A laser polishing control method and system for a platform

Abstract

The present application relates to the technical field of laser control, and particularly relates to a laser grinding control method and system for a platform, which quantifies the spatial distribution and material quantity difference of each type of area in the platform to be ground by comparing the target depth value with the actual depth value, extracting the convex area and the concave area from the surface image and calculating the volume, avoids the precision loss caused by blind processing, analyzes the mutual matching degree between the convex area and the concave area by comprehensively considering the positional relationship between the areas, the volume matching degree and other parameters, thereby quantitatively representing the feasibility and efficiency of the material transfer between the convex area and the concave area when heated, generates a grinding path based on the mutual matching degree, allocates the heating and cooling time sequence in combination with the material thermal characteristics, and converts the execution instructions of the laser grinding equipment, so that the laser beam motion path is rationalized and the energy output is accurately controlled, the idle stroke and thermal deformation are reduced, the thermal stress is controlled through periodic cooling, and the grinding efficiency and the grinding effect of the platform are improved.

Description

Technical Field

[0001] This invention relates to the field of laser control technology, and in particular to a laser polishing control method and system for a platform. Background Technology

[0002] Existing laser grinding control methods primarily rely on layered processing of images acquired from the side of the platform. This involves recording the position, contour, and relationships between adjacent layers of each layer to form workpiece layering data. This layering information is then fused with processing information to calculate and control the specific processing path, laser treatment time, and grinding passes for each layer based on processing accuracy requirements. However, these methods rely on fixed thresholds or simple geometric features for layering, failing to consider the material's flow characteristics after heating and the height of bulges around the heat-affected zone caused by laser heating. They cannot quantify the potential for material transfer and compensation between raised and recessed areas, making global optimization difficult. Furthermore, this results in higher platform roughness, leading to poor efficiency and effectiveness in laser grinding.

[0003] Therefore, how to control the laser grinding process to improve laser grinding efficiency and platform grinding effect has become an urgent problem to be solved. Summary of the Invention

[0004] To address the aforementioned technical problems, the present invention provides a laser polishing control method for a platform, which includes the following steps:

[0005] S1. Based on the target depth value, surface image, and point cloud data corresponding to the platform to be ground, obtain several raised areas, several recessed areas, the raised volume corresponding to each raised area, and the recessed volume corresponding to each recessed area in the surface image. The point cloud data includes the actual depth value corresponding to each pixel in the surface image.

[0006] S2, based on the position of each raised region and each recessed region in the surface image, the raised volume corresponding to each raised region and the recessed volume corresponding to each recessed region, obtain the degree of mutual matching between each raised region and each recessed region, where the degree of mutual matching is a value greater than or equal to 0.

[0007] S3. Based on the degree of matching between each raised area and each recessed area, the grinding path of the laser beam and the corresponding timing control data are obtained. The timing control data includes the state mode of the laser beam at each unit time point when grinding the platform to be ground according to the grinding path. The state mode includes heating mode and cooling mode.

[0008] S4, Grind the platform to be ground according to the grinding path and control data.

[0009] The present invention also provides a laser polishing control system for a platform, the control system comprising:

[0010] The region extraction module is used to obtain several raised regions, several recessed regions, the raised volume of each raised region, and the recessed volume of each recessed region in the surface image based on the target depth value, surface image, and point cloud data corresponding to the platform to be ground. The point cloud data includes the actual depth value corresponding to each pixel in the surface image.

[0011] The mutual matching degree acquisition module is used to obtain the mutual matching degree between each convex region and each concave region based on the position of each convex region and each concave region in the surface image, the convex volume corresponding to each convex region and the concave volume corresponding to each concave region, where the mutual matching degree is a value greater than or equal to 0.

[0012] The control data acquisition module is used to acquire the grinding path of the laser beam and the corresponding timing control data based on the degree of matching between each raised area and each recessed area. The timing control data includes the state mode of the laser beam at each unit time point when grinding the platform to be ground according to the grinding path. The state mode includes heating mode and cooling mode.

[0013] The grinding control module is used to grind the grinding platform according to the grinding path and control data.

[0014] This invention has at least the following beneficial effects: By comparing the target depth value with the actual depth value in the point cloud data, and extracting the raised and recessed areas from the surface image and calculating the volume, the spatial distribution and material quantity differences of various types of areas in the platform to be ground are quantified, providing a geometric and physical basis for subsequent material transfer, avoiding the loss of precision caused by blind processing. By integrating multi-dimensional parameters such as the positional relationship and volume matching degree of the raised and recessed areas, the degree of mutual matching between the raised and recessed areas is analyzed, thereby quantifying the feasibility and efficiency of material transfer between the raised and recessed areas when heated, facilitating the priority matching of high-value area pairs, improving the rationality of the overall processing plan and resource utilization, generating a grinding path based on the degree of mutual matching, allocating heating / cooling sequence based on the thermal properties of the material, and converting the planned grinding path and timing control data into execution instructions for the laser grinding equipment, making the laser beam movement path rational and the energy output precise, reducing idle travel and thermal deformation, and controlling thermal stress through periodic cooling, thereby improving grinding efficiency and the grinding effect of the platform. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0016] Figure 1 A flowchart of a laser polishing control method for a platform provided in Embodiment 1 of the present invention;

[0017] Figure 2 This is a schematic diagram of a laser polishing control system for a platform provided in Embodiment 2 of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It is understood that, where appropriate, the terms used to distinguish similar objects can be interchanged so that the invention can also be implemented in other embodiments besides the illustrated or described embodiments. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products, or devices.

[0020] Example 1

[0021] This embodiment provides a laser polishing control method for a platform, such as... Figure 1 As shown, the laser polishing control method for the platform includes the following steps:

[0022] S1. Based on the target depth value, surface image, and point cloud data corresponding to the platform to be ground, obtain several raised areas, several recessed areas, the raised volume corresponding to each raised area, and the recessed volume corresponding to each recessed area in the surface image. The point cloud data includes the actual depth value corresponding to each pixel in the surface image.

[0023] The platform to be ground is a workpiece carrier whose surface geometry or roughness does not meet design requirements and needs to be repaired and smoothed using laser technology. Laser grinders utilize the high energy density of the output laser beam to scan and flow molten material in the raised areas of the platform to be ground, filling the recessed areas. After cooling, the surface is smoothed, making the platform meet design requirements.

[0024] The target depth value is the ideal design depth of the platform to be ground, such as the standard thickness of a mold, which can be determined by the implementer based on the actual situation such as the CAD model or process requirements.

[0025] The surface image is a two-dimensional image of the upper surface of the platform to be ground, captured by an image acquisition device such as a camera. Point cloud data is obtained from point cloud data acquisition devices such as laser scanners and structured light cameras, providing the geometric shape information of the upper surface of the platform. The point cloud data is aligned with the surface image to ensure that each pixel corresponds to a unique actual depth value, serving as the basis for laser control.

[0026] The raised area is a continuous region where the actual depth is lower than the target depth. It can be considered the "material source" in the laser polishing process and can be filled by melting the material in the raised area with laser heat. The depressed area is a continuous region where the actual depth is higher than the target depth. It can be considered the "repair target" in laser polishing and needs to be filled with the material melted from the raised area to reach the target depth.

[0027] The volume of a protrusion can characterize the amount of material available in the corresponding protruding area, while the volume of a depression can characterize the amount of material required in the corresponding depression area.

[0028] In one specific embodiment, S1 includes the following steps:

[0029] S11, obtain the target platform thickness, surface image, point cloud data, and the first distance between the bottom surface of the platform to be ground and the point cloud data acquisition device. The actual depth value of each pixel in the point cloud data is the distance between the corresponding position of each pixel on the platform to be ground and the point cloud data acquisition device.

[0030] S12, the difference between the first distance and the thickness of the target platform is determined as the target depth value corresponding to the platform to be ground.

[0031] S13, the difference between the target depth value and the actual depth value corresponding to each pixel is determined as the height difference corresponding to each pixel.

[0032] S14. Based on the comparison between the height difference corresponding to each pixel and the preset height threshold, obtain several raised areas and several recessed areas in the surface image.

[0033] S15: Obtain the protrusion volume corresponding to each protrusion region based on the height difference corresponding to each pixel in each protrusion region.

[0034] S16, obtain the depression volume corresponding to each depression region based on the height difference corresponding to each pixel in each depression region.

[0035] The preset height threshold is a limit value for the height difference between raised and recessed areas, which can be set by the implementer based on factors such as platform accuracy requirements and material characteristics.

[0036] The target platform thickness is an ideal parameter set based on the platform design requirements. The target depth value represents the standard distance that should be reached between each point on the upper surface of the platform and the point cloud data acquisition device after laser grinding, providing a unified quantitative benchmark for the regional division of the entire platform.

[0037] For each identified raised / recessed region, the height differences of each pixel within that region are accumulated, and combined with the actual area of ​​each pixel, the total volume of the raised / recessed region is calculated through integration or summation. The raised volume is the material volume of the portion of each raised region below the target depth, used to quantify the amount of material in the raised region. The recessed volume is the spatial volume of the portion of each recessed region above the target depth, used to quantify the amount of material needed to fill the recessed region, providing a basis for material removal and transfer during the grinding process. The actual area corresponding to each pixel is determined by the accuracy of the point cloud data acquisition.

[0038] The above-mentioned method integrates point cloud data, surface images, target platform thickness, and equipment distance to achieve complementary analysis of the three-dimensional geometric features and two-dimensional visual features of the surface of the platform to be ground. This overcomes the limitations of a single data source, improves the accuracy of identifying raised and recessed areas, and enhances the accuracy of calculating raised and recessed volumes.

[0039] In one specific embodiment, S14 includes the following steps:

[0040] S141, for any pixel, if the height difference corresponding to the current pixel is greater than a preset height threshold, then the first marker value corresponding to the pixel is determined as the first category value, wherein the category of the pixel corresponding to the first category value is a raised pixel.

[0041] S142, if the height difference corresponding to the current pixel is less than the preset height threshold, then the first marker value corresponding to the pixel is determined as the second category value, wherein the category of the pixel corresponding to the second category value is a concave pixel.

[0042] S143, Update the first label value corresponding to all pixels according to the region growing algorithm to obtain the second label value corresponding to each pixel.

[0043] S144, if the second marker value corresponding to the current pixel is the first category value, then the category corresponding to the current pixel is determined as a raised pixel; if the second marker value corresponding to the current pixel is the second category value, then the category corresponding to the current pixel is determined as a recessed pixel.

[0044] S145: Based on the category corresponding to each pixel, obtain several raised areas and several recessed areas.

[0045] The specific values ​​of the first and second category values ​​can be set by the implementer. For example, a first category value of 1 and a second category value of 0 are used to distinguish between raised and recessed pixels. The preset height threshold is 0.

[0046] If the height difference corresponding to the current pixel is greater than the preset height threshold, that is, the target depth value is greater than the actual depth value corresponding to the current pixel, that is, the distance between the current pixel and the point cloud data acquisition device is greater than the standard distance that should be reached, then the current pixel is a raised pixel.

[0047] If the height difference corresponding to the current pixel is less than the preset height threshold, that is, the target depth value is less than the actual depth value corresponding to the current pixel, that is, the distance between the current pixel and the point cloud data acquisition device is less than the standard distance that should be reached, then the current pixel is a concave pixel.

[0048] The first label value corresponding to all pixels is updated using the region growing algorithm to obtain the second label value corresponding to each pixel.

[0049] The region growing algorithm optimizes and clusters the first label values ​​of the initially classified pixels, eliminating isolated points or noise interference, and obtaining continuous and complete raised or recessed regions, thus making the division of raised and recessed regions more accurate. Those skilled in the art will recognize that the region growing algorithm in the prior art falls within the protection scope of this invention, and will not be elaborated upon here.

[0050] Furthermore, after obtaining several raised and recessed regions through a region growing algorithm, morphological operations are performed on all regions based on the second marker value corresponding to each pixel. This further optimizes the region contours, eliminates noise interference, fills in minor holes or breaks, making the region boundaries smoother and the structure more complete, ultimately updating the resulting raised and recessed regions. The morphological operations include erosion, dilation, opening, and closing operations. Those skilled in the art will recognize that any morphological operation in the prior art falls within the protection scope of this invention, and will not be elaborated upon here.

[0051] The above method determines the category of a pixel by using the height difference and a preset height threshold, and updates the first label value corresponding to all pixels by using a region growing algorithm. This avoids misclassification caused by errors or noise in a single pixel, and improves the accuracy and completeness of region division.

[0052] S2, based on the position of each raised region and each recessed region in the surface image, the raised volume corresponding to each raised region and the recessed volume corresponding to each recessed region, obtain the degree of mutual matching between each raised region and each recessed region, where the degree of mutual matching is a value greater than or equal to 0.

[0053] The degree of mutual matching is used to measure the feasibility of a raised area to fill a specific depressed area with material. The degree of mutual matching is a value greater than or equal to 0. Correspondingly, the higher the degree of mutual matching, the higher the feasibility of filling the corresponding depressed area with material through the corresponding raised area.

[0054] In one specific embodiment, S2 includes the following steps:

[0055] S21, based on the position of each raised region and each recessed region in the surface image, obtain the second distance between each raised region and each recessed region, the first normal direction corresponding to each raised region, and the second normal direction corresponding to each recessed region.

[0056] S22, based on the position and actual depth value of each pixel in the surface image, obtain the fluid resistance matrix corresponding to the platform to be ground.

[0057] S23, for any convex region, based on the second distance between the current convex region and each concave region, obtain several associated concave regions corresponding to the current convex region from all the concave regions.

[0058] S24, for any associated concave region corresponding to the current convex region, obtain the target matching degree between the current convex region and the current associated concave region based on the second distance between the current convex region and the current associated concave region, the convex volume and the first normal direction corresponding to the current convex region, and the concave volume and the second normal direction corresponding to the current associated concave region.

[0059] S25. Based on the fluid resistance matrix corresponding to the platform to be ground, obtain the minimum resistance path length between the current protruding area and the current associated concave area.

[0060] S26. Based on the target matching degree between the current convex region and the current associated concave region and the minimum resistance path length, obtain the mutual matching degree between the current convex region and the current associated concave region.

[0061] Specifically, based on the coordinates of the positions of each raised region and each recessed region in the surface image, the centroid coordinates of each raised region and each recessed region are obtained. The Euclidean distance between the corresponding centroid coordinates is used as the second distance between each raised region and each recessed region to evaluate the physical distance cost of material transfer.

[0062] The normal direction, through local surface fitting of pixels within a region, reflects the spatial orientation of the region and can provide geometric constraints for subsequent optimization of laser energy transmission direction, avoiding laser energy loss. Specifically, the first normal direction is the normal vector of the surface of the convex region, used to characterize the diffusion direction of the material, and the second normal direction is the normal vector of the surface of the concave region, used to characterize the receiving direction of the material.

[0063] Based on the depth information of the surface image and the material properties of the platform to be ground, the resistance value corresponding to each pixel is evaluated. The position of each pixel in the fluid resistance matrix is ​​determined according to the horizontal and vertical coordinates of each pixel in the acquired image. The fluid resistance matrix is ​​then formed by combining the resistance values ​​to characterize the resistance distribution encountered by material particles as they diffuse and move on the platform surface. This reflects the influence of the platform surface morphology on the diffusion and movement of materials, and provides a data basis for the resistance dimension when planning the optimal path.

[0064] By setting a distance-related threshold, and filtering out depressions that are less than the threshold from the current protruding area based on the second distance, it is possible to transfer and compensate for the related depressions based on the current protruding area.

[0065] By comprehensively considering spatial distance, regional volume, and normal direction, the degree of target matching between each raised region and each corresponding associated recessed region is measured, and the feasibility and efficiency of material transfer are fully evaluated.

[0066] Based on the fluid resistance matrix and the optimal path search algorithm, the minimum resistance path length from the center of the current convex region to the center of the current associated concave region is obtained. This is used to plan a more reasonable material transfer path, avoid obstacles or high resistance areas that may be encountered by traditional straight paths, and improve the efficiency of material transfer, that is, improve the grinding efficiency and grinding effect of the platform.

[0067] In one specific implementation, the target matching degree M between the current raised region and the currently associated recessed region is used as a basis. ij and minimum resistance path length L ij The degree of matching C between the current raised region and the currently associated recessed region is obtained. ij =M ij ×exp(-α×L ij / L0), where M ij L represents the target matching degree between the i-th protruding region and the corresponding j-th associated concave region. ij Let be the minimum resistance path length between the i-th raised region and the corresponding j-th associated recessed region, α be a preset resistance influence coefficient, L0 be the average of the minimum resistance path lengths corresponding to all raised regions, and exp() be the natural exponential function. α is used to adjust the influence intensity of the fluid resistance value on the feasibility of material transfer. The specific value can be set by the implementer according to the actual situation such as material characteristics, and the value range is usually [0.1, 5].

[0068] Based on the distance, volume information, and normal direction of the raised and recessed areas, and considering the physical distance cost and resistance distribution of material transfer, the material filling potential of each raised area to the associated recessed area is measured by multi-dimensional parameter fusion. This yields the target matching degree between each raised and recessed area, providing a data basis for planning more reasonable material transfer paths and control parameters to control the laser grinding process, thereby improving the platform's grinding efficiency and grinding effect.

[0069] In one specific embodiment, S21 includes the following steps:

[0070] S211, For any raised region, construct the covariance matrix corresponding to the current raised region based on the coordinates of all pixels in the current raised region in the surface image.

[0071] S212, perform eigenvalue decomposition on the covariance matrix corresponding to the current convex region, and determine the eigenvector corresponding to the minimum eigenvalue as the first normal direction corresponding to the current convex region.

[0072] In one specific embodiment, S21 includes the following steps:

[0073] S213, For any concave region, construct the covariance matrix corresponding to the current concave region based on the coordinates of all pixels in the current concave region in the surface image.

[0074] S214, perform eigenvalue decomposition on the covariance matrix corresponding to the current concave region, and determine the eigenvector corresponding to the minimum eigenvalue as the second normal direction corresponding to the current concave region.

[0075] For each set of pixels within a region, the corresponding covariance matrix represents the geometric characteristics of the data distribution. Specifically, the eigenvectors of the covariance matrix represent the main direction of the data distribution, while the eigenvalues ​​represent the degree of dispersion of the data in the corresponding direction.

[0076] For a point cloud region that is approximately planar, since the point cloud has the least dispersion in the normal direction, the direction of the eigenvector corresponding to the smallest eigenvalue of the covariance matrix of the convex region is consistent with the first normal direction of the convex region, and the direction of the eigenvector corresponding to the smallest eigenvalue of the covariance matrix of the concave region is consistent with the second normal direction of the concave region.

[0077] The above method uses the pixels of each region as a basis and obtains the normal direction of each region based on the covariance matrix method. It takes into account the spatial relationship of all points in each region, improves the data's noise resistance, improves the accuracy of obtaining the normal direction, and thus improves the accuracy of measuring the degree of target matching.

[0078] In one specific embodiment, S22 includes the following steps:

[0079] S221: Based on the position and actual depth value of each pixel, obtain the gradient value, Gaussian curvature and average curvature of each pixel.

[0080] S222: Obtain the terrain resistance corresponding to each pixel based on the gradient value corresponding to each pixel and the preset terrain resistance coefficient.

[0081] S223, based on the Gaussian curvature and average curvature of each pixel, obtain the curvature resistance corresponding to each pixel.

[0082] S224: Based on the terrain resistance and curvature resistance corresponding to each pixel, obtain the comprehensive resistance corresponding to each pixel.

[0083] S225: Smooth the overall resistance corresponding to each pixel according to the Gaussian kernel function to obtain the smoothed resistance corresponding to each pixel.

[0084] S226. Based on the smoothing resistance corresponding to all pixels and the coordinates of each pixel in the surface image, the fluid resistance matrix is ​​obtained.

[0085] Specifically, based on the two-dimensional coordinates of each pixel on the acquired image and the corresponding actual depth value, a three-dimensional coordinate system is constructed for each pixel. Based on these three-dimensional coordinates, the gradient value ▽z for each pixel is obtained to reflect the surface slope, the Gaussian slope K for each pixel is obtained to reflect the surface curvature type, and the average curvature H for each pixel is obtained to reflect the degree of surface curvature. Those skilled in the art will recognize that the calculation methods for gradient values, Gaussian curvature, and average curvature in the prior art fall within the scope of this invention, and will not be elaborated upon here. Specifically, when K > 0, the surface curvature type is elliptic, and material particles flowing in the corresponding region will experience a convergence effect; when K < 0, the surface curvature type is hyperbolic, and material particles flowing in the corresponding region will experience anisotropic drag; when K = 0, the surface curvature type is parabolic, and material particles flowing in the corresponding region will experience unidirectional drag or drag without curvature.

[0086] Based on the gradient value ▽z corresponding to each pixel and the preset terrain resistance coefficient μ, the terrain resistance corresponding to each pixel is obtained. For example, the terrain resistance R1 = μ × ‖▽z‖ × cosθ1, which is used to characterize the gravitational component of the material particles on the inclined plane. Here, θ1 is the angle between the direction of gravity and the normal direction corresponding to the pixel, ‖▽z‖ is the magnitude of the gradient, and the specific value of the terrain resistance coefficient μ is set by the implementer according to the material properties.

[0087] Based on the Gaussian curvature K and average curvature H corresponding to each pixel, the curvature resistance for each pixel is obtained. For example, the curvature resistance R2 = β × |K| + γ × |H|, which is used to characterize the resistance to material transfer / filling in protruding / depressed areas. The specific values ​​of the weights β and γ can be set by the implementer according to the actual situation. For example, β = γ = 0.5.

[0088] Based on the terrain resistance R1 and curvature resistance R2 corresponding to each pixel, the overall resistance for each pixel is obtained. For example, the overall resistance R3 = w1 × R1 + w2 × R2, where the specific values ​​of the weights w1 and w2 can be set by the implementer according to the actual situation. For example, w1 = 0.7, w2 = 0.3.

[0089] The position of each pixel in the drag matrix is ​​determined by the horizontal and vertical coordinates of the acquired image. The combined drag matrix is ​​then formed by combining the comprehensive drag. The drag matrix is ​​then smoothed using a Gaussian kernel function to obtain a smoothed drag matrix composed of smoothed drag. This process eliminates abrupt changes in drag caused by local noise and improves the accuracy of the fluid drag matrix construction.

[0090] The above analysis, based on the position and actual depth of each pixel, analyzes the surface slope, surface curvature type and surface curvature represented by each pixel. Then, it combines the terrain resistance and curvature resistance corresponding to each pixel to obtain the comprehensive resistance. After further smoothing, it obtains the fluid resistance matrix, which improves the accuracy of the fluid resistance matrix construction and thus improves the accuracy of the measurement of the degree of target matching.

[0091] In one specific embodiment, S23 includes the following steps:

[0092] S231, for any convex region, obtain the equivalent diameter corresponding to the current convex region based on the area of ​​the connected domain corresponding to the current convex region.

[0093] S232, based on the maximum convex volume among all convex volumes, the maximum concave volume among all concave volumes, and the equivalent diameter corresponding to the current convex region, obtain the associated search diameter corresponding to the current convex region.

[0094] S233, for any concave region, if the second distance between the current convex region and the current concave region is less than the associated search diameter corresponding to the current convex region, then the current concave region is determined as the associated concave region corresponding to the current convex region.

[0095] S234, traverse all the concave regions and obtain all the associated concave regions corresponding to the current convex region.

[0096] In this process, the area of ​​the connected domain of the convex region is equivalent to the area of ​​a circle, and the equivalent diameter is calculated. For example, the equivalent diameter D1 = 2 × (A / π). 0.5 The equivalent diameter, as the geometric dimension of the raised region, is used to evaluate the effective coverage area of ​​the material sprayed after the raised region is heated by laser. Here, A is the area of ​​the connected region of the raised region, which is the total area of ​​all pixels in the raised region, and π is pi.

[0097] The largest bulge volume V among all bulge volumes max1 The largest depression volume V among all depression volumes max2 The equivalent diameter D1 corresponding to the current protrusion region is dynamically set to determine the correlation search diameter. This avoids the problem of "excessively large search range for small protrusions" or "insufficient search range for large protrusions" caused by a fixed threshold, thereby balancing computational efficiency and matching completeness and reducing invalid searches. For example, the correlation search diameter D2 = w3 × D1 + w4 × ((V max1 +V max2 ) / 2) 1 / 3 Among them, D1 can dominate the close-range search, ((V max1 +V max2 ) / 2) 1 / 3This allows for the filling of distant, extremely concave areas, such as large-volume concave areas that may require material from distant convex areas. The specific values ​​of weights w3 and w4 can be set by the implementer based on the actual situation. For example, w3=0.4, w4=0.6.

[0098] The above-mentioned method dynamically sets the associated search diameter by combining the largest convex volume among all convex volumes, the largest concave volume among all concave volumes, and the equivalent diameter corresponding to the current convex region. This is used to filter out all associated concave regions corresponding to the convex region, balancing computational efficiency and matching completeness, and reducing the workload of invalid searches.

[0099] In one specific embodiment, S24 includes the following steps:

[0100] S241, based on the preset distance attenuation coefficient and the second distance between the current protruding region and the current associated concave region, obtain the distance matching degree between the current protruding region and the current associated concave region.

[0101] S242, based on the convex volume corresponding to the current convex region and the concave volume corresponding to the current associated concave region, obtain the volume matching degree between the current convex region and the current associated concave region.

[0102] S243, based on the angle between the first normal direction corresponding to the current protruding region and the second normal direction corresponding to the current associated concave region, obtain the degree of directional consistency between the current protruding region and the current associated concave region.

[0103] S244, based on the distance matching degree, volume matching degree, and orientation consistency degree between the current raised region and the current associated recessed region, obtain the target matching degree between the current raised region and the current associated recessed region.

[0104] Among these, the closer the second distance, the higher the degree of distance matching between the corresponding regions. For example, the degree of distance matching E between the i-th protruding region and the corresponding j-th associated concave region. ij =exp(-λ×d ij ), used to characterize the exponential decay of material transfer efficiency with increasing distance, avoiding energy loss caused by long-distance transfer. Wherein, d ij λ is the second distance between the i-th protruding region and the corresponding j-th associated concave region. λ is a preset distance attenuation coefficient. The specific value of λ can be set by the implementer according to the material characteristics. The value range can be [0.5, 2]. Specifically, the larger the value of λ, the faster the long-distance matching degree attenuates, which is suitable for high viscosity materials such as ceramics. The smaller the value of λ, the more distant the matching is allowed, which is suitable for low viscosity materials such as aluminum alloys.

[0105] The closer the volumes of the protrusions and depressions are, the higher the material transfer efficiency, and the higher the degree of volume matching between the corresponding protrusion regions and associated depression regions. For example, the degree of volume matching F between the i-th protrusion region and the corresponding j-th associated depression region... ij =exp(-|(V i / V ij )-1| σ ), where V i V is the volume of the convex region corresponding to the i-th convex region. ij σ is the depression volume of the j-th associated depression corresponding to the i-th protruding region. σ is an adjustment index used to control the penalty intensity for volume mismatch, that is, the imbalance between material supply and demand. The larger the value of σ, the higher the sensitivity to volume difference. The specific value of σ can be set by the implementer according to the actual situation. For example, σ=2.

[0106] When the first normal direction corresponding to the i-th protruding region is consistent with the second normal direction corresponding to the j-th associated concave region, that is, when the angle θ between the first normal direction corresponding to the i-th protruding region and the second normal direction corresponding to the j-th associated concave region is consistent, ij When G = 0°, it indicates that the material melted by laser heating of the i-th protrusion region has a high filling effectiveness for the material of the j-th associated depression region, and the corresponding G... ij =1. When the first normal direction corresponding to the i-th convex region is opposite to the second normal direction corresponding to the j-th associated concave region, that is, θ ij When the angle is 180°, it indicates that the material melted by laser heating of the i-th protruding region has low effectiveness in filling the material of the j-th associated recessed region, and the corresponding G... ij =0. Therefore, the directional consistency G between the first normal direction corresponding to the i-th protruding region and the corresponding j-th associated concave region is obtained. ij =(1+cosθ ij ) / 2.

[0107] Based on the degree of distance matching, volume matching, and directional consistency between the current raised region and the currently associated recessed region, a weighted sum is performed using the weights corresponding to each of these factors to obtain the target matching degree between the current raised region and the currently associated recessed region. This summation characterizes the material transfer adaptability between the raised and recessed regions. The specific values ​​of the weights corresponding to the degree of distance matching, volume matching, and directional consistency can be set by the implementer according to the actual situation; for example, the corresponding weights could be 0.3, 0.4, and 0.3, respectively.

[0108] The above methods are used to analyze the physical law of material transfer efficiency decay with distance by obtaining the degree of distance matching, to analyze the material supply and demand balance by obtaining the degree of volume matching, and to analyze the influence of the normal direction on the laser energy transmission efficiency and effectiveness by obtaining the degree of directional consistency. The target matching degree is obtained by combining the three dimensions, which characterizes the material transfer adaptability between the raised and recessed areas. This serves as the data basis for planning a more reasonable material transfer path and control parameters, so as to control the laser grinding process and improve the grinding efficiency and grinding effect of the platform.

[0109] S3. Based on the degree of matching between each raised area and each recessed area, the grinding path of the laser beam and the corresponding timing control data are obtained. The timing control data includes the state mode of the laser beam at each unit time point when grinding the platform to be ground according to the grinding path. The state mode includes heating mode and cooling mode.

[0110] In this embodiment, the laser polisher operates cyclically according to a heating and cooling state model. When the laser beam acts on the raised area in heating mode, the material temperature is rapidly increased through high energy density. When the temperature reaches the melting or vaporization point of the material, the raised part is evaporated or melted, and the molten material is transferred to the recessed area to fill it through flow. After heating, the model switches to cooling mode, for example, by introducing an inert gas, which allows the molten material to solidify rapidly, avoiding excessive melting or thermal deformation of the platform due to continuous heating. At the same time, thermal stress is generated in the material during heating (expansion) and cooling (contraction). Alternating operation uniformizes the stress distribution through periodic temperature changes, avoiding material cracking or delamination caused by a single drastic heating / cooling. Thus, dynamic thermal management balances material removal efficiency and morphology control, thereby improving the polishing effect of the laser beam on the platform to be polished.

[0111] In one specific embodiment, S3 includes the following steps:

[0112] S31. Sort all the mutual matching degrees in descending order, and obtain the priority matrix based on the sorting result. In the priority matrix, the row priority is the convex region arranged in descending order according to the corresponding maximum mutual matching degree, and the column priority is the concave region arranged in descending order according to the sum of the corresponding maximum mutual matching degrees. The value of each element in the priority matrix is ​​the mutual matching degree between the convex region of the corresponding row and the concave region of the corresponding column.

[0113] S32, based on the priority matrix, use the optimal path algorithm to optimize the movement path of the laser beam between each raised area and each recessed area, and obtain the grinding path corresponding to the laser beam in the laser polisher.

[0114] S33, based on the protrusion volume corresponding to each protrusion area, the preset material properties, the preset laser power, and the preset efficiency of the laser polisher, the total heating time of the laser beam for each protrusion area is obtained.

[0115] S34. Based on the order of each raised area in the grinding path, the total heating time and preset cooling time of each raised area, the state mode of the laser beam at each unit time point is obtained.

[0116] In the priority matrix, the row priority is determined by the convex regions being sorted in descending order of their maximum mutual matching degree, and the column priority is determined by the concave regions being sorted in descending order of their maximum mutual matching degree sum. The element value Y in the u-th row and v-th column of the priority matrix is... uv This represents the degree of matching between the raised regions of the corresponding row and the recessed regions of the corresponding column. If the recessed region of the column corresponding to an element value is not a related recessed region of the raised region of the corresponding row, then the element value is 0. The number of rows in the priority matrix is ​​the same as the total number of raised regions, and the number of columns in the priority matrix is ​​the same as the total number of recessed regions.

[0117] Based on the priority matrix, an optimal path algorithm is used to optimize the movement path of the laser beam between each raised and recessed region. For example, an improved Traveling Salesman Problem algorithm is used to generate the optimal movement path: min(Σ u,v (Z uv ×x uv )+Σ u,v (x uv / Y uv ), where Z uv x is equal to the second distance between the raised region in row u and the recessed region in column v. uv Path selection variables are used. By introducing the degree of mutual matching as a heuristic factor, the path order is dynamically adjusted to obtain the moving path with the highest global mutual matching degree, which serves as the grinding path corresponding to the laser beam in the laser polisher. Those skilled in the art will understand that any optimal path algorithm in the prior art falls within the protection scope of this invention, and will not be elaborated further here.

[0118] In one specific embodiment, the preset material properties include material density ρ, specific heat capacity c, melting temperature difference ΔT, and latent heat of vaporization J, based on the protrusion volume V corresponding to the i-th protrusion region. iBased on preset material properties, preset laser power P, and preset efficiency η of the laser polisher, the total heating time corresponding to the i-th raised region is obtained. By limiting the total heating time, the material in the raised region can be effectively transferred to the associated recessed region while preventing platform substrate deformation due to excessive energy and preventing excessively high bulge height in the heated region, thereby improving the platform polishing effect. For example, the total heating time Ψ corresponding to the i-th raised region... i =(ρ×V i ×c×ΔT+ρ×V i ×J) / (P×η). The specific values ​​of the preset laser power P and the preset efficiency η of the laser polisher are set by the implementer according to the actual situation. For example, η=0.5.

[0119] Arrange the raised regions in the order of the grinding path and cycle through them according to the heating and cooling state model. This yields the state mode of the laser beam at each unit time point when grinding the platform according to the grinding path. For example, a value of 1 represents the heating mode, and a value of 0 represents the cooling mode. If the total heating time Ψ1 corresponding to the first raised region in the grinding path includes N1 unit time points, and the preset cooling time includes N0 unit time points, then the timing control sub-data SX1 corresponding to grinding the first raised region is {(t1, 1), (t2, 1), ..., (t... N1 ,1),(t N1+1 ,0),……,(t N1+N0 The total heating time Ψ corresponding to the i-th raised area in the grinding path is 0). i Including Ni unit time points, the timing control sub-data SX corresponding to the i-th protrusion region. i For {(t) δ+1 ,1),(t δ+2 ,1),……,(t δ+Ni ,1),(t δ+Ni+1 ,0),……,(t δ+Ni+N0 ,0)}, where δ is the sum of the total number of unit time points of the heating mode and the total number of unit time points of the cooling mode corresponding to the first i-1 protruding regions.

[0120] The above-mentioned method transforms the multi-region matching problem into a computable structured model by forming a priority matrix based on the degree of mutual matching. It focuses on high-value regions, reduces ineffective processing, and improves overall grinding efficiency. Based on the priority matrix, the path order is dynamically adjusted to obtain the grinding path with the highest degree of global mutual matching. The total heating time of the laser beam for each protruding region is obtained by using the protrusion volume and preset parameters. Thus, each protruding region is arranged in the order of the grinding path and works in a cycle according to the heating and cooling state model. This yields the state mode of the laser beam at each unit time point when grinding the platform to be ground according to the grinding path. This method solves the problems of low efficiency of multi-region collaborative processing and difficulty in controlling thermal deformation in laser grinding, thereby improving grinding efficiency and grinding effect.

[0121] S4, Grind the platform to be ground according to the grinding path and control data.

[0122] The laser head of the laser polisher is driven to move along the polishing path based on the polishing path and control data, ensuring coverage of all raised areas to be heated. The heating and cooling states are switched according to the control data. Specifically, the laser beam is controlled to heat the area at specific times during heating mode, and the laser head of the laser polisher is controlled to move along the polishing path to the next raised area to be heated at specific times during cooling mode. This balances material removal and thermal deformation control, improving the flatness of the polishing platform.

[0123] In one specific embodiment, the control method further includes:

[0124] S5, after grinding the platform to be ground according to the grinding path and control data, obtains the updated surface image and updated point cloud data corresponding to the platform to be ground.

[0125] S6. Obtain the smoothness of the platform to be ground based on the point cloud data. If the smoothness does not meet the preset stopping condition, repeat step S1 until the new smoothness of the platform to be ground meets the preset stopping condition and the target platform is obtained.

[0126] The smoothness of the platform to be ground is evaluated based on surface images and point cloud data. For example, the surface images and point cloud data are input into a preset roughness evaluation model to obtain the smoothness of the platform. The preset roughness evaluation model can be constructed based on a convolutional neural network model and trained using labeled surface images and point cloud data to improve the accuracy of the smoothness assessment. Those skilled in the art will recognize that any existing roughness evaluation model and its corresponding training method fall within the scope of this invention, and will not be elaborated upon here.

[0127] Through multiple rounds of iterative control and grinding, the raised areas and the material bulges caused by the previous heating can be removed during the next controlled grinding, and the recessed areas can be filled more finely, thus gradually improving the smoothness of the grinding platform and enhancing the grinding effect.

[0128] The above describes a process that compares the target depth value with the actual depth value in the point cloud data, extracts raised and recessed areas from the surface image, and calculates their volume. This quantifies the spatial distribution and material quantity differences of various types of areas in the grinding platform, providing a geometric and physical basis for subsequent material transfer and avoiding precision loss caused by blind processing. By integrating multi-dimensional parameters such as the positional relationship and volume matching degree of raised and recessed areas, the degree of matching between them is analyzed. This quantifies the feasibility and efficiency of material transfer between raised and recessed areas under heating, facilitating the priority matching of high-value area pairs, improving the rationality of the overall processing plan and resource utilization. A grinding path is generated based on the degree of matching, and the heating / cooling sequence is allocated based on the thermal properties of the material. The planned grinding path and timing control data are then converted into execution instructions for the laser grinding equipment, making the laser beam movement path more rational and the energy output more precise, reducing idle travel and thermal deformation. At the same time, periodic cooling controls thermal stress, improving grinding efficiency and the grinding effect of the platform.

[0129] Example 2

[0130] This second embodiment provides a laser polishing control system for a platform, such as... Figure 2 As shown, the laser polishing control system for the platform includes:

[0131] The region extraction module 21 is used to obtain, based on the target depth value, surface image and point cloud data corresponding to the platform to be ground, a number of raised regions, a number of recessed regions, the raised volume corresponding to each raised region and the recessed volume corresponding to each recessed region in the surface image. The point cloud data includes the actual depth value corresponding to each pixel in the surface image.

[0132] The mutual matching degree acquisition module 22 is used to obtain the mutual matching degree between each convex region and each concave region based on the position of each convex region and each concave region in the surface image, the convex volume corresponding to each convex region and the concave volume corresponding to each concave region, wherein the mutual matching degree is a value greater than or equal to 0.

[0133] The control data acquisition module 23 is used to acquire the grinding path of the laser beam and the corresponding timing control data according to the degree of matching between each raised area and each recessed area. The timing control data includes the state mode of the laser beam at each unit time point when grinding the platform to be ground according to the grinding path. The state mode includes heating mode and cooling mode.

[0134] The grinding control module 24 is used to grind the grinding platform according to the grinding path and control data.

[0135] In one specific embodiment, the region extraction module 21 includes:

[0136] The first distance acquisition submodule is used to acquire the target platform thickness, surface image, point cloud data, and the first distance between the bottom surface of the platform to be ground and the point cloud data acquisition device. The actual depth value corresponding to each pixel in the point cloud data is the distance between the corresponding position of each pixel on the platform to be ground and the point cloud data acquisition device.

[0137] The target depth value acquisition submodule is used to determine the difference between the first distance and the thickness of the target platform as the target depth value corresponding to the platform to be ground.

[0138] The height difference acquisition submodule is used to determine the height difference for each pixel by the difference between the target depth value and the actual depth value corresponding to each pixel.

[0139] The region extraction submodule is used to obtain several raised regions and several recessed regions in the surface image based on the comparison result between the height difference corresponding to each pixel and the preset height threshold.

[0140] The bump volume acquisition submodule is used to obtain the bump volume corresponding to each bump region based on the height difference of each pixel in each bump region.

[0141] The concave volume acquisition submodule is used to obtain the concave volume of each concave region based on the height difference of each pixel in each concave region.

[0142] In one specific implementation, the region extraction submodule includes:

[0143] The first comparison unit is used to determine the first label value corresponding to any pixel as the first category value if the height difference corresponding to the current pixel is greater than a preset height threshold, wherein the category of the pixel corresponding to the first category value is a raised pixel.

[0144] The second comparison unit is used to determine the first marker value corresponding to the pixel as the second category value if the height difference corresponding to the current pixel is less than a preset height threshold, wherein the category of the pixel corresponding to the second category value is a concave pixel.

[0145] The marker value update unit is used to update the first marker value corresponding to all pixels according to the region growing algorithm, and obtain the second marker value corresponding to each pixel.

[0146] The pixel classification unit is used to determine the category of the current pixel as a raised pixel if the second label value corresponding to the current pixel is the first category value, and to determine the category of the current pixel as a recessed pixel if the second label value corresponding to the current pixel is the second category value.

[0147] The region extraction unit is used to obtain several raised regions and several recessed regions according to the category corresponding to each pixel.

[0148] In one specific embodiment, the mutual matching degree acquisition module 22 includes:

[0149] The data analysis submodule is used to obtain the second distance between each raised area and each recessed area, the first normal direction corresponding to each raised area, and the second normal direction corresponding to each recessed area based on the position of each raised area and each recessed area in the surface image.

[0150] The fluid resistance matrix acquisition submodule is used to obtain the fluid resistance matrix corresponding to the platform to be ground based on the position and actual depth value of each pixel in the surface image.

[0151] The associated concave region filtering submodule is used to obtain several associated concave regions corresponding to the current convex region from all concave regions for any convex region, based on the second distance between the current convex region and each concave region.

[0152] The target matching degree acquisition submodule is used to obtain the target matching degree between the current convex region and the current associated concave region for any associated concave region corresponding to the current convex region, based on the second distance between the current convex region and the current associated concave region, the convex volume and the first normal direction corresponding to the current convex region, and the concave volume and the second normal direction corresponding to the current associated concave region.

[0153] The minimum resistance path length acquisition submodule is used to obtain the minimum resistance path length between the current protruding region and the current associated concave region based on the fluid resistance matrix corresponding to the platform to be ground.

[0154] The mutual matching degree acquisition submodule is used to obtain the mutual matching degree between the current convex region and the current associated concave region based on the target matching degree between the current convex region and the current associated concave region and the minimum resistance path length.

[0155] In one specific implementation, the data analysis submodule includes:

[0156] The covariance matrix construction unit is used to construct the covariance matrix corresponding to any convex region based on the coordinates of all pixels in the surface image within the current convex region.

[0157] The first normal direction determination unit is used to perform eigenvalue decomposition on the covariance matrix corresponding to the current convex region, and determine the eigenvector corresponding to the minimum eigenvalue as the first normal direction corresponding to the current convex region.

[0158] In one specific embodiment, the fluid resistance matrix acquisition submodule includes:

[0159] The curvature acquisition unit is used to obtain the gradient value, Gaussian curvature and average curvature of each pixel based on the position and actual depth value of each pixel.

[0160] The terrain resistance acquisition unit is used to obtain the terrain resistance corresponding to each pixel based on the gradient value corresponding to each pixel and the preset terrain resistance coefficient.

[0161] The curvature resistance acquisition unit is used to obtain the curvature resistance corresponding to each pixel based on the Gaussian curvature and average curvature corresponding to each pixel.

[0162] The comprehensive resistance acquisition unit is used to obtain the comprehensive resistance corresponding to each pixel based on the terrain resistance and curvature resistance corresponding to each pixel.

[0163] The smoothing resistance acquisition unit is used to smooth the comprehensive resistance corresponding to each pixel according to the Gaussian kernel function, and obtain the smoothing resistance corresponding to each pixel.

[0164] The fluid resistance matrix acquisition unit is used to obtain the fluid resistance matrix based on the smoothing resistance corresponding to all pixels and the coordinates of each pixel in the surface image.

[0165] In one specific implementation, the associated recessed area filtering submodule includes:

[0166] The equivalent diameter acquisition unit is used to obtain the equivalent diameter of any convex region based on the area of ​​the connected domain corresponding to the current convex region.

[0167] The associated search diameter acquisition unit is used to obtain the associated search diameter corresponding to the current convex region based on the largest convex volume among all convex volumes, the largest concave volume among all concave volumes, and the equivalent diameter corresponding to the current convex region.

[0168] The associated concave region filtering unit is used to determine the current concave region as the associated concave region corresponding to the current convex region if, for any concave region, the second distance between the current convex region and the current concave region is less than the associated search diameter corresponding to the current convex region.

[0169] The region traversal unit is used to traverse all the concave regions and obtain all the associated concave regions corresponding to the current convex region.

[0170] In one specific implementation, the target matching degree acquisition submodule includes:

[0171] The distance matching degree acquisition unit is used to obtain the distance matching degree between the current convex region and the current associated concave region based on the preset distance attenuation coefficient and the second distance between the current convex region and the current associated concave region.

[0172] The volume matching degree acquisition unit is used to obtain the volume matching degree between the current convex region and the current associated concave region based on the convex volume corresponding to the current convex region and the concave volume corresponding to the current associated concave region.

[0173] The orientation consistency degree acquisition unit is used to obtain the orientation consistency degree between the current convex region and the current associated concave region based on the angle between the first normal direction corresponding to the current convex region and the second normal direction corresponding to the current associated concave region.

[0174] The target matching degree acquisition unit is used to obtain the target matching degree between the current convex region and the current associated concave region based on the distance matching degree, volume matching degree, and orientation consistency degree between the current convex region and the current associated concave region.

[0175] In one specific embodiment, the control data acquisition module 23 includes:

[0176] The priority matrix acquisition submodule is used to sort all the mutual matching degrees in descending order and obtain the priority matrix based on the sorting result. In the priority matrix, the row priority is the convex region arranged in descending order according to the corresponding maximum mutual matching degree, and the column priority is the concave region arranged in descending order according to the sum of the corresponding maximum mutual matching degrees. The value of each element in the priority matrix is ​​the mutual matching degree between the convex region of the corresponding row and the concave region of the corresponding column.

[0177] The grinding path acquisition submodule is used to optimize the movement path of the laser beam between each raised area and each recessed area using the optimal path algorithm based on the priority matrix, thereby obtaining the grinding path corresponding to the laser beam in the laser grinder.

[0178] The total heating time acquisition submodule is used to obtain the total heating time of the laser beam for each raised area based on the raised volume corresponding to each raised area, the preset material properties, the preset laser power, and the preset efficiency of the laser polisher.

[0179] The state mode acquisition submodule is used to obtain the state mode of the laser beam at each unit time point based on the order of each protrusion in the grinding path, the total heating time of each protrusion, and the preset cooling time.

[0180] It should be noted that the information interaction and execution process between the above modules are based on the same concept as the method embodiments of the present invention. For details on their specific functions and technical effects, please refer to the method embodiments section, which will not be repeated here.

[0181] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A laser polishing control method for a platform, characterized in that, The control method includes the following steps: S1. Based on the target depth value, surface image and point cloud data corresponding to the platform to be ground, obtain a number of raised areas, a number of recessed areas, the raised volume corresponding to each raised area and the recessed volume corresponding to each recessed area in the surface image. The point cloud data includes the actual depth value corresponding to each pixel in the surface image. S2, based on the position of each raised area and each recessed area in the surface image, the raised volume corresponding to each raised area and the recessed volume corresponding to each recessed area, the degree of mutual matching between each raised area and each recessed area is obtained, wherein the degree of mutual matching is a value greater than or equal to 0. S3. Based on the degree of matching between each raised area and each recessed area, the grinding path of the laser beam and the corresponding timing control data are obtained. The timing control data includes the state mode of the laser beam at each unit time point when grinding the platform to be ground according to the grinding path. The state mode includes heating mode and cooling mode. S4, Grind the platform to be ground according to the grinding path and the control data.

2. The laser polishing control method for a platform according to claim 1, characterized in that, S1 includes the following steps: S11, obtain the target platform thickness, surface image, point cloud data, and the first distance between the bottom surface of the platform to be ground and the point cloud data acquisition device corresponding to the platform to be ground. The actual depth value corresponding to each pixel in the point cloud data is the distance between the corresponding position of each pixel on the platform to be ground and the point cloud data acquisition device. S12, the difference between the first distance and the thickness of the target platform is determined as the target depth value corresponding to the platform to be ground; S13, the difference between the target depth value and the actual depth value corresponding to each pixel is determined as the height difference corresponding to each pixel; S14. Based on the comparison result between the height difference corresponding to each pixel and the preset height threshold, obtain a number of raised areas and a number of recessed areas in the surface image. S15, based on the height difference of each pixel in each raised region, obtain the raised volume corresponding to each raised region; S16, obtain the depression volume corresponding to each depression region based on the height difference corresponding to each pixel in each depression region.

3. The laser polishing control method for a platform according to claim 2, characterized in that, S14 includes the following steps: S141, for any pixel, if the height difference corresponding to the current pixel is greater than the preset height threshold, then the first marker value corresponding to the pixel is determined as the first category value, wherein the category of the pixel corresponding to the first category value is a raised pixel. S142, if the height difference corresponding to the current pixel is less than the preset height threshold, then the first marker value corresponding to the pixel is determined as the second category value, wherein the category of the pixel corresponding to the second category value is a concave pixel. S143, Update the first label value corresponding to all pixels according to the region growing algorithm to obtain the second label value corresponding to each pixel; S144, if the second marker value corresponding to the current pixel is the first category value, then the category corresponding to the current pixel is determined as a raised pixel; if the second marker value corresponding to the current pixel is the second category value, then the category corresponding to the current pixel is determined as a recessed pixel. S145: Based on the category corresponding to each pixel, obtain several raised areas and several recessed areas.

4. The laser polishing control method for a platform according to claim 1, characterized in that, S2 includes the following steps: S21, based on the position of each raised area and each recessed area in the surface image, obtain the second distance between each raised area and each recessed area, the first normal direction corresponding to each raised area and the second normal direction corresponding to each recessed area, wherein the first normal direction and the second normal direction are used to quantify the spatial orientation consistency between the raised area and the recessed area. S22, based on the position and actual depth value of each pixel in the surface image, obtain the fluid resistance matrix corresponding to the platform to be ground; S23, for any raised area, based on the second distance between the current raised area and each concave area, obtain several associated concave areas corresponding to the current raised area from all the concave areas; S24, for any associated concave region corresponding to the current protruding region, the target matching degree between the current protruding region and the current associated concave region is obtained based on the second distance between the current protruding region and the current associated concave region, the protruding volume and the first normal direction corresponding to the current protruding region, and the concave volume and the second normal direction corresponding to the current associated concave region. S25, based on the fluid resistance matrix corresponding to the platform to be ground, obtain the minimum resistance path length between the current protruding area and the current associated concave area; S26. Based on the target matching degree between the current convex region and the current associated concave region and the minimum resistance path length, obtain the mutual matching degree between the current convex region and the current associated concave region.

5. The laser polishing control method for a platform according to claim 4, characterized in that, S21 includes the following steps: S211, For any raised region, construct the covariance matrix corresponding to the current raised region based on the coordinates of all pixels in the current raised region in the surface image; S212, perform eigenvalue decomposition on the covariance matrix corresponding to the current convex region, and determine the eigenvector corresponding to the minimum eigenvalue as the first normal direction corresponding to the current convex region.

6. The laser polishing control method for a platform according to claim 4, characterized in that, S22 includes the following steps: S221, based on the position and actual depth value of each pixel, obtain the gradient value, Gaussian curvature and average curvature of each pixel; S222, based on the gradient value corresponding to each pixel and the preset terrain resistance coefficient, the terrain resistance corresponding to each pixel is obtained. The terrain resistance is used to characterize the gravitational component of the material particles on the inclined plane. The terrain resistance R1 = μ × ‖▽z‖ × cosθ1, where θ1 is the angle between the gravity direction and the normal direction corresponding to the pixel, ▽z is the gradient value corresponding to the pixel, ‖▽z‖ is the magnitude of the gradient, and μ is the preset terrain resistance coefficient. S223, based on the Gaussian curvature and average curvature of each pixel, obtain the curvature resistance corresponding to each pixel, wherein the curvature resistance is used to characterize the resistance to material transfer / filling in the protruding / depressed area, and the curvature resistance R2=β×|K|+γ×|H|, where K is the Gaussian curvature K corresponding to the pixel, H is the average curvature corresponding to the pixel, β is the weight corresponding to the Gaussian curvature, and γ is the weight corresponding to the average curvature weight; S224, based on the terrain resistance and curvature resistance corresponding to each pixel, obtain the comprehensive resistance corresponding to each pixel; S225, smooth the overall resistance corresponding to each pixel according to the Gaussian kernel function to obtain the smoothed resistance corresponding to each pixel; S226. Based on the smoothing resistance corresponding to all pixels and the coordinates of each pixel in the surface image, the fluid resistance matrix is ​​obtained.

7. The laser polishing control method for a platform according to claim 4, characterized in that, S23 includes the following steps: S231, For any convex region, obtain the equivalent diameter corresponding to the current convex region based on the area of ​​the connected region corresponding to the current convex region; S232, based on the maximum convex volume among all convex volumes, the maximum concave volume among all concave volumes, and the equivalent diameter corresponding to the current convex region, obtain the associated search diameter corresponding to the current convex region; S233, for any concave region, if the second distance between the current convex region and the current concave region is less than the associated search diameter corresponding to the current convex region, then the current concave region is determined as the associated concave region corresponding to the current convex region. S234, traverse all the concave regions and obtain all the associated concave regions corresponding to the current convex region.

8. The laser polishing control method for a platform according to claim 4, characterized in that, S24 includes the following steps: S241, based on the preset distance attenuation coefficient and the second distance between the current protruding region and the current associated concave region, obtain the distance matching degree between the current protruding region and the current associated concave region; S242, Based on the protrusion volume corresponding to the current protrusion region and the depression volume corresponding to the current associated depression region, obtain the volume matching degree between the current protrusion region and the current associated depression region; S243, based on the angle between the first normal direction corresponding to the current protruding region and the second normal direction corresponding to the current associated concave region, the degree of directional consistency between the current protruding region and the current associated concave region is obtained; S244, based on the distance matching degree, volume matching degree, and orientation consistency degree between the current raised region and the current associated recessed region, obtain the target matching degree between the current raised region and the current associated recessed region.

9. The laser polishing control method for a platform according to claim 1, characterized in that, S3 includes the following steps: S31. Sort all the mutual matching degrees in descending order, and obtain the priority matrix based on the sorting result. In the priority matrix, the row priority is the convex region arranged in descending order according to the corresponding maximum mutual matching degree, and the column priority is the concave region arranged in descending order according to the sum of the corresponding maximum mutual matching degree. The value of each element in the priority matrix is ​​the mutual matching degree between the convex region of the corresponding row and the concave region of the corresponding column. S32, according to the priority matrix, the optimal path algorithm is used to optimize the movement path of the laser beam between each raised area and each recessed area to obtain the grinding path corresponding to the laser beam in the laser polisher. S33, based on the protrusion volume corresponding to each protrusion area, the preset material properties, the preset laser power, and the preset efficiency of the laser polisher, the total heating time of the laser beam for each protrusion area is obtained; S34. Based on the order of each protruding region in the grinding path, the total heating time and preset cooling time corresponding to each protruding region, the state mode of the laser beam at each unit time point is obtained.

10. A laser polishing control system for a platform, characterized in that, The control system includes: The region extraction module is used to obtain, based on the target depth value, surface image and point cloud data corresponding to the platform to be ground, a number of raised regions, a number of recessed regions, the raised volume corresponding to each raised region and the recessed volume corresponding to each recessed region in the surface image. The point cloud data includes the actual depth value corresponding to each pixel in the surface image. The mutual matching degree acquisition module is used to obtain the mutual matching degree between each protruding region and each concave region based on the position of each protruding region and each concave region in the surface image, the protruding volume corresponding to each protruding region and the concave volume corresponding to each concave region, wherein the mutual matching degree is a value greater than or equal to 0. The control data acquisition module is used to acquire the polishing path of the laser beam and the corresponding timing control data according to the degree of matching between each raised area and each recessed area. The timing control data includes the state mode of the laser beam at each unit time point when polishing the platform to be polished according to the polishing path. The state mode includes a heating mode and a cooling mode. The grinding control module is used to grind the platform to be ground according to the grinding path and the control data.

Images