A co-cavity lubricating oil iron filings visual detection method and system

Abstract

The present application relates to the technical field of image recognition, in particular to a kind of common cavity lubricating oil iron filings visual inspection method and system, wherein, method includes: extracting iron filings area from oil image and groove image;Determine the iron filings density index corresponding to each detection site, and determine the groove accumulation change rate;For each machine, according to the change between the iron filings density index corresponding to the upstream detection site and downstream detection site of each machine and the groove accumulation change rate, determine the relative deterioration index of each machine;According to the comparative relationship of the relative deterioration index of machine and upstream adjacent machine and the distance between machines, the relative deterioration index is corrected to obtain the corrected deterioration index.The present application eliminates the masking effect of upstream iron filings conduction on downstream machine in series common cavity oil circuit, realizes the accurate evaluation of the real loss of the filter device of each machine, and provides a reliable basis for the maintenance decision of each machine.

Description

Technical Field

[0001] This invention relates to the field of image recognition technology, specifically to a visual detection method and system for metal filings in a common cavity lubricating oil. Background Technology

[0002] In fields such as electric drive systems for new energy vehicles, gearboxes and flat-wire motors typically employ a common-cavity lubrication system. This means the gearbox gears and motor share the same lubrication circuit, with multiple machines connected in series within this circuit. Each machine is independently equipped with a fine filter and a strong magnetic adsorption device. The oil flows sequentially through each machine and its filter before converging into a storage trench at the end of the oil circuit. Iron filings generated by gearbox gear wear, if carried by the oil into the flat-wire motor, can adhere to the stator and rotor surfaces and cut the enameled wire, ultimately leading to insulation failure or even burnout. To address this, existing technologies typically employ multi-layered protection measures, including titanium alloy or ceramic-plated gears, fine filters combined with strong magnetic adsorption, and storage trenches within the oil circuit, to intercept iron filings in the lubricating oil. With the development of machine vision technology, existing technologies have also emerged that use image enhancement and static impurity classification to visually detect impurities in lubricating oil, obtaining the overall content of impurities such as iron filings in the oil.

[0003] However, the aforementioned existing technologies have at least the following drawbacks: Existing visual inspection solutions for lubricating oil can only identify and classify impurities in the oil as a whole, making it difficult to accurately assess the actual filtration loss of each machine's filtration system in the common cavity oil circuit. In a common cavity oil circuit, multiple machines are connected in series, and oil flows between them. If the protective mechanism of any upstream machine fails, the resulting metal filings will be carried by the oil to the downstream machine, increasing the metal filings at the downstream machine's inlet. This causes the degradation of the downstream machine's own protective capabilities to be masked by apparent interception data. Therefore, existing visual inspection methods cannot distinguish whether the increase in metal filings originates from the machine's own filtration degradation or is carried upstream, thus failing to accurately assess the actual filtration status of each specific machine in the common cavity oil circuit and failing to provide a reliable basis for maintenance decisions for each machine. Summary of the Invention

[0004] This invention provides a visual inspection method and system for common cavity lubricating oil iron filings to solve the above-mentioned problems.

[0005] The visual inspection method and system for common cavity lubricating oil iron filings in this invention adopts the following technical solution: In a first aspect, one embodiment of the present invention provides a visual inspection method for metal filings in a common cavity lubricating oil, the method comprising: For multiple machines connected in series in the common cavity oil circuit, oil images are acquired at the upstream and downstream detection positions of the filter device of each machine, and trench images are acquired at the storage trench at the end of the common cavity oil circuit. The iron filings area is extracted from the oil images and the trench images. The iron filings density index corresponding to each detection position is determined based on the distribution characteristics of the iron filings region in the image corresponding to each detection position, and the trench accumulation change rate is determined based on the change of the degree of iron filings accumulation in the trench image over time; wherein, the iron filings density index is used to characterize the distribution density of iron filings in the oil at the corresponding detection position. For each machine, the relative deterioration index is determined based on the change between the iron filings density index corresponding to the upstream and downstream detection positions and the trench accumulation change rate; wherein, the relative deterioration index is used to characterize the degree of deterioration of the protective capability of the corresponding machine's filter device. Based on the comparison between the relative deterioration index of the machine and its upstream neighboring machine, and the distance between the machines, the relative deterioration index is corrected to obtain a corrected deterioration index; wherein, the corrected deterioration index is used to characterize the actual wear and tear of the machine's filtration device.

[0006] Further, determining the relative deterioration index of the machine based on the change in the chip density index between the upstream and downstream detection positions and the trench accumulation change rate includes: For each machine, the contribution of each machine to the accumulation of iron filings in the storage trench is determined based on the change between the iron filing density index corresponding to the upstream and downstream detection positions of each machine and the trench accumulation change rate; wherein, the contribution degree is used to characterize the magnitude of the contribution of the iron filing change of the corresponding machine to the change of iron filings accumulation in the storage trench. Based on the changes in the contribution level of each machine over time, the machine failure rate of each machine is determined; wherein, the machine failure rate is used to characterize the rate at which the protective capability of the corresponding machine's filtration device deteriorates over time. Using the machine with the smallest absolute value of the machine failure rate among all machines as the benchmark, the machine failure rate of each machine is compared with the machine failure rate of the benchmark machine to obtain the relative deterioration index of each machine.

[0007] Further, determining the trench accumulation change rate based on the change in the degree of iron filings accumulation in the trench image over time includes: The trench fill rate is determined based on the coverage of the iron filings region in the trench image of the storage trench; wherein, the trench fill rate is used to characterize the degree to which the storage trench is covered by iron filings; When the trench fill rate indicates that the storage trench is not filled with iron filings, the degree of iron filings accumulation is determined based on the trench fill rate; when the trench fill rate indicates that the storage trench is filled with iron filings, the degree of iron filings accumulation is determined based on the grayscale difference between the accumulated protruding area and the surrounding area in the trench image. The trench accumulation rate is determined based on the change in the degree of iron filings accumulation over time, and the trench accumulation change rate is determined based on the change in the trench accumulation rate over time; wherein, the trench accumulation rate is used to characterize the speed at which iron filings accumulate in the storage trench per unit time, and the trench accumulation change rate is used to characterize the overall failure severity of the filter device of each machine.

[0008] Further, determining the degree of iron filings accumulation based on the grayscale difference between the accumulated protruding area and the surrounding area in the trench image includes: Using the region with the highest grayscale in the trench image as the center, determine the position with the greatest grayscale change along the direction away from the center, and determine the boundary between the piled-up protrusion region and the surrounding region based on the position with the greatest grayscale change; The degree of iron filings accumulation is determined based on the difference between the grayscale of the accumulated protruding area inside the boundary and the grayscale of the outer perimeter area outside the boundary.

[0009] Furthermore, the method also includes: The failure of the pre-filtration stage is determined based on the trench accumulation change rate; when the trench accumulation change rate indicates that the iron filings in the storage trench are accumulating abnormally at an accelerated rate, the pre-filtration stage is determined to be in failure. When it is determined that the pre-filtering process is in failure, the failure of the pre-filtering process is traced back to the source machine by machine according to the relative deterioration index of each machine. The pre-filtration stage includes the filtration devices of each machine.

[0010] Further, the step of correcting the relative deterioration index based on the comparison between the relative deterioration index of the machine and its upstream neighboring machine, and the distance between the machines, to obtain a corrected deterioration index, includes: Based on the comparison of the relative deterioration index of the machine with that of the upstream adjacent machine, it is determined whether the machine is covered by the upstream adjacent machine; wherein, when the comparison indicates that the relative deterioration index of the machine is abnormally low relative to the relative deterioration index of the upstream adjacent machine, it is determined that the machine is covered by the upstream adjacent machine. For a machine that is obscured by an upstream neighboring machine, the relative deterioration index of the machine is corrected based on the distance between the machine and the upstream neighboring machine to obtain the corrected deterioration index of the machine.

[0011] Furthermore, the correction of the relative deterioration index of the machine includes: The weight of the loss impact of each upstream machine on the machine is determined based on the distance between the machine and each upstream machine; wherein, the closer the upstream machine is to the machine, the greater the weight of the loss impact. Based on the loss impact weight of each upstream machine and the masking status of each upstream machine, the loss impact factor of the machine is determined; wherein, the loss impact factor is used to characterize the cumulative impact of the failure of each upstream machine on the machine. The relative deterioration index of the machine is corrected based on the loss influence factor to obtain the corrected deterioration index of the machine.

[0012] Furthermore, the method also includes: Each machine is classified according to its modified deterioration index; wherein the classification includes at least: When the machine's corrected deterioration index meets the first determination condition, the machine is determined to be in a normal state. When the machine's correction deterioration index meets the second determination condition, a planned maintenance prompt for the machine is output; wherein, the degree of wear corresponding to the second determination condition is higher than the degree of wear corresponding to the first determination condition; When the machine's correction deterioration index meets the third determination condition, a shutdown and maintenance prompt is output for the machine; wherein the degree of wear and tear corresponding to the third determination condition is higher than the degree of wear and tear corresponding to the second determination condition.

[0013] Furthermore, the method also includes: For machines that output a shutdown and maintenance prompt, control the solenoid valve in the corresponding lubricating oil circulation circuit of the machine to cut off the oil flow in the lubricating oil circulation circuit, so as to prevent iron filings from spreading to downstream machines; Generate a maintenance work order for the machine; wherein the maintenance work order includes at least the parts of the machine that need to be replaced and the recommended maintenance window time, and the parts that need to be replaced include at least one of a fine filter element and a strong magnetic adsorption device.

[0014] Secondly, another embodiment of the present invention provides a visual inspection system for metal scrap in a common cavity lubricating oil, comprising an image acquisition unit and a data processing unit, wherein: The image acquisition unit is used to acquire oil images at the upstream and downstream detection positions of the filter device of each of the multiple machines connected in series in the common cavity oil circuit, and to acquire trench images at the storage trench at the end of the common cavity oil circuit. The data processing unit is used to extract iron filings regions from the oil image and the trench image; determine the iron filings density index corresponding to each detection position based on the distribution characteristics of the iron filings regions in the image corresponding to each detection position, and determine the trench accumulation change rate based on the change in the degree of iron filings accumulation in the trench image over time; wherein, the iron filings density index is used to characterize the distribution density of iron filings in the oil at the corresponding detection position; for each machine, a relative deterioration index is determined based on the change between the iron filings density index corresponding to the upstream and downstream detection positions of each machine and the trench accumulation change rate; wherein, the relative deterioration index is used to characterize the degree of deterioration of the protective capability of the filter device of the corresponding machine; and correct the relative deterioration index based on the comparison relationship between the relative deterioration index of each machine and the upstream adjacent machine and the distance between the machines to obtain a corrected deterioration index; wherein, the corrected deterioration index is used to characterize the actual wear and tear of the filter device of the corresponding machine.

[0015] The beneficial effects of the technical solution of the present invention are: In this embodiment of the invention, for multiple machines connected in series in a common cavity oil circuit, oil images are acquired at the upstream and downstream detection positions of the filter device of each machine, and a trench image is acquired at the storage trench at the end of the common cavity oil circuit. Iron filings regions are extracted from the oil images and trench images. The iron filings density index corresponding to each detection position is determined based on the distribution characteristics of the iron filings region in the image corresponding to each detection position, and the trench accumulation change rate is determined based on the change in the degree of iron filings accumulation in the trench image over time. The iron filings density index is used to characterize the density of iron filings distribution in the oil at the corresponding detection position. For each machine, a relative deterioration index is determined based on the change between the iron filings density index corresponding to the upstream and downstream detection positions and the trench accumulation change rate. The relative deterioration index is used to characterize the degree of deterioration in the protective capability of the filter device of the corresponding machine. The relative deterioration index is corrected based on the comparison between the relative deterioration index of the machine and its upstream adjacent machine, as well as the distance between the machines, to obtain a corrected deterioration index. The corrected deterioration index is used to characterize the actual wear and tear of the machine's filter device.

[0016] This invention, by acquiring oil images at the upstream and downstream detection positions of each machine's filter device and at the storage trench at the end of the common cavity oil circuit, determines the iron filings density index based on the distribution characteristics of the iron filings area at each detection position. It then combines this with the trench accumulation change rate to determine the relative deterioration index of each machine. Further corrections are made based on the comparison of the relative deterioration indices of adjacent machines and the distance between machines. This eliminates the masking effect of upstream iron filings conduction on downstream machines in the series common cavity oil circuit, achieving an accurate assessment of the true wear and tear of each machine's filter device and providing a reliable basis for maintenance decisions. Furthermore, by combining the changes in the iron filings density index between the upstream and downstream detection positions of each machine with the trench accumulation change rate to determine the contribution level, and obtaining the machine failure rate based on the change in contribution level over time, this invention compares the machines with the lowest absolute failure rates, eliminating the influence of global fluctuations in the common cavity oil circuit and individual differences between machines on wear assessment. Interference allows for the accurate location of machines with deteriorated protection capabilities. Furthermore, by quantifying the degree of iron filings accumulation based on the trench filling rate when the storage trench is not full, and quantifying the degree of iron filings accumulation based on the grayscale difference between the raised accumulation area and the surrounding area after the trench is full, a continuous quantitative index is constructed, ranging from planar fullness to three-dimensional overflow. Based on the trench accumulation change rate, the failure of the pre-filter is determined, enabling real-time perception and machine-by-machine traceability of the failure status of multi-level pre-filters. Additionally, by comparing the relative deterioration index of adjacent machines to determine whether downstream machines are covered, corrections are only made for covered machines according to the distance-related loss impact weight, avoiding false corrections caused by excessive propagation of upstream abnormal signals downstream. Combined with graded alarms based on the corrected deterioration index and the cut-off control of the oil circuit of the failed machine, the diffusion of iron filings to downstream machines and motor cavities is blocked, forming a complete closed loop from detection, location, early warning to maintenance intervention. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic flowchart of the visual inspection method for co-cavity lubricating oil iron filings provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the oil circuit structure and detection position arrangement of the common cavity lubricating oil metal chip visual inspection system provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the architecture of the visual inspection system for common cavity lubricating oil and iron filings provided in an embodiment of the present invention.

[0019] The numbers in the diagram are as follows: 201. First motor system; 202. First fine filter; 203. First strong magnetic adsorption device; 204. Second motor system; 205. Second fine filter; 206. Second strong magnetic adsorption device; 207. Motor stator oil return chamber; 208. Storage trench; 209. First camera; 210. Second camera; 211. Third camera; 212. Fourth camera; 300. Vision inspection system for lubricating oil chips in common cavity; 310. Image acquisition unit; 320. Data processing unit. Detailed Implementation

[0020] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following detailed description, in conjunction with the accompanying drawings and preferred embodiments, provides a detailed account of the specific implementation methods, structures, features, and effects of the method proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments may be combined in any suitable form. Unless otherwise defined, 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 pertains.

[0021] The following describes in detail, with reference to the accompanying drawings, a specific scheme for a visual inspection method and system for common cavity lubricating oil iron filings provided by the present invention.

[0022] like Figure 1 As shown, this embodiment of the invention provides a visual inspection method for metal chips in a common cavity lubricating oil, including: Step S110: For multiple machines connected in series in the common cavity oil circuit, oil images are collected at the upstream and downstream detection positions of the filter device of each machine, and a groove image is collected at the storage groove 208 at the end of the common cavity oil circuit. The iron filings area is extracted from the oil image and the groove image.

[0023] A common-cavity oil circuit refers to an oil circuit structure where the gearbox gears and the flat wire motor share the same lubrication oil circuit. Figure 2Taking the illustrated oil circuit structure as an example, the first motor system 201, the first fine filter 202, the first strong magnetic adsorption device 203, the second motor system 204, the second fine filter 205, and the second strong magnetic adsorption device 206 are arranged in series along the oil flow direction. The oil flows sequentially through the machine, the fine filter, and the strong magnetic adsorption device before entering the next machine, and finally flows into the storage trench 208 located at the end of the oil circuit. There is only one storage trench 208, located in the motor stator return oil chamber 207 at the very end. The oil in the motor stator return oil chamber 207 is resupplyed to each machine through an external return oil pipeline, forming a closed loop. The upstream detection position refers to the detection position located upstream of the filter inlet of the machine, used to collect images of the oil before it enters the filter; the downstream detection position refers to the detection position located downstream of the filter outlet, used to collect images of the oil after filtration and adsorption; the iron filings area refers to the connected pixel area in the image that is distinguishable from the oil background in grayscale, corresponding to iron filings.

[0024] In the image acquisition stage, you can refer to Figure 2 The diagram shows the installation of a detection camera on the corresponding pipeline for each machine. For each machine, a transparent detection window is installed upstream of the inlet pipeline of the fine filter, and a dark-field illumination industrial camera is configured, for example: Figure 2 The first camera 209 is used to capture images of the oil before it enters the second fine filter 205; the second camera 210 is used to capture images of the oil after it has been processed by the second fine filter 205; the third camera 211 is used to capture images of the oil after it has been intercepted and strongly magnetically adsorbed by the fine filter; in addition, a separate camera can be set for the storage groove 208 of the motor stator oil return chamber 207, such as... Figure 2 As shown in the fourth camera 212, the fourth camera 212 is used to focus on the trench area to acquire trench images. The reason for using dark field illumination is that the lubricating oil appears as a dark background under dark field illumination, while the iron filings appear as bright targets due to surface scattering, forming a high contrast between the particles and the background, which provides favorable imaging conditions for subsequent grayscale-based target extraction; the transparent detection window provides an optical path for real-time imaging of the oil flowing in the pipeline, so that the acquisition process does not require interruption of the oil circulation.

[0025] When extracting the iron filings region from the acquired image, denoising and contrast enhancement processing can be performed on the grayscale image first. One optional implementation is to use a 5×5 window median filter to remove salt-and-pepper noise and sensor noise from the image. Median filtering suppresses impulse noise while preserving the edge information of the particle target. Subsequently, histogram equalization is used to stretch the grayscale dynamic range of the image, improving the overall contrast and widening the grayscale difference between the low-brightness fine iron filings and the background against a dark background. It should be noted that the above denoising and enhancement methods are only optional implementations. In practice, alternative methods such as Gaussian filtering, bilateral filtering, and adaptive histogram equalization can also be used, as long as they can improve the distinction between the particle target and the background.

[0026] On the preprocessed image, a morphological white-hat transformation can be further performed to highlight the iron filings target. The white-hat transformation is defined as subtracting the opening operation result from the original image. The opening operation uses a structuring element to erode and then dilate the image, removing bright details smaller than the structuring element. After subtracting the opening operation result from the original image, the removed bright details are extracted, thus highlighting bright areas smaller than the structuring element. Under dark lighting, iron filings appear as bright, fine particles, while the oil background and pipe reflections appear as large, low-brightness areas. The white-hat transformation can enhance the particle targets while suppressing large background areas. The size of the structuring element can be selected according to the typical particle size of the iron filings, ensuring that iron filings within the particle size range are completely preserved.

[0027] The scrap metal region can be obtained by performing threshold segmentation on the white-hat transform result. One optional implementation is local adaptive threshold segmentation: a 15×15 window is set with each pixel in the image as the center. A local threshold is determined based on the gray-level statistics within the window. Pixels with gray levels higher than the local window threshold are extracted, and spatially connected pixels are aggregated into connected regions. All connected regions constitute the initial candidate set of suspected scrap metal. The reason for using a local window threshold instead of a global threshold is that the illumination intensity and background gray level differ at different locations of the transparent detection window. The local threshold can adapt to the imaging conditions of uneven illumination within the image, avoiding the global threshold's missed detection in dark areas and false detection in bright areas. The extracted scrap metal region will be used simultaneously for calculating the scrap metal density index at each detection position and analyzing the accumulation degree of the groove region, providing a unified data foundation for subsequent machine-by-machine loss localization and true loss correction.

[0028] Step S120: Determine the iron filings density index corresponding to each detection position based on the distribution characteristics of the iron filings region in the image corresponding to each detection position, and determine the trench accumulation change rate based on the change of the degree of iron filings accumulation in the trench image over time; wherein, the iron filings density index is used to characterize the distribution density of iron filings in the oil at the corresponding detection position.

[0029] The aforementioned iron filings density index is used to characterize the density of iron filings distribution in the oil at the corresponding detection location, and its target is the iron filings region extracted from the image of each detection location. ,in For the first A connected region, The total number of connected regions is the number of spatially interconnected aggregates of iron filings pixels after thresholding. The density of iron filings in oil contains two independent dimensions: the coverage area of ​​the iron filings in the image and the dispersion between the iron filings. A single dimension cannot fully characterize the density. For example, the density of the same number of particles clustered together is significantly different from that of uniformly scattered particles. Therefore, the iron filings density index is constructed from both the coverage area and the distribution spacing.

[0030] Coverage dimension based on area ratio Characterization, area ratio The method for determining the area ratio includes: calculating the pixel area of ​​each connected region, and recording the ratio of the sum of the pixel areas of all connected regions to the total pixel area of ​​the image as the area ratio. Area ratio The larger the value of , the wider the area covered by iron filings. Before filtration, the oil contains both flaky and granular iron filings, resulting in a larger connected area and a higher area ratio. The value is relatively large; the filtered oil contains fewer iron filings, the total area of ​​the connected region is small, and the area is smaller than that of the filtered oil. The value of decreases accordingly, so the area ratio can directly reflect the overall content level of iron filings in the oil.

[0031] Distribution spacing dimension with average spacing Characterization: Determine the center position of each connected region, calculate the Euclidean distance between the centers of any two connected regions, and take the arithmetic mean of all distances as the average spacing. A larger average spacing indicates a sparser distribution of iron filings, while a smaller average spacing indicates a denser distribution. Before filtration, there are many iron filings that are concentrated in one area, with a small average distance between their centers. After filtration, only a few scattered particles remain, which are dispersed and have a large average spacing. Therefore, the average spacing, from a geometric distribution perspective, supplements the information about particle aggregation and dispersion that the area ratio cannot reflect.

[0032] area ratio and average spacing Combined, record the density index of iron filings. Area ratio Positively correlated with density, average spacing The density index of iron filings is negatively correlated with the density of the distribution, and the division between the two results in a monotonically positive correlation. Before filtration, the iron filings cover a wide and dense area with a large area ratio and a small average spacing, resulting in a large iron filings density index. After filtration, the iron filings are sparse with a small area ratio and a large average spacing, resulting in a small iron filings density index. Therefore, the iron filings density indices corresponding to the upstream and downstream detection positions of the same machine can characterize the change in iron filings content in the oil before and after filtration with a unified dimension, providing a quantitative basis for horizontal comparison in determining the relative deterioration index of each machine. In actual implementation, the center of the connected region can be determined by the centroid or the center of the circumscribed rectangle, the area ratio can be replaced by the total number of pixels in the connected region, and the average spacing can be replaced by the nearest neighbor distance or the median distance.

[0033] The aforementioned trench buildup change rate refers to the change in trench buildup rate over time, used to characterize the severity of failure in the pre-filtration stage. The trench buildup rate reflects how quickly iron filings accumulate in the trench per unit time, while the trench buildup change rate further characterizes the acceleration or deceleration trend of the buildup rate. A higher value indicates a more pronounced acceleration in iron filings accumulation in the trench, and a more severe degree of failure in the pre-filtration stage.

[0034] Understandably, storage trench 208, located at the very end of the common cavity oil circuit, is the final convergence point for all escaped iron filings from the series-connected oil circuits of multiple machines. When the fine filters and strong magnetic adsorption devices of each machine are functioning normally, very few escaped iron filings are observed, and the iron filings in storage trench 208 accumulate at an approximately uniform and slow rate, with the trench accumulation change rate remaining near zero. However, when any pre-filtration stage fails, the number of escaped iron filings increases, and the iron filings in storage trench 208 accumulate rapidly in a short period of time, causing the trench accumulation change rate to rise accordingly. Therefore, the trench buildup change rate is equivalent to aggregating the failure information of the individual filtration stages of multiple machines into a global dynamic label. On the one hand, it can reflect the overall failure state of the multi-stage upstream protection stages in real time with a single parameter, providing a reliable triggering premise for machine-by-machine tracing. On the other hand, the total amount of iron filings in the common cavity oil circuit changes continuously over time. It is difficult to distinguish whether the increase in iron filings is due to the deterioration of the machine's protection or the increase in the global iron filings background based solely on the difference in iron filings density before and after filtration of a single machine. The trench buildup change rate provides a global reference benchmark for the iron filings density changes of each machine, enabling the change in the iron filings density index of each machine before and after filtration to be correlated with the global buildup dynamics. This lays the foundation for accurately determining the relative deterioration index of each machine and restoring the true wear and tear of each machine. The following introduces an optional determination scheme for the trench buildup change rate: Optionally, determining the trench accumulation change rate based on the change in the degree of iron filings accumulation in the trench image over time includes: determining the trench filling rate based on the coverage of the iron filings area on the storage trench 208 in the trench image; wherein, the trench filling rate is used to characterize the degree to which the storage trench 208 is covered by iron filings; when the trench filling rate indicates that the storage trench 208 is not completely covered by iron filings, the degree of iron filings accumulation is determined based on the trench filling rate; when the trench filling rate indicates that the storage trench 208 is completely covered by iron filings, the degree of iron filings accumulation is determined based on the grayscale difference between the accumulated protruding area and the surrounding area in the trench image; determining the trench accumulation rate based on the change in the degree of iron filings accumulation over time, and determining the trench accumulation change rate based on the change in the trench accumulation rate over time; wherein, the trench accumulation rate is used to characterize the speed at which iron filings accumulate in the storage trench 208 per unit time, and the trench accumulation change rate is used to characterize the overall failure severity of the filter device of each machine.

[0035] The storage trench 208 is located at the very end of the common cavity oil passage and serves as the final physical barrier for all the series-connected machines. Iron filings escaping from the fine filters and strong magnetic adsorption devices of each machine ultimately flow into the trench with the oil and deposit there. Therefore, the deposition state of the iron filings in the trench comprehensively reflects the overall working status of the multi-stage pre-existing protection system. The trench fill rate characterizes the degree to which the storage trench 208 is covered by iron filings. It is determined by: for the trench image, counting the total number of pixels with high brightness in the iron filings region of the image. and the total pixel area of ​​the image The trench filling rate is calculated according to the following formula. :

[0036] In the formula, This represents the total number of pixels highlighted within the iron filings region of the trench image; The total pixel area of ​​the groove image; The trench fill rate ranges from [0,1]. When the upstream filtration and adsorption stage is functioning properly, very few iron filings escape, and the sedimentation in the trench is slow, resulting in a high trench fill rate. The level remains low; when any preceding stage fails, the amount of escaped iron filings increases, the pixel area ratio of the iron filings region rises rapidly, and the trench fill rate decreases. As the number of chips increases, the groove filling rate can directly reflect the degree of planar coverage of the bottom surface of the groove by the iron filings.

[0037] The trench fill rate has an inherent upper limit; once the iron filings cover the bottom of the trench, the trench fill rate... The fill rate approaches 1 and saturates; thereafter, even if iron filings continue to accumulate, the trench fill rate... It can no longer reflect the increase in the amount of accumulation. As iron filings continue to accumulate after filling the trench, constrained by gravity and the trench boundaries, they will form a raised area at the center of the trench, higher than the edges. Under dark lighting, the top of the raised area exhibits high brightness due to its large curvature and strong scattering, while the brightness of the surrounding flat area is relatively low. The grayscale difference between the raised area and the surrounding area increases with the height of the raised area. Therefore, this grayscale difference can replace the fill rate to continue quantifying the degree of iron filings accumulation after the fill rate is saturated. An optional implementation method is to calculate the grayscale ratio of the inner and outer rings of the trench image. :

[0038] In the formula, This represents the average gray level of the piled-up protrusions in the trench image; The average gray level of the outer region; To prevent extremely small positive numbers with a denominator of zero, Units and The units are consistent. When the groove is not full, the iron filings are evenly spread, and the average gray level of the inner and outer areas is approximately equal. The gray level ratio between the inner and outer rings is... Approximately equal to 1; after the overflow protrusion is formed, the inner circle is brighter than the outer circle, and the grayscale ratio between the inner and outer circles is... It is greater than 1 and increases with the stacking height, thereby converting the three-dimensional stacking into gray-level differences in a two-dimensional image for quantification.

[0039] To ensure a continuous correlation between the quantitative results of the planar fill stage and the three-dimensional overflow stage, the product of the trench fill rate and the grayscale ratio of the inner and outer rings can be used as the degree of iron filings accumulation. ,Right now During the incomplete coverage phase, there was no difference in average grayscale between the inner and outer rings; however, the grayscale ratio between the inner and outer rings was... The value is 1, and the degree of accumulation is determined by the trench fill rate. Dominant; when the trench fill rate When the grayscale ratio of the inner and outer rings approaches 1 and the change slows down, As it begins to increase, the degree of accumulation shifts to the grayscale ratio between the inner and outer rings. Dominant, therefore the degree of iron filings accumulation It rises smoothly over time without any abrupt changes at the transition between two stages, achieving continuous quantization throughout the entire process from flat surface coverage to three-dimensional overflow.

[0040] After obtaining the continuous degree of iron filings accumulation, the trench accumulation rate can be determined according to the following formula. :

[0041] In the formula, For the current moment The degree of iron filings accumulation, The previous sampling time The degree of iron filings accumulation; The time interval between two consecutive samples is determined by the sampling frame rate; Characterization from Time's up The increment of trench buildup over time; The trench accumulation rate characterizes how quickly iron filings accumulate in the storage trench 208 per unit time. A higher value indicates faster accumulation and more severe failure of the pre-protection mechanism. Based on this, the trench accumulation change rate is determined according to the following formula. :

[0042] In the formula, The current trench buildup rate. The trench accumulation rate at the previous sampling time; The trench accumulation rate is used to characterize the severity of filter failure in each machine. Under normal circumstances, a small amount of escaped iron filings are uniformly deposited in the trench, and the trench accumulation rate is... The rate of change of trench packing increases at a constant speed. The value remains near zero; when a certain part of the filtration system fails, iron filings accumulate rapidly in a short period of time, which is reflected in the trench accumulation rate. The rate of increase is faster, and the rate of change of trench deposits is faster. The increase in trench accumulation rate can serve as a dynamic label reflecting the overall failure state of the multi-level protection system, providing a global reference benchmark for subsequent machine-by-machine loss location based on the chip density index of each detection point.

[0043] In the implementation method of quantifying the degree of iron filings accumulation during the overflow stage using the gray-scale difference between inner and outer regions, the delineation of the boundary between the accumulated protruding area and the outer region directly affects the accuracy of the gray-scale difference calculation. Therefore, this embodiment of the invention also provides an implementation method for determining the above boundary based on radial gray-scale changes, as follows: Optionally, the above-mentioned determination of the degree of iron filings accumulation based on the grayscale difference between the accumulated protruding area and the surrounding area in the trench image includes: taking the area with the highest grayscale in the trench image as the center, determining the position with the greatest grayscale change along the direction away from the center, determining the boundary between the accumulated protruding area and the surrounding area based on the position with the greatest grayscale change, and determining the degree of iron filings accumulation based on the difference between the grayscale condition of the accumulated protruding area inside the boundary and the grayscale condition of the surrounding area outside the boundary.

[0044] It is understandable that when iron filings continue to accumulate after covering the bottom of the trench, they form a raised area at the center of the trench, higher than the edges, due to gravity and the constraints of the trench boundaries. Under dark lighting, the top of the raised area exhibits a high-brightness spot due to its large curvature and strong scattering, while the flat area around the raised area has relatively low brightness. Furthermore, the grayscale change is most dramatic at the boundary between the raised edge and the flat area. Utilizing these imaging characteristics, the raised area and the surrounding area can be automatically divided from the trench image: the raised area refers to the high-grayscale area corresponding to the three-dimensional accumulation of iron filings in the trench image, while the surrounding area refers to the relatively low-grayscale area corresponding to the planar spread of iron filings in the trench image. The division is based on the radial variation of grayscale, rather than a pre-defined fixed geometric range. Therefore, the boundary position can adaptively adjust with changes in the accumulation morphology, unaffected by differences in trench size and accumulation contour.

[0045] One possible method for determining the boundary is to establish a polar coordinate system centered on the pixel with the highest grayscale value in the trench image, and let the grayscale value of a point in the polar coordinate system of the grayscale image be . ,in The radius distance from that point to the center of the circle. This is the polar angle at that point. Calculate the average grayscale value for each ring of pixels along the radial direction. That is, the same radius grayscale at each polar angle Take the average. Then calculate the average grayscale difference between two adjacent circles, and record the radius corresponding to the point where the grayscale difference is the largest as . ,in, The radius is the boundary between the bulging region and the surrounding region, i.e., the distance from the edge of the bulge to the center of the circle: from the center to... The area between them is a region of accumulated protrusions. The area between the image edge and the edge is the outer region. The reason it can serve as a boundary is that the grayscale level is higher and changes gradually within the raised area, while the grayscale level is lower and changes similarly gradually in the surrounding flat area. Only at the boundary between the edge of the raised area and the flat area does the grayscale level drop sharply, and the average grayscale difference between adjacent rings reaches its maximum at this location. The point with the highest grayscale value is chosen as the center because the top of the raised area has the strongest scattering and the highest brightness, and the grayscale peak naturally coincides with the geometric center of the raised area. Establishing a polar coordinate system with this center ensures that the radial direction aligns with the actual extension direction of the raised area. Using ring-by-ring averaging instead of pixel-by-pixel comparison can suppress the interference of individual noise points or individual reflective points on boundary positioning, making... The determined results are stable and reliable. In determining the boundaries... Then, from the center of the circle to The area between them is designated as the inner circle region, i.e., the area of ​​accumulated protrusions, with... The area between the image edge and the outer edge is defined as the outer region, i.e., the peripheral region. The average gray level of the inner region is calculated separately. Average gray level of the outer ring area Then, according to the grayscale ratio of the inner and outer circles Determine the degree of iron filings accumulation.

[0046] The degree of iron filings accumulation mentioned above can also be determined using one of the following methods: The first determination method is to directly use the total number of pixels in the iron filings region of the trench image or the ratio of the area of ​​the iron filings region to the area of ​​the trench image as the degree of iron filings accumulation. Since the field of view and installation position of the trench camera are fixed, the pixel area of ​​the iron filings region corresponds strictly to the actual coverage of iron filings in the trench. This ratio increases monotonically with the increase of deposition amount, which is suitable for quantification in the stage when the trench is not yet fully covered.

[0047] The second determination method is to perform grayscale statistics on the region of interest in the trench image, and use the proportion of pixels with grayscale values ​​higher than a set grayscale threshold or the average grayscale value of the region as the degree of iron filings accumulation. Under dark lighting, iron filings appear as bright targets. The more iron filings are deposited, the greater the proportion of high grayscale pixels and the higher the average grayscale value. The monotonic correspondence between the statistical quantity and the deposition amount is established.

[0048] The third approach is to use three-dimensional measurement to directly obtain the volume or height of the iron filings in the trench as the degree of iron filings accumulation. For example, a binocular camera can be placed above the trench for stereoscopic vision reconstruction, or structured light stripes can be projected into the trench and the surface height distribution can be calculated based on the deformation of the stripes on the accumulation surface. Both binocular parallax and structured light triangulation are mature three-dimensional reconstruction technologies that can accurately restore the geometric dimensions of the three-dimensional accumulation stage after the surface is filled.

[0049] In addition to the above-mentioned solutions, embodiments of the present invention may also employ one of the following solutions to determine the trench packing change rate: The first approach is to perform curve fitting on the time-series records of the degree of iron filings accumulation. For example, a polynomial or piecewise function can be used to fit the curve of the degree of accumulation over time. The second derivative of the fitted curve is then calculated. The second derivative directly corresponds to the acceleration of the degree of accumulation change, which is mathematically equivalent to the rate of change of trench accumulation. Furthermore, the fitting process has a smoothing effect on the noise of a single sampling.

[0050] The second approach is to use a sliding time window statistical method. The increment of the degree of iron filings accumulation is counted in two adjacent sliding windows. The difference between the increments of the two windows is divided by the window interval to obtain the rate of change of trench accumulation. The smoothing effect of the window can reduce the fluctuations caused by instantaneous noise.

[0051] The third determination scheme is based on the inter-frame difference of the image sequence. The inter-frame increment of the pixel area of ​​the iron filings region in the groove image of each adjacent frame is calculated to obtain the deposition rate at each time. Then, the difference between the deposition rates at adjacent time moments is calculated and compared with the time interval to directly reflect the acceleration trend of deposition.

[0052] The trench buildup change rate can not only quantify the severity of failure in the pre-filtration stage, but also serve as a basis for failure determination. Based on the trench buildup change rate, this invention determines the failure of the pre-filtration stage and initiates a step-by-step tracing process when a failure is detected, promptly pinpointing the failure location. The specific scheme is as follows: Optionally, the above-mentioned visual inspection method for iron filings in common cavity lubricating oil further includes: determining the failure of the pre-filtration stage based on the groove accumulation change rate; when the groove accumulation change rate indicates that the iron filings in the storage groove 208 are abnormally accelerated, it is determined that the pre-filtration stage has failed; when the pre-filtration stage is determined to have failed, the failure of the pre-filtration stage is traced back to its source machine by machine based on the relative deterioration index of each machine; wherein, the pre-filtration stage includes the filter device of each machine.

[0053] The pre-filtration stage refers to the various protective structures located upstream of the storage trench 208, used to intercept iron filings before the oil reaches the trench. This includes independently configured filtration devices for each machine, namely the fine filter and strong magnetic adsorption device for each machine. The fine filter is responsible for intercepting large iron filings such as flakes and granules, while the strong magnetic adsorption device is responsible for capturing small magnetic particles. Both are connected in series in the downstream oil passage of each machine. Failure of any of these stages will increase the amount of iron filings escaping. The storage trench 208 is located at the very end of the common cavity oil passage and is the final convergence point for iron filings escaping from all the connected machines. Therefore, any change in the operating state of any stage in the pre-filtration stage will ultimately be reflected in the deposition dynamics of iron filings within the trench. This is the physical basis for using the trench accumulation change rate to determine failure.

[0054] The working principle of failure assessment includes: Under normal conditions, a small amount of escaped iron filings settle evenly in the trench, the trench accumulation rate remains stable, and the trench accumulation change rate remains near zero. When the fine filter of a certain machine is damaged, large flaky iron filings escape, causing a sudden increase in flaky particles downstream of the machine's strong magnetic adsorption device. When the magnetic force of the strong magnetic adsorption device weakens, the number of small, dot-like particles downstream increases. These escaped iron filings eventually flow into the trench, causing the iron filings in the trench to settle more rapidly in a short period of time, manifested as an accelerated trench accumulation rate and an increased trench accumulation change rate. Therefore, the extent to which the trench accumulation change rate deviates from zero directly characterizes the severity of the failure in the pre-filtration stage. The process of its transition from a stable state near zero to a continuous increase can be used as a criterion for abnormally accelerated iron filings accumulation.

[0055] One possible determination method is to calculate the trench deposition rate in real time. When the rate of change of trench accumulation A value greater than zero indicates a rapid increase in iron filings deposited in the trench within a short period, suggesting a failure in the pre-filtration stage. Besides comparing to zero, the trench buildup rate can also be used as a reference. Compared with the preset rate of change threshold, the threshold can be set according to the fluctuation range of historical normal operation data to avoid misjudgment caused by fluctuations in operating conditions; it can also be used to statistically analyze the rate of change of trench deposits. The duration of multiple consecutive sampling periods greater than zero is used to determine the existence of a failure only when the abnormal acceleration continues for a set duration, in order to filter out transient interference.

[0056] The trench buildup change rate only indicates an overall failure in the pre-filter stage, not which machine the failure occurred in. Therefore, after determining a failure, machine-by-machine tracing is required. The tracing is based on the relative degradation index of each machine: the relative degradation index characterizes the degree of deterioration in the protective capability of the corresponding machine's filter; the higher the value, the more severe the failure. When a failure in the pre-filter stage is determined, the relative degradation indices of each machine are compared. The machine with the highest relative degradation index is identified as the failed machine, meaning the failed pre-filter stage is the filter device of that machine. Maintenance personnel can then directly pinpoint the machine and component requiring repair, avoiding the need to disassemble and inspect all connected machines individually, thus achieving a seamless transition from overall failure detection to individual machine failure location.

[0057] Step S130: For each machine, determine the relative deterioration index of each machine based on the change between the iron filings density index corresponding to the upstream and downstream detection positions and the trench accumulation change rate; wherein, the relative deterioration index is used to characterize the degree of deterioration of the protective capability of the corresponding machine's filter device.

[0058] The aforementioned relative deterioration index is a quantitative indicator used to characterize the degree of deterioration in the protective capability of the corresponding machine's filter device. The larger the value, the more significant the decrease in the interception capability of the corresponding machine's fine filter and strong magnetic adsorption device, and the more severe the degree of wear and tear. When the value is at a low level, it indicates that the corresponding machine's protective capability is normal.

[0059] Understandably, the trench buildup change rate can only reflect the overall failure state of the multi-stage front-end protection system. It indicates the existence of a failure but cannot indicate which machine the failure occurred in. Since multiple machines connected in series in the common cavity oil circuit are each equipped with an independent filtration device, maintenance and repair must be carried out on a single machine-by-machine basis. Therefore, a loss metric that is granular at the single machine level, has uniform dimensions, and can be compared horizontally is needed to realize the overall failure perception as machine-by-machine location results.

[0060] The relative deterioration index combines the change in the metal chip density index before and after filtration for each machine with the global dynamic represented by the groove accumulation change rate. The total amount of metal chips in the common cavity oil circuit fluctuates over time. Relying solely on the difference in metal chip density before and after filtration for a single machine makes it impossible to distinguish whether the increase in metal chips is due to the deterioration of the machine's protection or an increase in the overall metal chip background. However, by introducing the groove accumulation change rate as a global reference, the metal chip density change of each machine can be correlated with the global accumulation dynamic, allowing normal and deteriorated machines to exhibit distinguishable values ​​under a unified scale. Furthermore, the relative deterioration index is obtained through relative comparison between machines, eliminating individual differences in gear wear rate, oil flow rate, and installation location, as well as the interference of global fluctuations in the common cavity oil circuit on loss assessment. This allows for direct horizontal comparison of the index values ​​of each machine, providing a quantitative basis for accurately locating deteriorated machines and providing a unified data foundation for subsequent masking judgment and actual loss correction based on the comparison relationship of adjacent machine indices, as well as for graded alarms and maintenance decisions based on the correction results. The following describes an optional scheme for determining the relative deterioration index: Optionally, determining the relative deterioration index of each machine based on the changes in the scrap density index between the upstream and downstream detection positions and the trench accumulation change rate includes: determining the contribution of each machine to the scrap accumulation in the storage trench 208 based on the changes in the scrap density index between the upstream and downstream detection positions and the trench accumulation change rate; wherein the contribution level is used to characterize the contribution of the scrap change of the corresponding machine to the scrap accumulation change in the storage trench 208; determining the machine failure rate of each machine based on the change of the contribution level of each machine over time; wherein the machine failure rate is used to characterize the rate of deterioration of the protective capability of the filter device of the corresponding machine over time; and comparing the machine failure rate of each machine with the machine failure rate of the benchmark machine, using the machine with the smallest absolute value of the machine failure rate among all machines as the benchmark, to obtain the relative deterioration index of each machine.

[0061] In a common-cavity oil circuit, multiple machines are connected in series. Oil flows sequentially through each machine and its filtration device before converging into the final storage trench 208. Under normal circumstances, each machine's fine filter intercepts large particles, and its strong magnetic adsorption device captures fine magnetic debris, resulting in very little iron filings entering the trench. When the filtration device of a particular machine deteriorates, its interception capacity decreases, the difference in iron filings content before and after filtration narrows, and the escape of iron filings increases, accelerating trench accumulation. The contribution level is a parameter that links the changes in individual machines to the overall dynamics, characterizing the contribution of the corresponding machine's iron filings change to the iron filings accumulation change in the storage trench 208: if the iron filings change of a particular machine can explain the change in trench accumulation, it indicates that the machine is functioning normally, and the two maintain a stable proportional relationship. If trench accumulation accelerates while the interception rate of a particular machine does not change synchronously, it indicates that the contribution relationship of that machine to trench accumulation has been disrupted, and that machine is the source of deterioration.

[0062] One possible scheme for determining the aforementioned level of contribution includes: [denotes the first...] The upstream and downstream detection positions of the machine are the starting detection positions. and last detection bit Corresponding time The density indices of the iron filings are respectively and ,according to Calculate the first The rate of change of total chip density from the initial detection position to the final detection position of the machine is given by the formula, where, and They are time points Next start detection bit and last detection bit The corresponding iron filings density index; For the first The rate of change of total iron chip density of the machine from the initial detection position to the last detection position; This represents the time interval between two adjacent samples.

[0063] Then according to Calculate the degree of contribution, where, The rate of change of trench deposits; To the degree of contribution; The preset baseline trench buildup change rate can be calibrated based on historical data under normal operating conditions, for example, by... The value is set to 1. Furthermore, Units and The units are consistent. This is understandable, given the degree of contribution. The physical meaning of this can be understood as: the rate of change in machine scrap density corresponding to each unit of groove packing change rate, that is, the contribution intensity of single-machine scrap change to the global groove packing change. Under normal circumstances, the difference in scrap density index before and after filtration for each machine is relatively stable, and the groove packing change rate remains near zero. It remains stable within a small, constant range; when a machine deteriorates, its interception capability decreases, leading to... Significantly reduced, the first Total chip density change rate of the machine from the initial detection position to the final detection position The rate of change in trench accumulation decreases due to the increase in escaped iron filings. Ascend, the machine's A sharp drop occurs; for downstream machines of deteriorating machines, the amount of incoming iron filings increases proportionally. Basically unchanged, for upstream machines, It may only decrease slightly. Therefore, the performance of each machine... It exhibits different change patterns in time sequence, and monitoring The real-time changes can identify which machine is experiencing degradation.

[0064] The static value of the contribution level is affected by individual machine differences. To characterize the degradation rate of the protection capability of each machine, the contribution level can be time-series differencing, according to... Calculate the first Machine failure rate of the machine ,in, For the current moment The degree of contribution, For the next sampling time The degree of contribution, Indicates the first The decrease in the contribution of each machine from the current moment to the next moment. Machine failure rate. The rate at which the protective capability of the corresponding machine's filter deteriorates over time: under normal circumstances, the contribution is equal, and the machine failure rate... The value is near zero; the higher the machine failure rate, the faster the protection capability weakens and the more severe the failure.

[0065] Different machines exhibit individual differences in gear wear rates, oil flow rates, etc. Directly comparing machine failure rates across different machines can be affected by system background interference. Therefore, a relative comparison mechanism is introduced: the machine failure rates of all machines are recorded in real time at the same moment, and the machine with the smallest absolute value of its failure rate at that moment is identified as the benchmark. This benchmark machine is considered the most stable at that moment, and its absolute failure rate is recorded as [value missing]. Then according to Calculate the first The relative deterioration index of the machines, among which, To prevent extremely small positive numbers with a denominator of zero, Units and The units are consistent. Taking the absolute value ensures that the denominator remains positive even when the failure rate of the most stable machine is negative (i.e., its contribution increases slightly). Comparing the failure rates with the machine having the smallest absolute value is equivalent to using the healthiest machine in the system as a reference point. The degradation rate of each machine is measured relative to this reference point, thus eliminating the interference of global fluctuations and individual differences in the common cavity oil circuit. The relative deterioration index of the machine It can be further normalized to the (0,1) interval using the Sigmoid function. The larger the value, the greater the degree of wear and tear on the machine. This allows for horizontal comparisons on the same scale, providing a quantitative basis for locating deteriorated machines and subsequent cover-up and correction.

[0066] In addition to the above-mentioned schemes, the relative deterioration index can also be determined using one of the following schemes: The first approach is to perform a sliding window trend analysis on the contribution of each machine over time, using the rate of decline of the contribution within the window as the rate of degradation. The greater the rate of decline, the greater the relative degradation index. The sliding window has a smoothing effect on the noise of a single sampling.

[0067] The second approach is to use statistical standardization to calculate the mean and standard deviation of the machine failure rate of all machines at the same time. The standardized deviation of the machine failure rate of each machine from the mean is used as the relative deterioration index. The mean reflects the overall background of the system, and the deviation reflects the relative deterioration of a single machine.

[0068] Step S140: Based on the comparison between the relative deterioration index of the machine and the upstream adjacent machine and the distance between the machines, the relative deterioration index is corrected to obtain the corrected deterioration index; wherein, the corrected deterioration index is used to characterize the actual wear and tear of the machine's filtration device.

[0069] Understandably, the reason for this correction is that the series structure of the common cavity oil circuit can mask the true losses of downstream machines from those of upstream machines. In the common cavity oil circuit, oil flows sequentially through each machine. If the protective mechanism of any upstream machine fails, the resulting escaped metal filings will enter the downstream machine with the oil. When an upstream machine is severely damaged, the amount of metal filings at its outlet increases significantly, abnormally amplifying the base amount of metal filings at the inlet of the downstream machine. Even if the downstream machine's own protective capabilities have clearly decreased, the increased base amount of inlet metal filings may still result in a high level of intercepted metal filings, and the apparent data will not show significant changes. The relative deterioration index will thus remain low, masking the true losses of the downstream machine. If fault location and maintenance decisions are made directly based on the uncorrected relative deterioration index, deteriorated machines masked by upstream defects will be misjudged as normal machines, leading to missed diagnoses. This allows metal filings to continue spreading to further downstream machines and motor cavities, resulting in a loss of early warning opportunities. The comparison of relative deterioration indices and the distance between machines provide the information needed to reconstruct the true wear and tear. The comparison of relative deterioration indices between adjacent machines reflects the degree to which the index value of downstream machines is suppressed by upstream-transmitted iron filings. The distance between machines reflects the transmission and attenuation law of upstream failure effects along the oil path; the closer the upstream machine, the stronger its masking effect on the downstream. Therefore, by combining the comparison of relative deterioration indices between adjacent machines with the distance between machines to correct the relative deterioration index, the part of the apparent data masked by the upstream transmission effect can be restored, resulting in a corrected deterioration index that characterizes the true wear and tear of the filter devices of each machine. This allows for the accurate identification of masked deteriorated machines, ensuring that subsequent fault location, graded alarms, and maintenance decisions are based on the true wear and tear status of each machine.

[0070] Optionally, the above-mentioned method of correcting the relative deterioration index based on the comparison between the machine and its upstream neighboring machine and the distance between the machines to obtain a corrected deterioration index includes: determining whether the machine is covered by the upstream neighboring machine based on the comparison between the machine and its upstream neighboring machine; wherein, when the comparison indicates that the relative deterioration index of the machine is abnormally low relative to the relative deterioration index of the upstream neighboring machine, it is determined that the machine is covered by the upstream neighboring machine; for the machine covered by the upstream neighboring machine, the relative deterioration index of the machine is corrected based on the distance between the machine and the upstream neighboring machine to obtain the corrected deterioration index of the machine.

[0071] When metal scrap from an upstream machine is carried to a downstream machine by the oil, the initial metal scrap quantity at the downstream machine's inlet is abnormally amplified. Even if the downstream machine's own protection capabilities have decreased, its apparent interception rate may still remain high, and its relative degradation index is suppressed. This phenomenon is called masking. The task of masking determination is to identify machines whose relative degradation indices are suppressed by the upstream transmission effect from among the various machines. The basis for this is that if a machine is masked by an upstream adjacent machine, its relative degradation index will be disproportionately low relative to the upstream adjacent machine, while the relative degradation indices of two normally functioning adjacent machines should be at similar levels. Therefore, the comparison between the relative degradation indices of two adjacent machines constitutes the direct basis for masking determination.

[0072] One possible approach to masking determination is to define the first... The machine is relative to its upstream first The degree of concealment of the machine ,in, For the first The machine is at all times The relative deterioration index, For downstream The relative deterioration index of the machines; To prevent extremely small positive numbers with a denominator of zero, Units and The units are the same; For the first The machine is relative to its upstream first The degree to which the machine was concealed. When When the value is close to 1, it indicates that the relative deterioration index of two adjacent machines has similar values, and the two machines are in the same state, both in normal condition; when When the value of is close to 0, it indicates that the relative deterioration index of the downstream machine has dropped sharply relative to the upstream machine. That is, the index value of the downstream machine is abnormally suppressed, the proportion of iron filings at the inlet of the downstream machine comes from the upstream, and the machine's own iron filings production and actual losses are easily masked by the upstream. At this time, the first _____ is determined to be _____. The downstream machine is obscured by the upstream adjacent machine. In addition to calculating the ratio, the difference or decrease in the relative deterioration index of the adjacent machines can also be compared with the preset criteria, as long as it can identify the situation where the index value of the downstream machine is abnormally low relative to the upstream.

[0073] Based on the masking determination, the correction operation is only performed on the machines identified as masked. Machines not masked are not corrected, and their relative deterioration index is directly output as the correction deterioration index. The reason for selective correction is that if all downstream machines are indiscriminately corrected when upstream abnormal signals propagate downstream, the index values ​​of normal machines will be unnecessarily inflated, leading to false corrections and causing normal machines to be falsely reported as damaged machines. Limiting the correction to masked machines allows for precise focus on the location of the masking, preventing excessive propagation of upstream abnormal signals downstream. When correcting masked machines, the correction magnitude is related to the distance between the machines: iron filings generated by upstream failures are continuously intercepted by filters along the oil flow path, and their masking effect on downstream machines gradually decreases along the oil path. The closer the distance between upstream and downstream machines, the stronger the masking effect and the larger the required correction magnitude; the farther the distance, the weaker the masking effect and the smaller the correction magnitude. The relative deterioration index of the masked machine is adjusted by increasing the distance-related correction range described above, restoring the suppressed index value to a level consistent with the actual wear and tear. The result is the corrected deterioration index. The corrected deterioration index characterizes the true wear and tear of the machine's filter system. It eliminates the interference of upstream-transmitted iron filings on the apparent data, ensuring that the index value of each machine, regardless of its position in the common cavity oil circuit, accurately reflects the actual deterioration state of its filter system. This provides a reliable basis for subsequent graded alarm and maintenance decisions based on actual wear and tear.

[0074] The correction magnitude for the masked machine needs to comprehensively consider the transmission and accumulation of the failure effects of each upstream machine. However, the influence intensity of upstream machines decreases with distance, and simple adjacent comparisons are insufficient to fully characterize the cumulative effect. Therefore, this invention also provides a correction implementation method based on distance weighting and cumulative influence, as detailed below: Optionally, the above-mentioned correction of the relative deterioration index of the machine includes: determining the weight of the damage impact of each upstream machine on the machine based on the distance between the machine and each upstream machine; wherein, the closer the distance between the upstream machines, the greater the weight of the damage impact; determining the damage impact factor of the machine based on the damage impact weight of each upstream machine and the masking situation of each upstream machine; wherein, the damage impact factor is used to characterize the cumulative impact of the failure of each upstream machine on the machine; and correcting the relative deterioration index of the machine based on the damage impact factor to obtain the corrected deterioration index of the machine.

[0075] In a common-cavity oil circuit, the masking effect on a machine is not solely from the adjacent upstream machine, but rather a result of the cumulative effect of failures in each upstream machine propagating along the oil circuit. The intensity of the impact of each upstream machine on the same downstream machine varies: the more severe the failure of an upstream machine, the more metal filings escape, and the stronger the masking effect on the downstream machine; the closer the distance between the upstream and downstream machines, the smaller the proportion of escaped metal filings intercepted by the filtering device during transmission, resulting in more metal filings reaching the downstream machine and a stronger impact; the greater the distance, the less the impact diminishes along the oil circuit. The loss impact weight characterizes the differentiated impact intensity of each upstream machine on the downstream machine due to distance differences, while the loss impact factor characterizes the cumulative impact of upstream machine failures on the downstream machine. Both together correlate the correction magnitude of the masked machine with the spatial structure of the oil circuit.

[0076] One possible scheme for determining the weighting of loss effects includes: using the serial numbering of the machines in the common cavity oil circuit to represent the positional relationship between the machines, and denoting the first... The machine number on the platform is The revised first The machine number is The numbers increase sequentially along the oil flow direction, and the difference in numbers corresponds to the distance of the oil passage between machines. As the first Upstream machines on the platform Distance weights for each machine. The closer the number of the upstream machine is to the first The closer the distance between the machine numbers, the better. The larger the value, the greater the distance weight, which conforms to the transmission law that the influence of the upstream decreases with distance. In addition to the numbering ratio, the reciprocal of the numbering difference or the attenuation coefficient calibrated according to the actual path length of the oil pipeline can also be used to determine the distance weight, as long as the monotonic relationship that the closer the distance, the greater the weight is is satisfied.

[0077] Based on distance weighting, the loss impact factor is determined by considering the masking situation of each upstream machine. One possible method for determination is: for the th For each machine, calculate the loss impact factor using the following formula:

[0078] In the formula, For the first The degree of cover of a machine relative to its upstream neighboring machine, when The first machine had no upstream machine. Take 1; To conceal the inverse form of degree, The closer the value is to 0, the more severe the damage to the upstream machine at the corresponding location, and the more urgent the need for correction at that location. To prevent extremely small positive numbers with a denominator of zero, Units and The units are the same; For the first The loss impact factor of each machine. Summation is performed by iterating through machines 1 to 12. The severity of failures of each upstream machine is inversely amplified by the degree of masking, and then weighted and accumulated according to distance to fully characterize the impact of all upstream failures on the first machine. The cumulative impact of the machines.

[0079] In practical engineering, the masking effect of upstream machine damage on downstream machines is mainly reflected in the machine immediately following the damaged one, i.e., the location where the masking degree changes abruptly. If the absolute value of the loss impact factor is used for correction directly, the large weight generated by severe upstream damage may propagate excessively downstream along the summation chain, causing machines far from the damage location to be over-corrected. Therefore, the ratio of the loss impact factors of two adjacent machines can be used as a correction coefficient, and the following formula can be applied to the... The relative deterioration index of the machines was adjusted:

[0080] In the formula, For the first The relative deterioration index of the machines; For the first The wear and tear factors of the machine; To correct the deterioration index. When When the value is approximately 1, it indicates that the cumulative upstream impact on two adjacent machines is similar. No new masking changes occurred at the machine location; only minor adjustments are needed. When it is much greater than 1, it indicates that the first... The cumulative impact at a certain machine experiences a new jump relative to its upstream location, indicating severe masking at this point, requiring strong correction. Using the ratio of the impact coefficients of adjacent machines as the correction factor, the correction magnitude depends only on the relative increment of the cumulative impact at the jump location, rather than the absolute value of the cumulative impact. Therefore, the correction is precisely focused on the location of the masking jump, avoiding erroneous corrections caused by excessive propagation of upstream abnormal signals downstream. This ensures that the correction degradation index accurately reflects the true wear and tear status of each machine's filter device.

[0081] The corrected degradation index accurately reflects the actual wear and tear of the filter devices in each machine. Only by translating the wear assessment results into actionable maintenance actions can the testing value be truly realized. Therefore, this embodiment of the invention, based on the corrected degradation index, classifies each machine to match differentiated maintenance strategies. The specific scheme is as follows: Optionally, the above-mentioned visual detection method for metal chips in common cavity lubricating oil further includes: classifying each machine according to its correction deterioration index; wherein the classification judgment includes at least: when the correction deterioration index of a machine meets a first judgment condition, determining that the machine is in a normal state; when the correction deterioration index of a machine meets a second judgment condition, outputting a planned maintenance prompt for the machine; wherein the degree of wear corresponding to the second judgment condition is higher than the degree of wear corresponding to the first judgment condition; when the correction deterioration index of a machine meets a third judgment condition, outputting a shutdown and maintenance prompt for the machine; wherein the degree of wear corresponding to the third judgment condition is higher than the degree of wear corresponding to the second judgment condition.

[0082] The modified deterioration index reflects the actual wear and tear of the filtration devices on each machine in a continuous, quantitative manner. However, the investment of maintenance resources is discrete, and different treatment strategies should be matched to machines with different levels of wear: if all machines with minor wear are shut down for maintenance, it will waste maintenance resources and cause unnecessary production interruptions; if machines with severe wear are only subjected to routine inspections, the opportunity for intervention will be missed, allowing iron filings to enter the motor winding area. The purpose of the classification is to establish a mapping relationship between the modified deterioration index and maintenance strategies, converting continuous wear measurements into differentiated treatment instructions, so that the maintenance intensity matches the actual wear level.

[0083] The system sets progressively increasing levels of wear and tear for the first, second, and third judgment conditions. When the machine's correction deterioration index meets the first condition (lowest level of wear), it indicates that the machine's filter protection capability is normal, and the machine is considered to be in normal condition, with no output of any warning. When the correction deterioration index meets the second condition (moderate level of wear), it indicates that the machine's protection capability has deteriorated but has not yet failed severely, and a planned maintenance warning is output, suggesting that the machine be included in the maintenance plan and scheduled for repair. When the correction deterioration index meets the third condition (highest level of wear), it indicates that the machine's filter has failed severely, and a shutdown warning is output, requiring immediate shutdown for repair. The distinction between planned maintenance warnings and shutdown warnings allows for tiered timing and urgency of maintenance intervention: the former allows for coordinated maintenance without shutting down the machine, while the latter requires immediate response, thus achieving a balance between maintenance costs and failure risks.

[0084] Set alarm threshold and ,and , and The value of corresponds to the range of values ​​for the modified deterioration index. When the modified deterioration index... When, corresponding to the first judgment condition, the machine judges it as normal; when When, corresponding to the second judgment condition, a level one warning is triggered; when When the third judgment condition is met, a level two alarm is triggered. The threshold value can be calibrated based on the operating conditions, historical wear data, and maintenance cycle of each machine, or it can be set separately for different types of machines to adapt to the differences in wear characteristics of machines in a series oil circuit. In addition to comparing two thresholds, the value range of the corrected deterioration index can be divided into multiple numerical intervals, each interval corresponding to a judgment level, or a comprehensive judgment can be made by combining the duration of the corrected deterioration index. For example, an alert can be output only when the corrected deterioration index meets the corresponding judgment condition for multiple consecutive sampling cycles to suppress false alarms caused by instantaneous fluctuations. The output of the graded judgment includes the machine number and the corresponding wear level information, which can be directly used to guide maintenance personnel to determine the maintenance object and maintenance time, realizing the transformation from actual wear assessment to maintenance decision.

[0085] Optionally, the above-mentioned visual detection method for iron filings in common cavity lubricating oil further includes: for machines that output a shutdown maintenance prompt, controlling the solenoid valve in the corresponding lubricating oil circulation circuit of the machine to cut off the oil flow in the lubricating oil circulation circuit to block the spread of iron filings to downstream machines; generating a maintenance work order for the machine; wherein the maintenance work order includes at least the parts that need to be replaced on the machine and the recommended maintenance window time, and the parts that need to be replaced include at least one of a fine filter element and a strong magnetic adsorption device.

[0086] The shutdown and maintenance prompt only completes the output of the maintenance decision. Before the maintenance personnel actually arrive and handle the situation, the failed machine continues to operate along the common cavity oil circuit. Its filter device has been severely damaged, and the iron filings generated by machine wear are no longer effectively intercepted. They will continue to spread with the oil to downstream machines and the motor cavity, causing the hazard to continue to expand during the interval between maintenance. Executing the oil circuit cut-off for the machine that outputs the shutdown and maintenance prompt is an immediate intervention measure to stop the spread of hazard before maintenance is completed: Each machine has its own branch in the lubricating oil circulation loop, and each branch is equipped with a solenoid valve. The opening and closing of the solenoid valve determines the oil flow in that branch. By controlling the closing of the solenoid valve corresponding to the failed machine's branch, the oil flow in that branch is cut off. The iron filings generated by the failed machine cannot enter the downstream machine with the oil, and the stator, rotor, and enameled wire in the motor cavity are no longer exposed to the threat of newly added iron filings.

[0087] One optional control method includes: after receiving a shutdown and maintenance prompt, the maintenance management platform sends a control command to the programmable logic controller (PLC). The PLC locks the solenoid valve in the lubricating oil circulation loop corresponding to the failed machine, keeping the solenoid valve closed and cutting off the oil flow in that branch. Using a locking method instead of an instantaneous closing action prevents the valve from resetting due to disturbance or misoperation before the intervention is completed, ensuring the continuity of the blocked state. Other branches of the common cavity oil circuit remain unaffected, and the oil circulation of normal machines is maintained, thus limiting the intervention scope to the smallest oil circuit unit corresponding to the failed machine and avoiding collateral losses caused by a complete machine shutdown. The timing of the cut-off is linked to the grading judgment result, and is only executed on machines that trigger a shutdown and maintenance prompt; machines that trigger a planned maintenance prompt are not cut off, ensuring that the intervention intensity matches the degree of wear and tear.

[0088] While performing oil circuit intervention, maintenance work orders are generated for the failed machines, transforming the detection and evaluation results into executable maintenance tasks. The maintenance work order includes at least the parts that need to be replaced and the recommended maintenance window: Replaceable parts include at least one of the fine filter element and the strong magnetic adsorption device. Replace the filter element if the fine filter is damaged, and replace the strong magnetic adsorption device if its magnetic force weakens. If both fail, replace them together. The part information marked on the work order allows maintenance personnel to prepare spare parts in advance, avoiding prolonged downtime caused by on-site disassembly and relocation. The recommended maintenance window is a maintenance timing suggestion based on the machine's wear level and production plan, ensuring maintenance is scheduled within a time window that minimizes production impact. Furthermore, alarm signals, the failed machine's number, and the correction deterioration index can be sent to the maintenance management platform via industrial Ethernet. The maintenance management platform generates maintenance work orders based on this information and updates the status of each machine in real time on the local industrial screen, displaying real-time images and iron filings concentration trend curves for each detection point, providing on-site personnel with an intuitive monitoring method. The combination of oil circuit disconnection and maintenance work orders creates a complete closed loop for each machine in the common cavity oil circuit, from iron filings detection, failure location, actual loss assessment, graded early warning to maintenance intervention, fundamentally preventing iron filings from entering the motor winding area and causing insulation damage.

[0089] like Figure 3 As shown, based on the same inventive concept, this embodiment of the invention also provides a visual inspection system 300 for lubricating oil chips in a common cavity, including an image acquisition unit 310 and a data processing unit 320, wherein: The image acquisition unit 310 is used to acquire oil images at the upstream and downstream detection positions of the filter device of each of the multiple machines connected in series in the common cavity oil circuit, and to acquire trench images at the storage trench 208 at the end of the common cavity oil circuit. The data processing unit 320 is used to extract the iron filings region from the oil image and the trench image; determine the iron filings density index corresponding to each detection position based on the distribution characteristics of the iron filings region in the image corresponding to each detection position, and determine the trench accumulation change rate based on the change of the iron filings accumulation degree in the trench image over time; wherein, the iron filings density index is used to characterize the distribution density of iron filings in the oil at the corresponding detection position; for each machine, the relative deterioration index of each machine is determined based on the change between the iron filings density index corresponding to the upstream and downstream detection positions and the trench accumulation change rate; wherein, the relative deterioration index is used to characterize the degree of deterioration of the protective capability of the filter device of the corresponding machine; the relative deterioration index is corrected based on the comparison relationship between the relative deterioration index of each machine and the upstream adjacent machine and the distance between the machines to obtain a corrected deterioration index; wherein, the corrected deterioration index is used to characterize the actual wear and tear of the filter device of the corresponding machine.

[0090] The image acquisition unit 310 mentioned above is responsible for acquiring raw image data. Its physical implementation can refer to the arrangement of the aforementioned detection positions: a transparent detection window and a dark-field illumination industrial camera are set upstream of the inlet pipe of each machine's fine filter to acquire images of the oil before it enters the filtration device; a detection camera is set downstream of the outlet pipe of each machine's strong magnetic adsorption device to acquire images of the oil after filtration and adsorption; and a separate camera is set directly above the storage trench 208 of the stator oil return chamber 207 at the very end of the machine to acquire trench images. Under dark-field illumination, iron filings appear as bright particles while the oil appears as a dark background. The high contrast between the particles and the background provides a good imaging basis for subsequent image processing. Each camera continuously acquires images at a set sampling frame rate, forming a time-continuous image sequence.

[0091] The data processing unit 320 is responsible for image processing and loss degree calculation. It is communicatively connected to the image acquisition unit 310. After receiving oil and trench images uploaded from each detection position, it sequentially performs processes such as scrap region extraction, scrap density index determination, trench accumulation change rate determination, relative deterioration index determination, and masking correction. First, scrap regions are extracted from various images. Then, based on the distribution characteristics of scrap regions in each detection position image, the scrap density index corresponding to each detection position is determined. The trench accumulation change rate is determined based on the change in scrap accumulation degree over time in the trench image. Subsequently, for each machine, the change in scrap density index between the upstream and downstream detection positions is combined with the trench accumulation change rate to determine the relative deterioration index for each machine. Finally, based on the comparison between the relative deterioration indices of each machine and its upstream neighboring machine, as well as the distance between the machines, the relative deterioration index is corrected to obtain a corrected deterioration index characterizing the actual loss of the filter device in each machine. The implementation methods of the above processes are consistent with the corresponding steps in the aforementioned method embodiments, and will not be repeated here. All optional solutions in the aforementioned method embodiments can be executed by the data processing unit 320.

[0092] The hardware form of the aforementioned data processing unit 320 can be varied, such as an industrial computer or edge computing device deployed on-site, or a remote server communicating with the maintenance management platform, as long as it has image reception and parallel processing capabilities. After obtaining the corrected deterioration index, the data processing unit 320 can further perform a graded judgment: comparing the corrected deterioration index of each machine with the judgment conditions, and outputting normal, planned maintenance prompts, or shutdown inspection prompts for machines at different wear levels; and sending alarm signals, the number of the failed machine, and the corrected deterioration index to the maintenance management platform via industrial Ethernet, while simultaneously updating the status of each machine in real time on the local industrial screen, displaying real-time images and iron filings concentration trend curves for each detection position. For machines that trigger a shutdown inspection prompt, the maintenance management platform can lock the solenoid valve in the corresponding lubricating oil circulation loop of the machine through a programmable logic controller, cutting off the oil flow in that branch to prevent iron filings from spreading to downstream machines and the motor cavity, and generating a maintenance work order indicating the parts to be replaced and the recommended maintenance window time. Therefore, the common cavity lubricating oil metal chip visual inspection system 300, with the image acquisition unit 310 and the data processing unit 320 as its core, together with the maintenance management platform and the oil circuit actuator, realizes a complete closed-loop management of each machine in the common cavity oil circuit from metal chip detection, failure location, actual loss assessment, graded early warning to maintenance intervention.

[0093] This invention is now complete.

[0094] In summary, in this embodiment of the invention, for multiple machines connected in series in a common cavity oil circuit, oil images are acquired at the upstream and downstream detection positions of the filter device of each machine, and a trench image is acquired at the storage trench 208 at the end of the common cavity oil circuit. Iron filings regions are extracted from the oil images and trench images. The iron filings density index corresponding to each detection position is determined based on the distribution characteristics of the iron filings region in the images corresponding to each detection position, and the trench accumulation change rate is determined based on the change in the degree of iron filings accumulation in the trench image over time. The iron filings density index is used to characterize the density of iron filings distribution in the oil at the corresponding detection position. For each machine, a relative deterioration index is determined based on the change between the iron filings density index corresponding to the upstream and downstream detection positions and the trench accumulation change rate. The relative deterioration index is used to characterize the degree of deterioration in the protective capability of the filter device of the corresponding machine. The relative deterioration index is corrected based on the comparison between the relative deterioration indices of the machine and its upstream adjacent machine, as well as the distance between the machines, to obtain a corrected deterioration index. The corrected deterioration index is used to characterize the actual wear and tear of the machine's filter device.

[0095] This invention acquires oil images at the upstream and downstream detection positions of each machine's filter device, and trench images at the storage trench 208 at the end of the common cavity oil circuit. Based on the distribution characteristics of the iron filings area at each detection position, it determines the iron filings density index, and combines this with the trench accumulation change rate to determine the relative deterioration index of each machine. Further corrections are made based on the comparison of the relative deterioration indices of adjacent machines and the distance between machines. This eliminates the masking effect of upstream iron filings conduction on downstream machines in the series common cavity oil circuit, achieving an accurate assessment of the actual wear and tear of each machine's filter device and providing a reliable basis for maintenance decisions. On the other hand, by combining the changes in the iron filings density index between the upstream and downstream detection positions of each machine with the trench accumulation change rate to determine the contribution level, and obtaining the machine failure rate based on the change in contribution level over time, a comparison is made using the machine with the smallest absolute failure rate as a benchmark. This eliminates the interference of global fluctuations in the common cavity oil circuit and individual differences in each machine on wear assessment. The system effectively addresses various challenges. Firstly, it accurately locates machines with deteriorated protective capabilities. Secondly, it quantifies the degree of iron filings accumulation based on the filling rate when the storage trench 208 is not full, and quantifies the degree of iron filings accumulation based on the grayscale difference between the raised accumulation area and the surrounding area after the trench is full. This constructs a continuous quantitative index from planar fullness to three-dimensional overflow. Furthermore, it determines the failure of the pre-filter based on the trench accumulation change rate, enabling real-time perception and machine-by-machine traceability of the failure status of multi-level pre-filters. Thirdly, it determines whether downstream machines are covered by comparing the relative deterioration indices of adjacent machines. Only covered machines are corrected according to the distance-related loss impact weight, avoiding false corrections caused by excessive propagation of upstream abnormal signals downstream. Combined with graded alarms based on the corrected deterioration index and the cut-off control of the oil circuit of the failed machine, it blocks the diffusion of iron filings to downstream machines and motor cavities, forming a complete closed loop from detection, location, early warning to maintenance intervention.

[0096] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A visual inspection method for iron filings in a common cavity lubricating oil, characterized in that, The method includes: For multiple machines connected in series in the common cavity oil circuit, oil images are acquired at the upstream and downstream detection positions of the filter device of each machine, and trench images are acquired at the storage trench at the end of the common cavity oil circuit. The iron filings area is extracted from the oil images and the trench images. The iron filings density index corresponding to each detection position is determined based on the distribution characteristics of the iron filings region in the image corresponding to each detection position, and the trench accumulation change rate is determined based on the change of the degree of iron filings accumulation in the trench image over time; wherein, the iron filings density index is used to characterize the distribution density of iron filings in the oil at the corresponding detection position. For each machine, the relative deterioration index is determined based on the change between the iron filings density index corresponding to the upstream and downstream detection positions and the trench accumulation change rate; wherein, the relative deterioration index is used to characterize the degree of deterioration of the protective capability of the corresponding machine's filter device. Based on the comparison between the relative deterioration index of the machine and its upstream neighboring machine, and the distance between the machines, the relative deterioration index is corrected to obtain a corrected deterioration index; wherein, the corrected deterioration index is used to characterize the actual wear and tear of the machine's filtration device.

2. The visual inspection method for iron filings in common cavity lubricating oil according to claim 1, characterized in that, The determination of the relative deterioration index for each machine based on the change in the chip density index between the upstream and downstream detection positions and the trench accumulation change rate includes: For each machine, the contribution of each machine to the accumulation of iron filings in the storage trench is determined based on the change between the iron filing density index corresponding to the upstream and downstream detection positions of each machine and the trench accumulation change rate; wherein, the contribution degree is used to characterize the magnitude of the contribution of the iron filing change of the corresponding machine to the change of iron filings accumulation in the storage trench. Based on the changes in the contribution level of each machine over time, the machine failure rate of each machine is determined; wherein, the machine failure rate is used to characterize the rate at which the protective capability of the corresponding machine's filtration device deteriorates over time. Using the machine with the smallest absolute value of the machine failure rate among all machines as the benchmark, the machine failure rate of each machine is compared with the machine failure rate of the benchmark machine to obtain the relative deterioration index of each machine.

3. The visual inspection method for iron filings in common cavity lubricating oil according to claim 1, characterized in that, The step of determining the trench accumulation change rate based on the change in the degree of iron filings accumulation in the trench image over time includes: The trench fill rate is determined based on the coverage of the iron filings region in the trench image of the storage trench; wherein, the trench fill rate is used to characterize the degree to which the storage trench is covered by iron filings; When the trench fill rate indicates that the storage trench is not filled with iron filings, the degree of iron filings accumulation is determined based on the trench fill rate; when the trench fill rate indicates that the storage trench is filled with iron filings, the degree of iron filings accumulation is determined based on the grayscale difference between the accumulated protruding area and the surrounding area in the trench image. The trench accumulation rate is determined based on the change in the degree of iron filings accumulation over time, and the trench accumulation change rate is determined based on the change in the trench accumulation rate over time; wherein, the trench accumulation rate is used to characterize the speed at which iron filings accumulate in the storage trench per unit time, and the trench accumulation change rate is used to characterize the overall failure severity of the filter device of each machine.

4. The visual inspection method for iron filings in common cavity lubricating oil according to claim 3, characterized in that, The step of determining the degree of iron filings accumulation based on the grayscale difference between the accumulated protruding area and the surrounding area in the trench image includes: Using the region with the highest grayscale in the trench image as the center, determine the position with the greatest grayscale change along the direction away from the center, and determine the boundary between the piled-up protrusion region and the surrounding region based on the position with the greatest grayscale change; The degree of iron filings accumulation is determined based on the difference between the grayscale of the accumulated protruding area inside the boundary and the grayscale of the outer perimeter area outside the boundary.

5. The visual inspection method for iron filings in common cavity lubricating oil according to claim 1, characterized in that, The method further includes: The failure of the pre-filtration stage is determined based on the trench accumulation change rate; when the trench accumulation change rate indicates that the iron filings in the storage trench are accumulating abnormally at an accelerated rate, the pre-filtration stage is determined to be in failure. When it is determined that the pre-filtering process is in failure, the failure of the pre-filtering process is traced back to the source machine by machine according to the relative deterioration index of each machine. The pre-filtration stage includes the filtration devices of each machine.

6. The visual inspection method for iron filings in common cavity lubricating oil according to claim 1, characterized in that, The step of correcting the relative deterioration index based on the comparison between the relative deterioration index of the machine and its upstream neighboring machine, and the distance between the machines, to obtain a corrected deterioration index, includes: Based on the comparison of the relative deterioration index of the machine with that of the upstream adjacent machine, it is determined whether the machine is covered by the upstream adjacent machine; wherein, when the comparison indicates that the relative deterioration index of the machine is abnormally low relative to the relative deterioration index of the upstream adjacent machine, it is determined that the machine is covered by the upstream adjacent machine. For a machine that is obscured by an upstream neighboring machine, the relative deterioration index of the machine is corrected based on the distance between the machine and the upstream neighboring machine to obtain the corrected deterioration index of the machine.

7. The visual inspection method for iron filings in common cavity lubricating oil according to claim 6, characterized in that, The correction of the relative deterioration index of the machine includes: The weight of the loss impact of each upstream machine on the machine is determined based on the distance between the machine and each upstream machine; wherein, the closer the upstream machine is to the machine, the greater the weight of the loss impact. Based on the loss impact weight of each upstream machine and the masking status of each upstream machine, the loss impact factor of the machine is determined; wherein, the loss impact factor is used to characterize the cumulative impact of the failure of each upstream machine on the machine. The relative deterioration index of the machine is corrected based on the loss influence factor to obtain the corrected deterioration index of the machine.

8. The visual inspection method for iron filings in co-cavity lubricating oil according to any one of claims 1-7, characterized in that, The method further includes: Each machine is classified according to its modified deterioration index; wherein the classification includes at least: When the machine's corrected deterioration index meets the first determination condition, the machine is determined to be in a normal state. When the machine's correction deterioration index meets the second determination condition, a planned maintenance prompt for the machine is output; wherein, the degree of wear corresponding to the second determination condition is higher than the degree of wear corresponding to the first determination condition; When the machine's correction deterioration index meets the third determination condition, a shutdown and maintenance prompt is output for the machine; wherein the degree of wear and tear corresponding to the third determination condition is higher than the degree of wear and tear corresponding to the second determination condition.

9. The visual inspection method for iron filings in common cavity lubricating oil according to claim 8, characterized in that, The method further includes: For machines that output a shutdown and maintenance prompt, control the solenoid valve in the corresponding lubricating oil circulation circuit of the machine to cut off the oil flow in the lubricating oil circulation circuit, so as to prevent iron filings from spreading to downstream machines; Generate a maintenance work order for the machine; wherein the maintenance work order includes at least the parts of the machine that need to be replaced and the recommended maintenance window time, and the parts that need to be replaced include at least one of a fine filter element and a strong magnetic adsorption device.

10. A visual inspection system for lubricating oil chips in a common cavity, characterized in that, It includes an image acquisition unit and a data processing unit, wherein: The image acquisition unit is used to acquire oil images at the upstream and downstream detection positions of the filter device of each of the multiple machines connected in series in the common cavity oil circuit, and to acquire trench images at the storage trench at the end of the common cavity oil circuit. The data processing unit is used to extract iron filings regions from the oil image and the trench image; determine the iron filings density index corresponding to each detection position based on the distribution characteristics of the iron filings regions in the image corresponding to each detection position, and determine the trench accumulation change rate based on the change in the degree of iron filings accumulation in the trench image over time; wherein, the iron filings density index is used to characterize the distribution density of iron filings in the oil at the corresponding detection position; for each machine, a relative deterioration index is determined based on the change between the iron filings density index corresponding to the upstream and downstream detection positions of each machine and the trench accumulation change rate; wherein, the relative deterioration index is used to characterize the degree of deterioration of the protective capability of the filter device of the corresponding machine; and correct the relative deterioration index based on the comparison relationship between the relative deterioration index of each machine and the upstream adjacent machine and the distance between the machines to obtain a corrected deterioration index; wherein, the corrected deterioration index is used to characterize the actual wear and tear of the filter device of the corresponding machine.

Images