering regions, such as old lubricating oil films, may also exist. A thermal excitation pulse thermally excites the identified monitoring area, causing its temperature to change over time. Since fluorescence intensity is highly correlated with the temperature of the fluorescent material, increased temperature typically leads to decreased fluorescence efficiency. Therefore, the dynamic response of the monitoring area under the thermal excitation pulse refers to the fluorescence intensity-time change caused by the time-varying fluorescence of the fluorescent material under the thermal excitation pulse, resulting in a dynamic response of the fluorescence signal as fluorescence intensity-time. By utilizing the difference between the time-varying characteristics of insulating oil under thermal excitation and those of other interfering substances, a distinction is made between the fluorescence intensity-time variation characteristics of insulating oil under thermal excitation and those of other fluorescent interfering substances. Using the fluorescence signal response characteristics of insulating oil, specific insulating oil regions are screened out, achieving more accurate insulating oil region screening.

[0032] Furthermore, the presence of insulating oil does not necessarily indicate a leakage area. For example, if leakage has occurred for some time, the leaking insulating oil will flow and create a leakage path. The leakage area will exist within the leakage path. By tracing the leakage area according to the leakage characteristic area defined by the oil-filled equipment, the final leakage area can be determined from the insulating oil area, thus achieving accurate leakage location.

[0033] Based on a first embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection of the present invention, in a second embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection of the present invention, step S50 includes: Step S51: The intelligent analysis module extracts the insulating oil region based on the fluorescence signal response characteristics; Step S52: Obtain the leakage feature area based on the key component area of ​​the oil-filled equipment surface in the image reference system; Step S53: Map the insulating oil area to the corresponding leakage feature area to trace the leakage area.

[0034] In this invention, a difference image is obtained by using fluorescence images and background images at the same angle to remove obvious static interference items. A common difference image is obtained by using difference images at different angles to eliminate interference from ambient light. Therefore, the extraction of the insulating oil region based on the fluorescence signal response characteristics is mainly to eliminate fluorescence interference items that have dried and solidified, such as old oil films (e.g., lubricating oil films).

[0035] Specifically, the principle of extracting insulating oil regions based on fluorescence signal response features under thermal excitation is as follows: Oil-filled equipment housings have high thermal conductivity (the thermal conductivity of metals is generally hundreds of times that of insulating oil), making them excellent conductors of heat. When a thermal pulse is applied, the surface temperature rises rapidly and instantaneously. After the pulse ends, the heat is quickly conducted to the entire oil-filled equipment housing, causing the surface temperature to drop sharply. Therefore, the time curve exhibits a sharp peak characteristic.

[0036] Insulating oil has low thermal conductivity and is a poor conductor of heat. Due to the slow thermal conduction of pulsed heat, the oil layer heats up slowly. After the pulse ends, the heat can only be slowly transferred through the oil layer to the metal substrate of the oil-filled equipment below. Therefore, the cooling process is slow, so the time curve shows a flat and broad characteristic. Similarly, the fluorescence intensity-time curve shows a flat and broad characteristic of slow decline in the early stage and slow rise after the pulse ends.

[0037] A film of insulating oil forms in the leakage area. This new oil film is liquid, wet, and spreadable, forming a continuous liquid film with a clear interface between the oil-filled equipment surface and the metal substrate, and the film thickness is relatively uniform. During thermal excitation, the heat is confined in the continuous oil film for a short time, and the thermal relaxation process is slow and exhibits a broad and gentle hump. Therefore, the obtained fluorescence signal response characteristics refer to the curve of the fluorescence intensity of the new oil film changing over time. This curve shows a slow decrease in the early stage and a slow increase in the later stage during the thermal excitation process, resulting in a broad and gentle curve. For the metal substrate of oil-filled equipment, due to the high thermal conductivity of metal, the thermal relaxation process of thermal excitation is extremely fast, exhibiting a sharp transient peak.

[0038] Interference items such as dried and solidified old oil stains, having already solidified on the surface of the oil-filled equipment, exhibit viscous, dried, and solid characteristics. Furthermore, due to their dried and solidified state, the old oil stains are generally discontinuous in physical state and have penetrated into the microscopic rough structure of the metal substrate on the oil-filled equipment surface, resulting in uneven thickness and oxidation into a solidified film. During thermal excitation, heat is rapidly conducted to the microscopic rough structure of the metal substrate on the oil-filled equipment surface, leading to a rapid thermal relaxation process with sharp peaks. Therefore, in the obtained fluorescence signal response characteristics, the fluorescence intensity curve of the solidified interference items such as old oil stains shows a rapid decrease in the early stage and a rapid increase in the later stage during the thermal excitation process, resulting in a sharp and steep curve.

[0039] That is, because the new oil film has low thermal expansion efficiency and slow thermal relaxation process, the fluorescence intensity change curve over time is flatter. The old oil stain, because it is fused with the metal substrate, has the thermal relaxation properties of the metal substrate, and the thermal relaxation process is faster and the curve is steeper. Therefore, by observing the curve characteristics of the fluorescence intensity change curve over time, the insulating oil area with the leakage of the new oil film can be distinguished from the old oil stain and other areas.

[0040] Furthermore, the system can simulate and generate a standard curve of the fluorescence intensity of the insulating oil film on the surface of the oil-filled equipment under thermal excitation over time. The actual change curve of the area to be monitored is compared with the standard curve to extract the insulating oil area that conforms to the change of the standard curve.

[0041] The insulating oil region includes the leakage area and the leakage path area of ​​the insulating oil leaking from the leakage area. In this invention, different key component areas on the surface of the oil-filled equipment are considered to have different leakage characteristics. Therefore, based on the position of the insulating oil region in the image reference system, the corresponding leakage characteristic area is found, and the leakage area is traced. This facilitates the direct output of the leakage point to the user, eliminating the need for the user to trace the leakage point themselves, and improving the accuracy of detection and positioning.

[0042] Based on a second embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection of the present invention, in a third embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection of the present invention, the image acquisition unit includes a first image acquisition unit disposed at a first acquisition angle and a second image acquisition unit disposed at a second acquisition angle; the first acquisition angle and the second acquisition angle are not equal; step S10 includes: Step S11: Obtain the first background image of the oil-filled equipment acquired by the first image acquisition unit when the fluorescence excitation unit is not turned on, and obtain the second background image of the oil-filled equipment acquired by the second image acquisition unit when the fluorescence excitation unit is not turned on. Step S20 includes: Step S21: Obtain the first fluorescence image of the oil-filled equipment acquired by the first image acquisition unit when the fluorescence excitation unit is turned on, and obtain the second fluorescence image of the oil-filled equipment acquired by the second image acquisition unit when the fluorescence excitation unit is turned on.

[0043] In this embodiment, two image acquisition units are used, positioned at different acquisition angles. A second filter unit is provided in front of each image acquisition unit.

[0044] Specifically, regarding the requirements for the target signal (fluorescence): fluorescence scatters in all directions. Therefore, when observing the same real oil stain fluorescent area from different angles, its brightness should be as close as possible. Thus, in this invention, the two viewing angles should not differ too much to avoid significant differences in the observed fluorescence brightness due to excessive tilting of the viewing angle.

[0045] Regarding the requirements for interference signals (reflected light from ambient light): specular reflection has extremely strong directionality. The positions of the two image acquisition units should be arranged to maximize the difference in the intensity of the reflected light they receive. That is, one image acquisition unit should avoid strong reflections as much as possible, while the other image acquisition unit may receive the reflected light.

[0046] In this invention, the angle between the first image acquisition unit and the second image acquisition unit is preferably between 30° and 40°. This angle range can effectively enhance the distinguishability of environmental reflection interference signals while ensuring the consistency of observation of the target fluorescence signal.

[0047] Furthermore, the fields of view of the two image acquisition units and the irradiation field of the ultraviolet light excited by the fluorescence excitation unit must cover the same area to be monitored, and the arrangement of the image acquisition units should avoid the specular reflection path of the ultraviolet light source. That is, the image acquisition units should not be directly facing the direction of ultraviolet light reflection on the surface of the equipment.

[0048] In this design, a first image acquisition unit can be used as the main acquisition unit, responsible for acquiring images with high signal intensity. A second image acquisition unit serves as an auxiliary acquisition unit, forming a distinct acquisition angle with the first image acquisition unit and the ultraviolet light source. This acquisition angle ensures that the reflection interference characteristics observed by the second image acquisition unit are significantly different from those observed by the first image acquisition unit. The fluorescence excitation unit is positioned close to the first image acquisition unit.

[0049] In a third embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection of the present invention, and in a fourth embodiment of the same method, step S30 includes: Step S31: Calculate the first difference image between the first fluorescence image and the first background image; Step S32: Calculate the second difference image between the second fluorescence image and the second background image; Step S33: Obtain the common difference image after the first difference image and the second difference image are registered to the same image reference frame.

[0050] ; in, For the pixels in the first difference image grayscale value, For the pixels in the first fluorescence image grayscale value, For the pixels in the first background image grayscale value; ; in, For the pixels in the second difference image grayscale value, For pixels in the second fluorescence image grayscale value, For pixels in the second background image grayscale value; Because the two image acquisition units are located at different spatial positions and angles, the same physical scene captured in the images exhibits translation, rotation, and perspective distortion. Therefore, the second difference image needs to be mapped to the same image reference system based on the first difference image using an image registration algorithm to obtain the registered second difference image. The purpose of this is to establish the spatial mapping relationship between corresponding pixels in the two images.

[0051] After being registered to the same image reference frame, the pixels in the first difference image and the second difference image are subjected to a minimum value operation to obtain a common difference image.

[0052] ; in, For pixels in the common difference image grayscale value, The second difference image is obtained by mapping the second difference image to the same image reference frame based on the first difference image.

[0053] In this embodiment, only pixels that exhibit significant brightness from both different viewing angles are retained. Interference such as specular reflection, due to its strong directionality, typically appears as bright pixels in the difference image from one viewing angle, while being weak or absent in the difference image from another. Therefore, its intensity is greatly suppressed after the minimum value operation. Conversely, the true fluorescence signal, due to its diffuse reflectance characteristics, exists in all viewing angles and can thus be well preserved in the common difference image.

[0054] In a second embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection according to the present invention, in a fifth embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection according to the present invention, step S52 includes: Step S521: In the image reference system, identify the key component areas on the surface of the oil-filled equipment; wherein, the key component areas include: weld area, flange connection area, bolt connection area, flat box wall area and sleeve base area; Step S522: Divide the leakage characteristic areas according to the areas of each key component.

[0055] Specifically, the pre-marked key component areas of the oil-filled equipment are mapped onto an image reference system to identify the various key component areas on the surface of the oil-filled equipment within the image reference system.

[0056] Based on the areas of each key component, leakage characteristic areas are divided, and standard leakage characteristics of each leakage characteristic area are defined according to the attributes and location of the leakage characteristic areas.

[0057] Based on the fifth embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection of the present invention, and the sixth embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection of the present invention: The standard leakage characteristics of the weld area are defined as: a linear or banded fluorescent distribution that is continuous or discontinuous in space and extends in the same direction as the weld, with a width within a preset pixel range. For flange connection areas, the standard leakage characteristic is defined as: fluorescent areas distributed in a ring or arc shape along the flange sealing ring trajectory; For bolted connection areas, the standard leakage characteristic is defined as: fluorescent areas that are dotted around the bolt holes; For flat box wall areas, the standard leakage characteristic is defined as: there is a high-risk component upstream of the gravity direction that has been identified as a leakage point, and the high-risk component has a leakage path that coincides with the fluorescent area.

[0058] In a second embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection according to the present invention, and in a seventh embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection according to the present invention, step S51 includes: Step S511: In the fluorescence image sequence of the area to be monitored acquired by the set image acquisition unit, the selected fluorescence image is globally segmented based on the fluorescence signal response to divide it into multiple grids. Step S512: Within each grid, select the pixel with the highest gray value as the candidate seed point; Step S513: For each candidate seed point, determine whether the gray value exceeds the seed point threshold; Step S514: Using candidate seed points that exceed the seed point threshold as starting points, and using a set growth threshold, perform region growth in the neighborhood of the starting point to extract the complete candidate region for insulating oil.

[0059] Specifically, in step S40, the controlled image acquisition unit acquires the fluorescence signal response characteristics of the area to be monitored. This is done by acquiring fluorescence images of the area to be monitored at different times before, during, and after the thermal excitation pulse is generated. The fluorescence intensity of the image at each time moment is obtained, thus generating a fluorescence image sequence of the area to be monitored. The fluorescence image sequence is arranged according to the acquisition time. The fluorescence intensity corresponding to the same pixel in each fluorescence image changes over time. When the fluorescence intensity change of this pixel conforms to the fluorescence intensity-time variation law of insulating oil, then this pixel corresponds to a pixel in the insulating oil region.

[0060] Therefore, in this embodiment, steps S511 to S514 are used to quickly find candidate regions of insulating oil, which is beneficial for subsequently judging whether the fluorescence intensity change of each pixel in the candidate region conforms to the fluorescence intensity-time change law of insulating oil, thereby finding the accurate insulating oil region.

[0061] Step S511 is used to perform global gridding of the selected fluorescence image, dividing it spatially into several rectangular grids of equal size. Specifically, in this embodiment, the selected fluorescence image is the fluorescence image of the area to be monitored acquired at the initial moment before the pulse thermal excitation module is started, because the fluorescence intensity is stronger at this time.

[0062] The grid size is optimized based on the image resolution and the expected minimum size of the leak spot. Specifically, the image can be divided into a 20x20 pixel or 50x50 pixel grid. Global grid segmentation helps avoid all seed points being concentrated in a certain part of the image, ensuring the ability to detect leaks over a large area.

[0063] In step S512, the core of the true fluorescent region typically has high fluorescence brightness. This step is equivalent to pre-selecting within each small region to find the point most likely to be the signal. Only one point is retained per grid, which filters out the vast majority of pixels within the grid, including most low-value noise and scattered noise, providing an extremely concise candidate point set for subsequent processing. All candidate seed points selected from the grids are traversed, and their gray values ​​are compared with a threshold. Only those points with gray values ​​exceeding the threshold are retained as high-confidence valid seed points, effectively reducing the amount of data processing. Furthermore, using candidate seed points exceeding the seed point threshold as starting points, a region growth is performed in the neighborhood of the starting point using a set growth threshold, thereby extracting the complete insulating oil candidate region. This also helps to obtain the variation of fluorescence signal intensity with the position of each pixel, thus aiding in the extraction of the starting point and direction of insulating oil leakage.

[0064] In the seventh embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection of the present invention, and in the eighth embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection of the present invention, step S51 further includes: Step S515: Determine the insulating oil region based on whether the fluorescence intensity change of each pixel in the candidate region of insulating oil in the fluorescence image sequence conforms to the fluorescence intensity-time change law of insulating oil.

[0065] In the eighth embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection of the present invention, and in the ninth embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection of the present invention, step S53 includes: Step S531: Obtain the spatial distribution characteristics of pixels in each insulating oil region; Step S532: Obtain the correlation between the fluorescence intensity distribution of pixels in each insulating oil region and the physical structure of the leakage feature region; Step S533: Based on the spatial distribution characteristics and the correlation with the physical structure, trace the leakage area.

[0066] Specifically, based on the spatial distribution characteristics of pixels in each insulating oil region and the leakage characteristic regions divided based on the key component areas on the surface of the oil-filled equipment, leakage paths can be simulated.

[0067] Based on the correlation between the fluorescence intensity distribution of pixels in each insulating oil region and the physical structure of the leakage feature region, it is helpful to correct the simulated leakage path to the physical structure of the leakage feature region, and to determine the leakage region based on the corrected leakage path, which is the leakage point.

[0068] In a ninth embodiment of the intelligent detection method for oil-filled equipment based on fluorescence reflection of the present invention, and in a tenth embodiment of the same invention, step S53 further includes: Step S535: Trace the leakage characteristic area based on the leakage area and the corrected leakage path; Step S536: When the corrected leakage path matches the standard leakage characteristics of the leakage area, output the current leakage area as the detection result. Step S537: When the corrected leakage path does not conform to the standard leakage characteristics of the leakage area, trace the direction of gravity to see if there is a continuous fluorescent signal based on the three-dimensional structural model of the oil-filled equipment and the direction of gravity. Step S538: If a continuous fluorescence signal exists, the correct leakage area is determined based on the trajectory of the fluorescence signal.

[0069] To achieve the above objectives, the present invention also proposes an intelligent detection system for oil-filled equipment based on fluorescence reflection, which is used to detect oil leakage. The intelligent detection system includes a control module, an optical path detection module, an intelligent analysis module, and a pulsed thermal excitation module. The optical path detection module includes a fluorescence excitation unit, a first filter unit, a second filter unit, and at least two image acquisition units set at different acquisition angles.

[0070] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms, or by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to enter the methods described in the various embodiments of the present invention.

[0071] In the description of this specification, references to terms such as "one embodiment," "another embodiment," "other embodiments," or "first embodiment to Xth embodiment," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, method steps, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0072] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0073] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0074] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A smart detection method for oil-filled equipment based on fluorescence reflection, characterized in that, The method employs an intelligent detection system for oil-filled equipment. This system includes a control module, an optical path detection module, an intelligent analysis module, and a pulsed thermal excitation module. The optical path detection module comprises a fluorescence excitation unit, a first filter unit, a second filter unit, and at least two image acquisition units positioned at different acquisition angles. The method includes the following steps: Acquire the background images of the oil-filled equipment collected by each image acquisition unit when the fluorescence excitation unit is not turned on; The fluorescence images acquired by each image acquisition unit on the oil-filled equipment when the fluorescence excitation unit is turned on are obtained. The fluorescence image is the equipment image acquired by the image acquisition unit after the ultraviolet light output by the fluorescence excitation unit is irradiated onto the oil-filled equipment by the first filter unit and the stray light is filtered out by the second filter unit. Calculate the difference image between the fluorescence image and the background image corresponding to each image acquisition unit, and register all difference images to the same image reference system to determine the common difference image; The monitoring area of ​​the oil-filled equipment is determined based on the common differential image. The pulse thermal excitation module is controlled to send thermal excitation pulses to the monitoring area, and the set image acquisition unit is controlled to collect the fluorescence signal response characteristics of the monitoring area. The intelligent analysis module traces the leakage area based on the fluorescent signal response characteristics and leakage feature areas of the insulating oil.

2. The intelligent detection method for oil-filled equipment based on fluorescence reflection according to claim 1, characterized in that, The intelligent analysis module traces the leakage area based on the fluorescent signal response characteristics and leakage characteristic areas of the insulating oil, including the following steps: The intelligent analysis module extracts the insulating oil region based on fluorescence signal response characteristics; Obtain the leakage feature area based on the key component area of ​​the oil-filled equipment surface in the image reference frame; The insulating oil area is mapped to the corresponding leakage feature area to trace the leakage area.

3. The intelligent detection method for oil-filled equipment based on fluorescence reflection according to claim 2, characterized in that, The image acquisition unit includes a first image acquisition unit disposed at a first acquisition angle and a second image acquisition unit disposed at a second acquisition angle; The first acquisition angle is not equal to the second acquisition angle; The step of acquiring background images of the oil-filled equipment by each image acquisition unit when the fluorescence excitation unit is not turned on includes: Acquire a first background image of the oil-filled equipment acquired by the first image acquisition unit when the fluorescence excitation unit is not turned on, and acquire a second background image of the oil-filled equipment acquired by the second image acquisition unit when the fluorescence excitation unit is not turned on; The step of acquiring fluorescence images of the oil-filled equipment acquired by each image acquisition unit when the fluorescence excitation unit is turned on includes: Acquire a first fluorescence image of the oil-filled equipment acquired by the first image acquisition unit when the fluorescence excitation unit is turned on, and acquire a second fluorescence image of the oil-filled equipment acquired by the second image acquisition unit when the fluorescence excitation unit is turned on.

4. The intelligent detection method for oil-filled equipment based on fluorescence reflection according to claim 3, characterized in that, The steps of calculating the difference image between the fluorescence image and the background image corresponding to each image acquisition unit, registering all difference images to the same image reference frame, and determining the common difference image include: Calculate the first difference image between the first fluorescence image and the first background image; Calculate the second difference image between the second fluorescence image and the second background image; Obtain the common difference image after the first difference image and the second difference image are registered to the same image reference frame.

5. The intelligent detection method for oil-filled equipment based on fluorescence reflection according to claim 2, characterized in that, The step of obtaining the leakage feature region based on the key component area of ​​the oil-filled equipment surface in the image reference frame includes: The key component areas on the surface of the oil-filled equipment are identified in the image reference system; wherein, the key component areas include: weld area, flange connection area, bolt connection area, flat box wall area, and sleeve base area; Based on the areas of each key component, the leakage characteristic areas are divided.

6. The intelligent detection method for oil-filled equipment based on fluorescence reflection according to claim 5, characterized in that: The standard leakage characteristics of the weld area are defined as: a linear or banded fluorescent distribution that is continuous or discontinuous in space and extends in the same direction as the weld, with a width within a preset pixel range. For flange connection areas, the standard leakage characteristic is defined as: fluorescent areas distributed in a ring or arc shape along the flange sealing ring trajectory; For bolted connection areas, the standard leakage characteristic is defined as: fluorescent areas that are dotted around the bolt holes; For flat box wall areas, the standard leakage characteristic is defined as: there is a high-risk component upstream of the gravity direction that has been identified as a leakage point, and the high-risk component has a leakage path that coincides with the fluorescent area.

7. The intelligent detection method for oil-filled equipment based on fluorescence reflection according to claim 2, characterized in that, The intelligent analysis module extracts the insulating oil region based on fluorescence signal response characteristics, including the following steps: In the sequence of fluorescence images of the area to be monitored acquired by the set image acquisition unit, the selected fluorescence images are globally segmented based on the fluorescence signal response to divide them into multiple grids; Within each grid, the pixel with the highest gray value is selected as the candidate seed point; For each candidate seed point, determine whether the gray value exceeds the seed point threshold; Candidate seed points that exceed the seed point threshold are taken as starting points. Using a set growth threshold, region growth is performed in the neighborhood of the starting point to extract complete candidate regions for insulating oil.

8. The intelligent detection method for oil-filled equipment based on fluorescence reflection according to claim 7, characterized in that, The intelligent analysis module's step of extracting the insulating oil region based on fluorescence signal response characteristics also includes: The insulating oil region is determined by whether the fluorescence intensity change of each pixel in the candidate region of insulating oil in the fluorescence image sequence conforms to the fluorescence intensity-time change law of insulating oil.

9. The intelligent detection method for oil-filled equipment based on fluorescence reflection according to claim 8, characterized in that, The step of mapping the insulating oil area to the corresponding leakage feature area to trace the leakage area includes: Obtain the spatial distribution characteristics of pixels in each insulating oil region; Obtain the correlation between the fluorescence intensity distribution of pixels in each insulating oil region and the physical structure of the leakage feature region; Based on the spatial distribution characteristics and the correlation with the physical structure, the leakage area is traced.

10. An intelligent detection system for oil-filled equipment based on fluorescence reflection, characterized in that, The method for intelligent detection of oil-filled equipment based on fluorescence reflection as described in any one of claims 1 to 9 is used for oil leakage detection; the intelligent detection system includes a control module, an optical path detection module, an intelligent analysis module, and a pulse thermal excitation module; the optical path detection module includes a fluorescence excitation unit, a first filter unit, a second filter unit, and at least two image acquisition units set at different acquisition angles.

Images