A method of cleaning a heterogeneous industrial component

Abstract

The present application relates to the field of industrial cleaning, and particularly relates to a heterogeneous industrial part cleaning method. The present application determines a heterogeneous region in a three-dimensional model, screens an induced scattering region by using a Gaussian curvature, and subsequently generates a plurality of candidate jet directions by using the induced scattering region as a constraint. A water flow excitation path index is determined by using a secondary scattering ray path of a primary jet ray in the heterogeneous region. A jet parameter is determined based on a mapping relationship between the water flow excitation path index and a jet intensity. A cleaning nozzle is controlled to cyclically switch to each candidate jet direction for cleaning. A cleaning verification is performed according to a bubble characteristic during the process. Cleaning optimization is performed based on a cleaning verification result. The present application quantitatively analyzes a geometric morphology of a heterogeneous region, controls a cleaning nozzle to cyclically switch to each candidate jet direction, and adaptively matches a jet parameter based on an analysis result. Impact turbulent flow is effectively excited in the heterogeneous region, the cleaning effect on a complex heterogeneous region is improved, and the cleaning efficiency is ensured.

Description

Technical Field

[0001] This invention relates to the field of industrial cleaning, and more particularly to a method for cleaning heterogeneous industrial components. Background Technology

[0002] In fields such as precision manufacturing, hydraulic components, engine parts, and non-standard process parts, a large number of irregularly shaped workpieces have complex internal cavity structures such as deep holes, blind cavities, intersecting holes, and grooves. These areas are prone to accumulating machining chips, oil stains, and particulate matter, which are difficult to remove effectively using traditional cleaning methods. Conventional soaking, spraying, or fixed-direction high-pressure jet cleaning often results in low-speed laminar flow of the cleaning fluid along the cavity wall for cavities with small openings and tortuous internal structures. The turbulent kinetic energy is insufficient to remove contaminants attached to the wall or dead corners. Although ultrasonic cleaning can improve the cleanliness of the internal cavity, its applicability to large or special material workpieces is limited, and it may cause surface cavitation damage.

[0003] For example, Chinese patent CN115722366A discloses a high-pressure cleaning nozzle assembly and a cleaning system. The high-pressure cleaning nozzle assembly includes: a mounting base for connecting a high-pressure liquid line and a compressed air line; a first nozzle and a second nozzle, alternatively connected to the mounting base; when the first nozzle is connected to the mounting base, it blocks the compressed air line and communicates with the high-pressure liquid line to spray liquid to clean the object; when the second nozzle is connected to the mounting base, it blocks the high-pressure liquid line and communicates with the compressed air line to spray gas to dry the object. This patent can flush localized cavities and has certain applicability.

[0004] However, the following problems still exist in the existing technology. Existing technologies for cleaning irregularly shaped parts do not consider the significant influence of the wall geometry of the local heterogeneous regions on the jet behavior. Therefore, they cannot specifically determine the jet direction and jet parameters of the impact turbulence in the excitation cavity, resulting in poor cleaning effect on the heterogeneous regions. Summary of the Invention

[0005] To address this issue, the present invention provides a method for cleaning heterogeneous industrial components, which overcomes the problem in the prior art that does not consider the significant influence of the wall geometry of the local heterogeneous region on the jet behavior, thus making it impossible to specifically determine the jet direction and jet parameters relative to the impact turbulence in the excitation cavity, resulting in poor cleaning effect on the heterogeneous region.

[0006] To achieve the above objectives, the present invention provides a method for cleaning heterogeneous industrial components, comprising: A three-dimensional model of the component to be cleaned is obtained to determine the heterogeneous region in the three-dimensional model, wherein the heterogeneous region is a cavity region including concave geometric features; The heterogeneous region is divided into several sub-regions, and the Gaussian curvature of each sub-region is determined in order to screen several induced scattering regions. Using the induced scattering regions as constraints, several candidate jet directions are generated so that the main jet ray corresponding to each candidate jet direction intersects with the induced scattering region; Based on the secondary scattered ray paths of each main jet ray in the heterogeneous region, the water flow excitation path index is determined, and the spray parameters of the cleaning nozzle are determined based on the pre-established mapping relationship between the water flow excitation path index and the spray force. The movement path of the cleaning nozzle is generated according to each of the candidate spray directions, so as to control the cleaning nozzle to cyclically switch to each of the candidate spray directions and clean the heterogeneous region with the spray parameters. The image data of the heterogeneous region is continuously monitored, bubble features are extracted from the image data, cleaning verification is performed based on the bubble features, and cleaning optimization is performed based on the cleaning verification results until the cleaning verification results are valid.

[0007] It should be noted that the process of determining heterogeneous regions in a 3D model includes, Determine the average curvature of several sub-regions on the surface of the three-dimensional model, and select concave edge regions based on the change ratio of the average curvature of the sub-regions relative to the adjacent sub-regions. Connect the recessed edge regions to form a boundary range, and select several reference points within each recessed edge region. Fit each reference point to determine the opening plane. Determine the maximum distance between the inner wall of the 3D model within the boundary range and the opening plane, in order to verify the internal region within the boundary range; The internal regions that pass the verification are identified as heterogeneous regions; Among them, the sub-regions with a change ratio greater than a predetermined change ratio threshold are defined as concave edge regions.

[0008] It should be noted that the verification of the internal region includes, If the maximum distance is greater than the distance threshold, the internal region verification passes.

[0009] It should be noted that the process of selecting several induced scattering regions and generating several candidate jet directions based on these induced scattering regions includes, The sub-regions are arranged in ascending order according to Gaussian curvature, and a predetermined proportion of sub-regions are selected from the beginning of the sequence as induced scattering regions. In three-dimensional space, the movable range of the nozzle is determined, and within the movable range, the starting point of the main jet ray is determined. The main jet ray is constructed based on the starting point, and the direction of the main jet ray is determined as the candidate jet direction. The main jet ray constructed must pass through the opening plane.

[0010] It should be noted that the process of determining the water flow excitation path index based on the secondary scattered ray paths of each main jet ray in the heterogeneous region includes, The point where the main jet ray intersects with the induced scattering region is determined as the impact point; Determine the wall normal vector at the location of the impact point, and based on the reflection path of the main jet ray after specular reflection relative to the wall normal vector, determine the reflection path as the secondary scattered ray path; A diffuse reflection cone is constructed with the secondary scattered ray path as the central axis and the impact point as the vertex. The inner wall of the heterogeneous region where the diffuse reflection cone intersects is determined, and several reference points of the inner wall of the heterogeneous region are selected. The average distance between the reference point and the impact point is determined as the water flow excitation path index.

[0011] It should be noted that the process of determining the spray parameters of the cleaning nozzle based on the pre-established mapping relationship between the water flow excitation path index and the spray force includes, Establish a mapping relationship between water flow excitation path indicators and jet force in advance, including the jet force mapped by different water flow excitation path indicator constraint ranges; Determine the water flow induction path index constraint range to which the water flow induction path belongs, in order to select the mapped injection parameters; The water flow excitation path index constraint range is mapped one-to-one with the jetting parameters, including the jetting force.

[0012] It should be noted that the cleaning nozzle moves continuously during the process of cyclically switching to each of the candidate spray directions, and the spray parameters of the cleaning nozzle need to be re-determined when reaching a new candidate moving path.

[0013] It should be noted that the process of continuously monitoring the image data of the heterogeneous region and extracting bubble features from the image data includes, Binarize consecutive frames of image data from the heterogeneous region and annotate the bubble outline; Bubble characteristics are determined based on the bubble profile, including the maximum bubble width and bubble density.

[0014] It should be noted that the process of cleaning verification based on bubble characteristics includes, Compare the bubble features with predefined bubble feature constraints; If the verification conditions are met, the cleaning verification result is deemed valid. The verification condition is that the bubble features do not deviate from the preset bubble feature constraints.

[0015] It should be noted that when optimizing the cleaning process based on the cleaning verification results, this includes increasing the spray speed or changing the spray direction.

[0016] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention obtains a three-dimensional model of the component to be cleaned, identifies heterogeneous regions within the model, uses Gaussian curvature to filter out induced scattering regions, and then generates several candidate jet directions using each induced scattering region as a constraint. Subsequently, it uses the secondary scattering ray paths of the main jet ray within the heterogeneous regions to determine the water flow excitation path index, and determines the jet parameters of the cleaning nozzle based on the mapping relationship between the water flow excitation path index and the jet force. The cleaning nozzle is then controlled to cyclically switch to each of the candidate jet directions for cleaning. During the process, cleaning verification is performed based on bubble characteristics, and cleaning optimization is performed based on the verification results. This invention, through quantitative analysis of the geometry of heterogeneous regions, controls the cleaning nozzle to cyclically switch to each candidate jet direction, and adaptively matches jet parameters based on the analysis results, effectively excites impact turbulence within heterogeneous regions, improving the cleaning effect on complex heterogeneous regions and ensuring cleaning efficiency.

[0017] Beneficially, this invention identifies heterogeneous regions and, based on their Gaussian curvature, selects several induced scattering regions. In practice, for regions with negative Gaussian curvature, the larger the absolute value of the Gaussian curvature, the more complex the surface morphology becomes, exhibiting bidirectional curvature and concavity. When a jet impacts such regions, the tearing and diversion effects of the wall on the fluid are significant, making it easier to induce jet splitting and generate initial vortices. This invention uses these induced scattering regions as target points to generate candidate jet directions, ensuring that the jet energy is precisely applied to the geometrically significant locations within the cavity most prone to turbulence. Compared to the traditional method of indiscriminately flushing the entire cavity, this invention achieves a stronger turbulence-inducing effect with less energy consumption, thereby improving the cleaning effect and ensuring cleaning efficiency.

[0018] Beneficially, this invention considers the secondary scattered ray paths of the main jet within the heterogeneous region to determine the water flow excitation path index. This index reflects the distance the water travels after being scattered at the impact point and then impacting the wall again. A smaller index indicates a stronger spatial constraint on the scattering water flow from the cavity wall, forcing the water flow to undergo secondary impacts and deflections within a shorter distance. This results in denser continuous reflections and vortex superposition within a limited space, significantly enhancing turbulence intensity. This invention establishes a mapping relationship between the water flow excitation path index and jet parameters, enabling the cleaning system to automatically match the jet force according to the actual geometric constraints of the cavity, thereby improving cleaning effectiveness and ensuring cleaning efficiency.

[0019] Beneficially, in the actual cleaning process of this invention, image data of the heterogeneous region is continuously monitored, bubble features are extracted, and cleaning verification is performed. When the water flow in the heterogeneous region is in a state of strong turbulence, a large number of microbubbles will be generated due to cavitation effect and gas entrainment; conversely, if the water flow is gentle and turbulence is insufficient, the amount of bubble generation is significantly reduced. Therefore, the density and size distribution of bubbles can serve as a visual proxy indicator of turbulence intensity, indirectly reflecting the cleaning effect. This invention utilizes physical properties to extract bubble features in real time through a vision system for cleaning verification, thereby optimizing the cleaning process and spraying parameters to improve the cleaning effect and ensure cleaning efficiency. Attached Figure Description

[0020] Figure 1 This is a step diagram of a method for cleaning heterogeneous industrial components according to an embodiment of the invention; Figure 2 A logic block diagram for determining heterogeneous regions of a three-dimensional model in an embodiment of the invention; Figure 3 This is a cross-sectional schematic diagram of the diffuse reflection cone according to an embodiment of the invention; Figure 4 This is a logic block diagram for performing cleaning verification according to an embodiment of the invention; In the diagram, 1: the part to be cleaned, 2: the main jet ray, 3: the diffuse reflection cone, and 4: the path of the secondary scattered rays. Detailed Implementation

[0021] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0022] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0023] Please see Figure 1 The diagram shows the steps of a heterogeneous industrial component cleaning method according to an embodiment of the invention. The heterogeneous industrial component cleaning method according to this embodiment includes: Step S1: Obtain a three-dimensional model of the component 1 to be cleaned, and determine the heterogeneous region in the three-dimensional model. The heterogeneous region is a cavity region including concave geometric features. Step S2: Divide the heterogeneous region into several sub-regions and determine the Gaussian curvature of each sub-region to screen several induced scattering regions; Step S3: Using each of the induced scattering regions as constraints, generate several candidate jet directions so that the main jet ray 2 corresponding to each of the candidate jet directions intersects with the induced scattering region; Step S4: Based on the secondary scattered ray paths of each of the main jet rays 2 in the heterogeneous region, determine the water flow excitation path index, and determine the spray parameters of the cleaning nozzle based on the pre-established mapping relationship between the water flow excitation path index and the spray force. Step S5: Generate a moving path for the cleaning nozzle based on each of the candidate spray directions, so as to control the cleaning nozzle to cyclically switch to each of the candidate spray directions and clean the heterogeneous region with the spray parameters. Step S6: Continuously monitor the image data of the heterogeneous region, extract bubble features from the image data, perform cleaning verification based on the bubble features, and optimize the cleaning based on the cleaning verification results until the cleaning verification results are valid.

[0024] In practice, there are no restrictions on how the cleaning nozzles are set up. For example, in typical modern industrial scenarios, the cleaning nozzles can be mounted on the end effector of a multi-degree-of-freedom robotic arm, and the spray direction can be changed by adjusting the position of the robotic arm. At the same time, the cleaning nozzles must have a spray force adjustment function to meet the spray parameter requirements corresponding to different water flow excitation path indicators; this will not be elaborated further.

[0025] In practice, there are no restrictions on the method of monitoring images of heterogeneous areas. For example, an industrial CCD camera can be used to acquire images of heterogeneous areas. The industrial CCD camera and the cleaning nozzle can both be mounted on the robotic arm to acquire images in the corresponding spray direction of the cleaning nozzle. This will not be elaborated further.

[0026] In practice, there are no restrictions on how the 3D model of the component to be cleaned 1 is acquired. For example, point cloud data can be collected online by a 3D scanner and the model can be reconstructed, or the pre-stored digital model of the industrial component can be directly imported. This will not be elaborated further.

[0027] In practice, the order of cleaning heterogeneous areas and overall cleaning is not limited. As a preferred embodiment, after completing the targeted fine cleaning of the heterogeneous areas, the component 1 to be cleaned can be sprayed or rinsed as a whole. First, the heterogeneous areas where pollutants are most likely to accumulate and are most difficult to clean are cleaned with high-intensity, multi-angle turbulent stimulation cleaning, which can ensure that stubborn stains are removed under the most favorable conditions. The subsequent overall spraying or rinsing can thoroughly remove these secondary pollutants and avoid cross-contamination.

[0028] Please see Figure 2 As shown, it is a logic block diagram for determining heterogeneous regions of a 3D model according to an embodiment of the invention. The process of determining heterogeneous regions of a 3D model includes, Determine the average curvature of several sub-regions on the surface of the three-dimensional model, and select concave edge regions based on the change ratio of the average curvature of the sub-regions relative to the adjacent sub-regions. Connect the recessed edge regions to form a boundary range, and select several reference points within each recessed edge region. Fit each reference point to determine the opening plane. Determine the maximum distance between the inner wall of the 3D model within the boundary range and the opening plane, in order to verify the internal region within the boundary range; The internal regions that pass the verification are identified as heterogeneous regions; Among them, the sub-regions with a change ratio greater than a predetermined change ratio threshold are defined as concave edge regions.

[0029] In practice, the inner and outer sides of the three-dimensional model can be defined in three-dimensional space, and the concavity and convexity can be determined based on the positive and negative values ​​of the average curvature of the sub-regions.

[0030] It is understandable that the edge of the cavity region is a transition zone between the external non-recessed surface and the internal concave surface, where the curvature undergoes a significant abrupt change. Therefore, the gradient of the average curvature change is much greater than that of a flat surface or a smooth area of ​​the cavity's inner wall. By selecting locally adjacent regions with relatively large changes in average curvature and connecting them sequentially, the opening edge of the cavity can be determined.

[0031] In practice, the predetermined change ratio threshold is set in advance. Several 3D model samples can be manually labeled in advance to determine the change ratio of the sub-regions at the edge of the cavity region relative to the adjacent sub-regions. The lower limit of the change ratio is used as the change ratio threshold.

[0032] During implementation, the recessed edge area needs to be verified. Only when the recessed edge area can be connected into a closed range, and the minimum diameter of the closed range is greater than 5cm, can it be regarded as the boundary range. The significance is to screen out and exclude small recessed structures, such as machining chamfers, threaded edges, surface scratches, or micropores. The minimum diameter threshold of 5cm also ensures that there is enough room for operation in the subsequent spray direction planning, avoiding the jet from not being able to enter effectively or the nozzle movement path not being able to be planned reasonably due to the cavity being too narrow.

[0033] It is understandable that the determined reference points are distributed at the opening positions of the heterogeneous region. Therefore, the reference points can be fitted to determine the opening plane. When selecting reference points, reference points can be selected in different induced scattering regions, and the number of reference points should not be less than 3. Specifically, when determining the opening plane, the criterion is to minimize the sum of the squares of the vertical distances from each reference point to the target plane, and find the opening plane that is closest to all points. Of course, other methods can also be used, which are existing technologies and will not be elaborated here.

[0034] To elaborate, the verification of the internal region includes, If the maximum distance is greater than the distance threshold, the internal region verification passes.

[0035] In practice, setting a distance threshold is significant in measuring the depth of the cavity. For shallower internal areas, deep cleaning is unnecessary. In practice, the distance threshold is set to 3cm. From a fluid dynamics perspective, for cavities with a depth of less than 3cm, the jet enters before the turbulent structure has fully developed and reaches the bottom of the cavity and reflects out. The constraint effect of secondary impact is weak, and the gain of directional fine cleaning compared to conventional rinsing is not significant. Secondly, from an industrial practice perspective, the depth of deep cavities and blind holes in common irregularly shaped workpieces such as hydraulic valve bodies, engine cylinder heads, and pump housings that require special cleaning is usually more than 3cm. However, for shallow grooves and countersunk holes with a depth of less than 3cm, the amount of processing residue is small and the adhesion is relatively weak, so conventional rinsing is sufficient to remove it.

[0036] To elaborate, the process of selecting several induced scattering regions and generating several candidate jet directions based on these induced scattering regions includes: The sub-regions are arranged in ascending order according to Gaussian curvature, and a predetermined proportion of sub-regions are selected from the beginning of the sequence as induced scattering regions. In three-dimensional space, the movable range of the nozzle is determined, and within the movable range, the starting point of the main jet ray 2 is determined. The main jet ray 2 is constructed based on the starting point, and the direction of the main jet ray 2 is determined as the candidate jet direction. The main jet ray 2 constructed must pass through the opening plane.

[0037] It is understandable that Gaussian curvature reflects the inherent bending properties of a surface. Therefore, water flow impacts a sub-region with lower Gaussian curvature are more likely to induce turbulence in that sub-region. For example, when the Gaussian curvature is small, especially when the value is negative, the surface exhibits a saddle-shaped bidirectional bending. When water flow impacts the surface, the fluid is stretched by the wall in two opposing directions simultaneously, generating strong shearing and diversion effects, which in turn makes it easier to induce turbulence.

[0038] The predetermined ratio can be set between 30% and 50% to select a sub-region with a relatively small Gaussian curvature as the induced scattering region.

[0039] Understandably, due to the spatial arrangement on site, the range of motion of the nozzle on the robotic arm is limited. Therefore, within the movable range corresponding to the nozzle, the starting point can be determined to construct the main jet ray 2, simulating the path of the water flow after it is ejected at the corresponding position. Thus, the main jet ray 2 needs to pass through the opening plane.

[0040] In practice, to simplify the calculation, for a single induced scattering region, only one corresponding main jet ray can be determined.

[0041] This invention identifies heterogeneous regions and, based on their Gaussian curvature, selects several induced scattering regions. In practice, for regions with negative Gaussian curvature, the larger the absolute value of the Gaussian curvature, the more complex the surface morphology, exhibiting bidirectional curvature and concavity. When a jet impacts such regions, the tearing and diversion effects of the wall on the fluid are significant, making it easier to induce jet splitting and generate initial vortices. This invention uses these induced scattering regions as target points to generate candidate jet directions, ensuring that the jet energy is precisely applied to the geometrically characteristic locations within the cavity most prone to turbulence. Compared to the traditional method of indiscriminately flushing the entire cavity, this invention achieves a stronger turbulence-inducing effect with less energy consumption, thereby improving the cleaning effect and ensuring cleaning efficiency.

[0042] Please see Figure 3 As shown, this is a cross-sectional schematic diagram of the diffuse reflection cone according to an embodiment of the invention. The process of determining the water flow excitation path index based on the secondary scattered ray path 4 of each of the main jet rays 2 in the heterogeneous region includes, The intersection of the main jet ray 2 and the induced scattering region is determined as the impact point; Determine the wall normal vector at the location of the impact point, and based on the reflection path of the main jet ray 2 after specular reflection relative to the wall normal vector, determine the reflection path as the secondary scattered ray path 4; A diffuse reflection cone 3 is constructed with the secondary scattered ray path 4 as the central axis and the impact point as the vertex. The inner wall of the heterogeneous region where the diffuse reflection cone 3 intersects is determined, and several reference points of the inner wall of the heterogeneous region are selected. The average distance between the reference point and the impact point is determined as the water flow excitation path index.

[0043] It should be noted that the actual diffuse reflection process of fluid is affected by many factors such as Reynolds number, surface tension, and wall roughness, and has a high degree of uncertainty. However, the spatial constraint effect of the cavity geometry on fluid motion is certain. Therefore, this invention adopts a simplified analysis method, only considering the theoretical impact point of water flow and the path after scattering. Then, the water flow excitation path index is determined by the second-order scattered rays, and the spatial constraint on fluid motion is quantified.

[0044] Understandably, the actual trajectory of a fluid after impacting a wall is highly uncertain and difficult to model accurately. This invention approximates the potential reflection range after fluid impact by constructing a diffuse reflection cone. This cone, with the specular reflection direction as its main axis, simplifies the complex fluid diffusion behavior into an envelope space determined by geometric parameters, thereby simplifying the analysis and allowing subsequent consideration of the spatial constraints on fluid motion.

[0045] In practice, the cone angle of the diffuse reflection cone 3 is inversely related to the incident angle, which is the angle between the jet direction and the wall normal vector; When the incident angle is large, the fluid spreads further along the mirror reflection direction after impacting the wall, and the energy is more concentrated. Therefore, a smaller cone angle is set to fully capture the secondary impact area at the far end. When the incident angle is small, the fluid impacts almost vertically, and the splash range is wider. Therefore, a wider cone angle is set.

[0046] In practice, optionally, when the incident angle is less than 30°, the cone angle is set to 90°; Optionally, when the incident angle is greater than 30°, the cone angle is set to 60°; Of course, those skilled in the art can make adaptive adjustments to the cone angle, which will not be elaborated here.

[0047] This invention considers the secondary scattered ray paths of the main jet within the heterogeneous region to determine the water flow excitation path index. This index reflects the distance the water travels after being scattered at the impact point and then impacting the wall again. A smaller index indicates a stronger spatial constraint on the scattering water flow from the cavity wall, forcing the water flow to undergo secondary impacts and deflections within a shorter distance. This results in denser continuous reflections and vortex superposition within a limited space, significantly enhancing turbulence intensity. This invention establishes a mapping relationship between the water flow excitation path index and jet parameters, enabling the cleaning system to automatically match the jet force according to the actual geometric constraints of the cavity, thereby improving cleaning effectiveness and ensuring cleaning efficiency.

[0048] To elaborate, the process of determining the spray parameters of the cleaning nozzle based on the pre-established mapping relationship between water flow excitation path indicators and spray force includes: Establish a mapping relationship between water flow excitation path indicators and jet force in advance, including the jet force mapped by different water flow excitation path indicator constraint ranges; Determine the water flow induction path index constraint range to which the water flow induction path belongs, in order to select the mapped injection parameters; The water flow excitation path index constraint range is mapped one-to-one with the jetting parameters, including the jetting force.

[0049] In practice, the constraint range of the water flow activation path index is a continuous numerical range.

[0050] As an optional implementation method, the mapping relationship can be established through prior experimental calibration. Specifically, a series of typical cavity samples with different depths and opening sizes are selected, and the water flow activation path index is calculated for each of them. On a controllable cleaning test bench, each cavity sample is cleaned with different spray intensities, and the minimum spray intensity required to achieve the cleaning standard is recorded. Based on the experimental records compiled according to the constraints of the water flow excitation path index, several minimum jetting intensities corresponding to different water flow excitation path index constraints were determined, and the average minimum jetting intensity corresponding to the water flow excitation path index constraints was determined. Then, the mapping relationship between the water flow excitation path index constraints and the minimum jetting intensity was established.

[0051] The constraint range of the water flow induction path index is essentially a closed interval. The length of the interval is determined based on the difference between the maximum and minimum values ​​of the water flow induction path index during the experiment. The length of the constraint range of the water flow induction path index is set to 20% of the difference.

[0052] To elaborate, the cleaning nozzle moves continuously as it cycles through the candidate spray directions, and the spray parameters of the cleaning nozzle need to be redefined when it reaches a new candidate moving path.

[0053] Understandably, continuous movement can eliminate cleaning dead zones, further stimulate turbulence superposition, and improve cleaning effect.

[0054] To elaborate, the process of continuously monitoring the image data of the heterogeneous region and extracting bubble features from the image data includes, Binarize consecutive frames of image data from the heterogeneous region and annotate the bubble outline; Bubble characteristics are determined based on the bubble profile, including the maximum bubble width and bubble density.

[0055] In practice, there are no restrictions on the method for determining the bubble outline. For example, an image segmentation algorithm can be used to process the image data of the heterogeneous region to separate the bubble region from the background fluid region and then extract the bubble outline. Of course, other methods can also be used, which will not be elaborated here.

[0056] The maximum width and density of the bubble can be determined based on the bubble outline. The bubble density is the ratio of the total area of ​​the bubble outline in the image data to the area of ​​the fluid region.

[0057] Please see Figure 4 As shown, it is a logic block diagram for cleaning verification according to an embodiment of the invention. The process of cleaning verification based on bubble characteristics includes, Compare the bubble features with predefined bubble feature constraints; If the verification conditions are met, the cleaning verification result is deemed valid. The verification condition is that the bubble features do not deviate from the preset bubble feature constraints.

[0058] During implementation, bubble characteristic constraints can be determined through pre-cleaning experiments. The specific process includes obtaining several samples to be cleaned, determining the water flow excitation path index in the heterogeneous region, artificially applying standard pollutants in the heterogeneous region and cleaning, and recording the time series data of the samples to be cleaned that meet the cleanliness standard and the corresponding bubble characteristics. The recorded data are classified according to the water flow excitation path index. For samples to be cleaned that belong to the same water flow excitation path index constraint range, the data recorded during the cleaning process are classified into one category. The normal distribution of the bubble characteristics after classification is calculated, and the 95% confidence interval is used as the bubble characteristic constraint.

[0059] In practice, bubble features include the maximum bubble width and bubble density. Therefore, bubble feature constraints include the maximum bubble width constraint and the bubble density constraint. Thus, it is necessary to determine the normal distribution of the maximum bubble width and the bubble density and the corresponding 95% confidence interval, respectively. Then, the 95% confidence interval of the normal distribution of the maximum bubble width is used as the maximum bubble width constraint, and the 95% confidence interval of the normal distribution of the bubble density is used as the bubble density constraint.

[0060] It is understandable that the bubble feature constraint corresponds to the range of water flow excitation path index constraint. In practice, when performing cleaning verification, the selected bubble feature constraint should match the water flow excitation path corresponding to the component 1 to be cleaned.

[0061] To elaborate, when optimizing cleaning based on the cleaning verification results, this includes increasing the spray speed or changing the spray direction.

[0062] In practice, the spraying speed can optionally be increased to 10% of the initial speed to avoid excessive adjustment at once and to facilitate repeated adjustments until the cleaning verification results are effective.

[0063] If the cleaning verification results are still invalid after adjusting the spray speed three times, the spray speed will no longer be adjusted. The spray direction can be changed appropriately. Based on the original candidate spray direction, it can be shifted 5° in any direction. Note that water flow should be avoided from spraying into non-heterogeneous areas.

[0064] In the actual cleaning process of this invention, image data of the heterogeneous region is continuously monitored, bubble features are extracted, and cleaning verification is performed. When the water flow in the heterogeneous region is in a state of strong turbulence, a large number of microbubbles will be generated due to cavitation effect and gas entrainment; conversely, if the water flow is gentle and turbulence is insufficient, the amount of bubble generation will be significantly reduced. Therefore, the density and size distribution of bubbles can serve as a visual proxy indicator of turbulence intensity, indirectly reflecting the cleaning effect. This invention utilizes physical properties to extract bubble features in real time through a vision system for cleaning verification, thereby optimizing the cleaning process and spraying parameters to improve the cleaning effect and ensure cleaning efficiency.

[0065] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A method for cleaning heterogeneous industrial components, characterized in that, include: A three-dimensional model of the component to be cleaned is obtained to determine the heterogeneous region in the three-dimensional model, wherein the heterogeneous region is a cavity region including concave geometric features; The heterogeneous region is divided into several sub-regions, and the Gaussian curvature of each sub-region is determined in order to screen several induced scattering regions. Using the induced scattering regions as constraints, several candidate jet directions are generated so that the main jet ray corresponding to each candidate jet direction intersects with the induced scattering region; Based on the secondary scattered ray paths of each main jet ray in the heterogeneous region, the water flow excitation path index is determined, and the spray parameters of the cleaning nozzle are determined based on the pre-established mapping relationship between the water flow excitation path index and the spray force. The movement path of the cleaning nozzle is generated according to each of the candidate spray directions, so as to control the cleaning nozzle to cyclically switch to each of the candidate spray directions and clean the heterogeneous region with the spray parameters. The image data of the heterogeneous region is continuously monitored, bubble features are extracted from the image data, cleaning verification is performed based on the bubble features, and cleaning optimization is performed based on the cleaning verification results until the cleaning verification results are valid.

2. The cleaning method for heterogeneous industrial components according to claim 1, characterized in that, The process of determining heterogeneous regions in a 3D model includes, Determine the average curvature of several sub-regions on the surface of the three-dimensional model, and select concave edge regions based on the change ratio of the average curvature of the sub-regions relative to the adjacent sub-regions. Connect the recessed edge regions to form a boundary range, and select several reference points within each recessed edge region. Fit each reference point to determine the opening plane. Determine the maximum distance between the inner wall of the 3D model within the boundary range and the opening plane, in order to verify the internal region within the boundary range; The internal regions that pass the verification are identified as heterogeneous regions; Among them, the sub-regions with a change ratio greater than a predetermined change ratio threshold are defined as concave edge regions.

3. The cleaning method for heterogeneous industrial components according to claim 2, characterized in that, Verification of the internal region includes... If the maximum distance is greater than the distance threshold, the internal region verification passes.

4. The cleaning method for heterogeneous industrial components according to claim 3, characterized in that, The process of selecting several induced scattering regions and generating several candidate jet directions based on each of these induced scattering regions includes: The sub-regions are arranged in ascending order according to Gaussian curvature, and a predetermined proportion of sub-regions are selected from the beginning of the sequence as induced scattering regions. In three-dimensional space, the movable range of the nozzle is determined, and within the movable range, the starting point of the main jet ray is determined. The main jet ray is constructed based on the starting point, and the direction of the main jet ray is determined as the candidate jet direction. The main jet ray constructed must pass through the opening plane.

5. The cleaning method for heterogeneous industrial components according to claim 1, characterized in that, The process of determining the water flow excitation path index based on the secondary scattered ray paths of each main jet ray in the heterogeneous region includes, The point where the main jet ray intersects with the induced scattering region is determined as the impact point; Determine the wall normal vector at the location of the impact point, and based on the reflection path of the main jet ray after specular reflection relative to the wall normal vector, determine the reflection path as the secondary scattered ray path; A diffuse reflection cone is constructed with the secondary scattered ray path as the central axis and the impact point as the vertex. The inner wall of the heterogeneous region where the diffuse reflection cone intersects is determined, and several reference points of the inner wall of the heterogeneous region are selected. The average distance between the reference point and the impact point is determined as the water flow excitation path index.

6. The cleaning method for heterogeneous industrial components according to claim 1, characterized in that, The process of determining the spray parameters of the cleaning nozzle based on the pre-established mapping relationship between water flow excitation path index and spray force includes: Establish a mapping relationship between water flow excitation path indicators and jet force in advance, including the jet force mapped by different water flow excitation path indicator constraint ranges; Determine the water flow induction path index constraint range to which the water flow induction path belongs, in order to select the mapped injection parameters; The water flow excitation path index constraint range is mapped one-to-one with the jetting parameters, including the jetting force.

7. The cleaning method for heterogeneous industrial components according to claim 6, characterized in that, The cleaning nozzle moves continuously as it cycles through the candidate spray directions. When it reaches a new candidate moving path, the spray parameters of the cleaning nozzle need to be redefined.

8. The cleaning method for heterogeneous industrial components according to claim 1, characterized in that, The process of continuously monitoring image data of the heterogeneous region and extracting bubble features from the image data includes, Binarize consecutive frames of image data from the heterogeneous region and annotate the bubble outline; Bubble characteristics are determined based on the bubble profile, including the maximum bubble width and bubble density.

9. The cleaning method for heterogeneous industrial components according to claim 1, characterized in that, The process of cleaning verification based on bubble characteristics includes, Compare the bubble features with predefined bubble feature constraints; If the verification conditions are met, the cleaning verification result is deemed valid. The verification condition is that the bubble features do not deviate from the preset bubble feature constraints.

10. The cleaning method for heterogeneous industrial components according to claim 1, characterized in that, When optimizing cleaning based on cleaning validation results, the following should be included: Increase the spray speed or change the spray direction.

Images