Laser grinding control method and system for platform

Abstract

The invention relates to the technical field of laser control, in particular to a laser grinding control method and system for a platform, which comprises the following steps: comparing a target depth value with an actual depth value, extracting a convex region and a concave region in combination with a surface image, calculating the volume, quantifying the space distribution and material quantity difference of various types of regions in a to-be-ground platform, and calculating the size of the to-be-ground platform. The precision loss caused by blind machining is avoided, the mutual matching degree between the convex area and the concave area is analyzed by integrating parameters such as the position relation and the volume matching degree between the areas, so that the feasibility and efficiency of transferring of the material between the convex area and the concave area during heating are quantitatively represented, and a grinding path is generated based on the mutual matching degree; heating and cooling time sequences are distributed in combination with thermal characteristics of materials and are converted into execution instructions of laser grinding equipment, so that the motion path of a laser beam is reasonable, energy output is precise, idle stroke and thermal deformation are reduced, thermal stress is controlled through periodic cooling, and meanwhile the grinding efficiency and the grinding effect of the platform are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of laser control, in particular to a laser polishing control method and system for a platform. BACKGROUND

[0002] The existing platform laser polishing control method mainly performs layering processing according to the side surface image of the platform, records the position information, contour shape information and relationship information between adjacent layers of each layer, forms the layering information data of the workpiece, and fuses and analyzes the layering information of the workpiece and the processing information, calculates and controls the specific processing path, laser action time and polishing times of each layer according to the processing precision requirement. However, the above method depends on fixed threshold or simple geometric features for layering, does not consider the flow characteristics of the material after heating and the height of the bulge around the heat affected zone caused by laser heating, cannot quantify the potential of the material to transfer and compensate between the convex and concave areas, is difficult to achieve global optimization, and can cause the platform to have high roughness, thereby causing the laser polishing efficiency and effect to be poor.

[0003] Therefore, how to control the polishing process of laser polishing so as to improve the laser polishing efficiency and the polishing effect of the platform becomes a problem to be solved. SUMMARY

[0004] In view of the above technical problems, the technical scheme adopted by the present application is a laser polishing control method for a platform, which comprises the following steps: S1, according to the target depth value corresponding to the platform to be polished, the surface image and the point cloud data, a plurality of convex regions, a plurality of concave regions, a convex volume corresponding to each convex region and a concave volume corresponding to each concave region in the surface image are obtained, wherein the point cloud data includes the actual depth value corresponding to each pixel point in the surface image.

[0005] S2, according to 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, the mutual matching degree between each convex region and each concave region is obtained, wherein the mutual matching degree is a value greater than or equal to 0.

[0006] S3, according to the mutual matching degree between each convex region and each concave region, the polishing path of the laser beam and the corresponding time sequence control data are obtained, wherein the time sequence control data includes the corresponding state mode of the laser beam at each unit time point when the platform to be polished is polished according to the polishing path, and the state mode includes a heating mode and a cooling mode.

[0007] S4, polishing the platform to be polished according to the polishing path and the control data.

[0008] The application also provides a laser polishing control system for a platform, which comprises: A region extraction module is configured to acquire a plurality of convex regions, a plurality of concave regions, a convex volume corresponding to each convex region and a concave volume corresponding to each concave region in the surface image according to the target depth value corresponding to the platform to be polished, the surface image and the point cloud data, wherein the point cloud data comprises an actual depth value corresponding to each pixel in the surface image.

[0009] A mutual matching degree acquisition module is configured to acquire a mutual matching degree between each convex region and each concave region according to the positions 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.

[0010] A control data acquisition module is configured to acquire a polishing path of the laser beam and corresponding time sequence control data according to the mutual matching degree between each convex region and each concave region, wherein the time sequence control data comprises a state mode corresponding to the laser beam at each unit time point when polishing the platform to be polished according to the polishing path, and the state mode comprises a heating mode and a cooling mode.

[0011] A control polishing module is configured to polish the platform to be polished according to the polishing path and the control data.

[0012] The application has at least the following beneficial effects: by comparing the target depth value with the actual depth value in the point cloud data, the convex regions and the concave regions are extracted in combination with the surface image, and the volumes are calculated, so as to quantify the spatial distribution and material quantity difference of each type of region in the platform to be polished, to provide a geometric and physical quantity basis for subsequent material transfer, to avoid precision loss caused by blind processing, to analyze the mutual matching degree between the convex regions and the concave regions by fusing the positional relationship, volume matching degree and other multi-dimensional parameters of the convex regions and the concave regions, to quantify and characterize the feasibility and efficiency of material transfer between the convex regions and the concave regions when heated, to facilitate preferential matching of high-value region pairs, to improve the rationality of the overall processing plan and the resource utilization rate, to generate a polishing path based on the mutual matching degree, to distribute the heating / cooling time sequence in combination with the material thermal properties, and to convert the planned polishing path and time sequence control data into execution instructions of the laser polishing equipment, so as to rationalize the laser beam motion path, to accurately output the energy, to reduce the idle stroke and thermal deformation, to control the thermal stress through periodic cooling, and to improve the polishing efficiency and the polishing effect of the platform. BRIEF DESCRIPTION OF DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0014] Figure 1 This is a flow chart of a laser grinding control method for a platform provided in Example 1 of the present invention; Figure 2 This is a schematic diagram of a laser grinding control system for a platform provided in Example 2 of the present invention. DETAILED DESCRIPTION

[0015] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0016] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings 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 above-mentioned terms used to distinguish similar objects can be interchanged so that the present invention can also implement other embodiments other than the above-mentioned illustrated embodiments or described embodiments. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. 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 clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices. Example 1

[0017] This embodiment provides a laser grinding control method for a platform, such as Figure 1 As shown, the laser grinding control method for the platform includes the following steps: S1. According to the target depth value, surface image and point cloud data corresponding to the platform to be ground, 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 are obtained, wherein the point cloud data includes the actual depth value corresponding to each pixel point in the surface image.

[0018] The to-be-ground platform is a workpiece carrier that needs to be repaired and flattened by laser technology due to the surface geometry or roughness not meeting the design requirements. The laser grinder can utilize the high energy density characteristics of the output laser beam, utilize laser scanning to make the molten material in the raised area of the to-be-ground platform flow and fill the recessed area, and realize surface flattening after cooling, so that the to-be-ground platform meets the design requirements.

[0019] The target depth value is the ideal design depth of the to-be-ground platform, for example, the standard thickness of a mold, which can be determined by the implementer according to the actual situation such as the CAD model or process requirements.

[0020] The surface image is a two-dimensional image of the upper surface of the to-be-ground platform that needs to be ground, which is captured by an image acquisition device such as a camera. The point cloud data is two-dimensional coordinate and depth value data of the upper surface of the to-be-ground platform that needs to be ground, which is obtained by a point cloud data acquisition device such as a laser scanner or a structured light camera, and is used to provide the geometric shape information of the upper surface of the to-be-ground platform. The point cloud data is aligned with the surface image to ensure that each pixel point corresponds to a unique actual depth value, which serves as the basis for laser control.

[0021] The raised area is a continuous area with an actual depth value lower than the target depth value, which can be regarded as a "material source" in the laser grinding process and can be filled by molten material after being heated by laser. The recessed area is a continuous area with an actual depth value higher than the target depth value, which can be regarded as a "repair target" of laser grinding and needs to be filled to the target depth by the molten material from the raised area.

[0022] The raised volume can represent the amount of material that the corresponding raised area can provide, and the recessed volume can represent the amount of material that the corresponding recessed area needs.

[0023] In a specific embodiment, S1 includes the following steps: S11, obtaining the target platform thickness corresponding to the to-be-ground platform, the surface image, the point cloud data, and the first distance between the bottom surface of the to-be-ground platform and the point cloud data acquisition device, wherein the actual depth value corresponding to each pixel point in the point cloud data is the distance between the corresponding position of each pixel point on the to-be-ground platform and the point cloud data acquisition device.

[0024] S12, determining the difference between the first distance and the target platform thickness as the target depth value corresponding to the to-be-ground platform.

[0025] S13, determining the difference between the target depth value and the actual depth value corresponding to each pixel point as the height difference value corresponding to each pixel point.

[0026] S14, obtaining a plurality of raised areas and a plurality of recessed areas in the surface image according to the comparison result between the height difference value corresponding to each pixel point and the preset height threshold.

[0027] S15, obtaining a protrusion volume corresponding to each protrusion region according to the height difference value corresponding to each pixel point in each protrusion region.

[0028] S16, obtaining a depression volume corresponding to each depression region according to the height difference value corresponding to each pixel point in each depression region.

[0029] The preset height threshold is a height difference value limit value for distinguishing the protrusion and depression regions, which can be set by the implementer according to platform accuracy requirements, material characteristics and other factors.

[0030] The target platform thickness is an ideal parameter set based on platform design requirements, and the target depth value represents a standard distance between each point on the upper surface of the platform and the point cloud data acquisition device after laser lapping processing, thereby providing a unified quantitative benchmark for the region division of the entire platform.

[0031] For each determined protrusion region / depression region, the height difference value of each pixel point in the protrusion region / depression region is accumulated, and the actual area corresponding to the pixel point is combined to calculate the total volume of the protrusion region / depression region by integration or summation. The protrusion volume is the volume of the material below the target depth value in each protrusion region, which is used to quantify the amount of material in the protrusion region, and the depression volume is the space volume above the target depth value in each depression region, which is used to quantify the amount of material needed to fill the depression region, thereby providing a basis for material removal and transfer during lapping. The actual area corresponding to the pixel point is determined by the acquisition accuracy of the point cloud data.

[0032] Through the fusion of point cloud data, surface image, target platform thickness and device distance, the above-mentioned method realizes complementary analysis of the three-dimensional geometric features and two-dimensional visual features of the upper surface of the platform to be lapped, breaks through the limitation of a single data source, and improves the identification accuracy of the protrusion region and the depression region, as well as the calculation accuracy of the protrusion volume and the depression volume.

[0033] In a specific embodiment, S14 includes the following steps: S141, for any pixel point, if the height difference value corresponding to the current pixel point is greater than the preset height threshold, the first mark value corresponding to the pixel point is determined as a first category value, wherein the category of the pixel point corresponding to the first category value is a protrusion pixel point.

[0034] S142, if the height difference value corresponding to the current pixel point is less than the preset height threshold, the first mark value corresponding to the pixel point is determined as a second category value, wherein the category of the pixel point corresponding to the second category value is a depression pixel point.

[0035] S143, the first label value corresponding to all pixel points is updated according to the region growing algorithm, and the second label value corresponding to each pixel point is obtained.

[0036] S144, if the second label value corresponding to the current pixel point is the first category value, the category corresponding to the current pixel point is determined as a convex pixel point, if the second label value corresponding to the current pixel point is the second category value, the category corresponding to the current pixel point is determined as a concave pixel point.

[0037] S145, according to the category corresponding to each pixel point, a plurality of convex regions and a plurality of concave regions are obtained.

[0038] The specific values of the first category value and the second category value can be set by the implementer. For example, the first category value is 1, and the second category value is 0, which is used to distinguish convex pixel points and concave pixel points. The preset height threshold is 0.

[0039] If the height difference value corresponding to the current pixel point 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 point, that is, the distance between the current pixel point and the point cloud data acquisition device is greater than the standard distance that should be reached, then the current pixel point is a convex pixel point.

[0040] If the height difference value corresponding to the current pixel point 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 point, that is, the distance between the current pixel point and the point cloud data acquisition device is less than the standard distance that should be reached, then the current pixel point is a concave pixel point.

[0041] The first label value corresponding to all pixel points is updated according to the region growing algorithm, and the second label value corresponding to each pixel point is obtained.

[0042] The first label value of the preliminarily classified pixel points is optimized and clustered by the region growing algorithm, and isolated points or noise interference is eliminated, so that continuous and complete convex regions or concave regions are obtained, so that the division of the convex regions and the concave regions is more accurate. Those skilled in the art know that the region growing algorithm in the prior art falls within the protection scope of the present application, and will not be described here.

[0043] Further, after obtaining a plurality of convex regions and a plurality of concave regions by the region growing algorithm, further morphological operations are performed on all regions according to the second label value corresponding to each pixel point, the region contour is further optimized, noise interference is eliminated, and small cavities or fractures are filled, so that the region boundary is smoother and the structure is more complete, and finally a plurality of convex regions and a plurality of concave regions are obtained. The morphological operations include erosion, dilation, opening operation and closing operation. Those skilled in the art know that any morphological operation in the prior art falls within the protection scope of the present application, and will not be described here.

[0044] The above, by the height difference value and the preset height threshold value determine the category of the pixel point, and the first mark value corresponding to all pixel points is updated through the region growing algorithm, which avoids the error classification caused by single pixel point error or noise, and improves the accuracy and integrity of the region division.

[0045] S2, according to 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, the mutual matching degree between each convex region and each concave region is obtained, wherein the mutual matching degree is greater than or equal to 0.

[0046] Wherein, the mutual matching degree is used to measure the material filling feasibility of a convex region to a specific concave region, and the mutual matching degree is greater than or equal to 0. Correspondingly, the larger the mutual matching degree value is, the higher the feasibility of material filling of the corresponding convex region to the corresponding concave region is.

[0047] In a specific embodiment, S2 includes the following steps: S21, according to the position of each convex region and each concave region in the surface image, the second distance between each convex region and each concave region, the first normal direction corresponding to each convex region and the second normal direction corresponding to each concave region are obtained.

[0048] S22, according to the position and actual depth value of each pixel point in the surface image, the fluid resistance matrix corresponding to the to-be-ground platform is obtained.

[0049] S23, for any convex region, according to the second distance between the current convex region and each concave region, a plurality of associated concave regions corresponding to the current convex region are obtained from all the concave regions.

[0050] S24, for any associated concave region corresponding to the current convex region, according to 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, the target matching degree between the current convex region and the current associated concave region is obtained.

[0051] S25, according to the fluid resistance matrix corresponding to the to-be-ground platform, the minimum resistance path length between the current convex region and the current associated concave region is obtained.

[0052] S26, according to the target matching degree and the minimum resistance path length between the current convex region and the current associated concave region, the mutual matching degree between the current convex region and the current associated concave region is obtained.

[0053] Among them, based on the corresponding coordinates of the positions of each raised area and each recessed area in the surface image, the centroid coordinates of each raised area and each recessed area are obtained, and the Euclidean distance between the corresponding centroid coordinates is used as the second distance between each raised area and each recessed area to evaluate the physical distance cost of material transfer.

[0054] The normal direction reflects the spatial orientation of the region by fitting the local surface of the pixels within the region. This provides geometric constraints for subsequent optimization of the laser energy transmission direction and avoids laser energy loss. Specifically, the first normal direction is the surface normal vector of the raised area, which is used to represent the diffusion direction of the material. The second normal direction is the surface normal vector of the recessed area, which is used to represent the receiving direction of the material.

[0055] 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, and 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 combined with the resistance values ​​to characterize the resistance distribution encountered by material particles when they diffuse and move on the platform surface, thereby reflecting the influence of the platform surface morphology on the diffusion and movement of the material, and providing a data basis for the resistance dimension when subsequently planning the optimal path.

[0056] A distance-related threshold is set, and based on the second distance, concave areas whose distance from the current convex area is less than the threshold are screened out as associated concave areas, indicating that material transfer and compensation can be performed on the associated concave areas based on the current convex area.

[0057] The degree of target matching between each raised area and each corresponding associated recessed area is measured by comprehensively considering spatial distance, area volume and normal direction, and the feasibility and efficiency of material transfer are comprehensively evaluated.

[0058] Based on the fluid resistance matrix and the optimal path search algorithm, the minimum resistance path length from the center of the current raised area to the center of the current associated recessed area is obtained for subsequent planning of a more reasonable material transfer path, avoiding obstacles or high resistance areas that may be encountered in traditional straight paths, and improving the efficiency of material transfer, that is, improving the platform grinding efficiency and grinding effect.

[0059] In a specific embodiment, according to the target matching degree M between the current convex area and the current associated concave area ij and the length of the path of least resistance, L ij , obtain the mutual matching degree C between the current convex area and the current associated concave area ij =M ij ×exp(-α×L0 / L ij ), where M ijis the target matching degree between the i-th convex area and the corresponding j-th associated concave area, L ij is the minimum resistance path length between the i-th raised area and the j-th associated recessed area, α is the preset resistance influence coefficient, L0 is the average minimum resistance path length corresponding to all raised areas, and exp() is the natural exponential function. α is used to adjust the impact of fluid resistance on material transfer feasibility. The specific value can be set by the implementer based on actual conditions such as material properties, and is generally in the range [0.1, 5].

[0060] As mentioned above, based on the distance, volume information, and normal direction of the raised and recessed areas, the physical distance cost and resistance distribution of material transfer are comprehensively considered. The material filling potential of each raised area for the associated recessed area is measured through multi-dimensional parameter fusion, and the degree of mutual matching between each raised area and each recessed area is obtained, which provides a data basis for the subsequent planning of more reasonable material transfer paths and control parameters to control the laser grinding process, thereby improving the platform grinding efficiency and grinding effect.

[0061] In a specific embodiment, S21 includes the following steps: S211 , for any convex region, construct a covariance matrix corresponding to the current convex region according to the coordinates of all pixels in the current convex region in the surface image.

[0062] S212 , performing eigendecomposition on the covariance matrix corresponding to the current convex region, and determining the eigenvector corresponding to the minimum eigenvalue as the first normal direction corresponding to the current convex region.

[0063] In a specific embodiment, S21 includes the following steps: S213 , for any concave area, construct a covariance matrix corresponding to the current concave area according to the coordinates of all pixels in the current concave area in the surface image.

[0064] S214 , performing eigendecomposition on the covariance matrix corresponding to the current concave area, and determining the eigenvector corresponding to the minimum eigenvalue as the second normal direction corresponding to the current concave area.

[0065] Among them, for the set of pixel points in each area, the corresponding covariance matrix can represent the geometric characteristics of the data distribution. Specifically, the eigenvector of the covariance matrix represents the main direction of the data distribution, and the eigenvalue represents the degree of dispersion of the data in the corresponding direction.

[0066] For the approximately planar point cloud region, since the discrete degree of the point cloud in the normal direction is minimum, the minimum eigenvalue of the covariance matrix corresponding to the convex region corresponds to the feature vector direction consistent with the first normal direction of the convex region, and the minimum eigenvalue of the covariance matrix corresponding to the concave region corresponds to the feature vector direction consistent with the second normal direction of the concave region.

[0067] The above, based on the pixel points of a region, the normal direction of each region is obtained based on the covariance matrix method, the spatial relationship of all points in each region is considered, the data noise resistance is improved, the accuracy of obtaining the normal direction is improved, and the measurement accuracy of the target mutual matching degree is improved.

[0068] In a specific embodiment, S22 includes the following steps: S221, according to the position and actual depth value corresponding to each pixel point, the gradient value, Gaussian curvature and average curvature corresponding to each pixel point are obtained.

[0069] S222, according to the gradient value corresponding to each pixel point and the preset terrain resistance coefficient, the terrain resistance corresponding to each pixel point is obtained.

[0070] S223, according to the Gaussian curvature and average curvature corresponding to each pixel point, the curvature resistance corresponding to each pixel point is obtained.

[0071] S224, according to the terrain resistance and curvature resistance corresponding to each pixel point, the comprehensive resistance corresponding to each pixel point is obtained.

[0072] S225, according to the Gaussian kernel function, the comprehensive resistance corresponding to each pixel point is smoothed, and the smoothed resistance corresponding to each pixel point is obtained.

[0073] S226, according to the smoothed resistance corresponding to all pixel points and the coordinates of each pixel point in the surface image, the fluid resistance matrix is obtained.

[0074] Wherein, according to the two-dimensional coordinates corresponding to each pixel point on the collected image, and adding the corresponding actual depth value, the three-dimensional coordinates corresponding to each pixel point are constructed. According to the three-dimensional coordinates, the gradient value ∇z corresponding to each pixel point is obtained to reflect the surface slope, the Gaussian slope K corresponding to each pixel point is obtained to reflect the surface bending type, and the average curvature H corresponding to each pixel point is obtained to reflect the surface bending degree. Those skilled in the art know that the calculation method of the gradient value, the Gaussian curvature and the average curvature in the prior art falls within the protection scope of the present application, and will not be repeated here. Wherein, when K>0, the corresponding surface bending type is an elliptic point type, the material particles will be subjected to a convergence effect when flowing in the corresponding area, when K<0, the corresponding surface bending type is a hyperbolic point type, the material particles will be subjected to an anisotropic resistance when flowing in the corresponding area, and when K=0, the corresponding surface bending type is a parabolic point type, the material particles will be subjected to a unidirectional anisotropic resistance or a curvature resistance when flowing in the corresponding area.

[0075] According to the gradient value ∇z corresponding to each pixel point and the preset terrain resistance coefficient μ, the terrain resistance corresponding to each pixel point is obtained, for example, the terrain resistance R1=μ×‖∇z‖×cosθ1, which is used to represent the hindering effect of the gravity component of the material particles on the slope. Wherein, θ1 is the included angle between the gravity direction and the normal direction corresponding to the pixel point, ‖∇z‖ is the modulus of the gradient, and the specific value of the terrain resistance coefficient μ is set by the implementer according to the material characteristics.

[0076] According to the Gaussian curvature K and the average curvature H corresponding to each pixel point, the curvature resistance corresponding to each pixel point is obtained, for example, the curvature resistance R2=β×|K|+γ×|H|, which is used to represent the resistance condition of the convex area / the concave area to the material transfer / filling. Wherein, 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.

[0077] According to the terrain resistance R1 and the curvature resistance R2 corresponding to each pixel point, the comprehensive resistance corresponding to each pixel point is obtained, for example, the comprehensive resistance R3=w1×R1+w2×R2, wherein 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.

[0078] According to the horizontal and vertical coordinates of each pixel point in the collected image, the position of each pixel point in the resistance matrix is determined, so as to combine the comprehensive resistance to form the resistance matrix, and then the resistance matrix is smoothed using the Gaussian kernel function to obtain the fluid resistance matrix composed of the smoothed smoothing resistance, so as to eliminate the resistance mutation caused by local noise and improve the construction accuracy of the fluid resistance matrix.

[0079] Based on the position and actual depth value of each pixel point, the surface slope, surface bending type and surface bending degree represented by each pixel point are analyzed, then the comprehensive resistance is obtained by comprehensively considering the terrain resistance and curvature resistance corresponding to each pixel point, and the fluid resistance matrix is obtained after smoothing, thereby improving the construction accuracy of the fluid resistance matrix, and further improving the measurement accuracy of the target mutual matching degree.

[0080] In a specific embodiment, S23 comprises the following steps: S231, for any protrusion region, the equivalent diameter corresponding to the current protrusion region is obtained according to the connected domain area corresponding to the current protrusion region.

[0081] S232, the associated search diameter corresponding to the current protrusion region is obtained according to the maximum protrusion volume in all protrusion volumes, the maximum recess volume in all recess volumes and the equivalent diameter corresponding to the current protrusion region.

[0082] S233, for any recess region, if the second distance between the current protrusion region and the current recess region is less than the associated search diameter corresponding to the current protrusion region, the current recess region is determined as the associated recess region corresponding to the current protrusion region.

[0083] S234, all associated recess regions corresponding to the current protrusion region are obtained by traversing all recess regions.

[0084] The connected domain area of the protrusion region is equivalent to a circular area, and the equivalent diameter is calculated, for example, the equivalent diameter D1=2×(A / π) 0.5 The equivalent diameter is used as the geometric size of the protrusion region to evaluate the effective coverage range of material ejection after the protrusion region is heated by laser. Wherein, A is the connected domain area of the protrusion region, that is, the total area of all pixel points in the protrusion region, and π is the circular constant.

[0085] The maximum protrusion volume V max1 in all protrusion volumes, the maximum recess volume V max2 in all recess volumes and the equivalent diameter D1 corresponding to the current protrusion region are combined to dynamically set the associated search diameter, so as to avoid the problem of "too large search range of small protrusion" or "insufficient search range of large protrusion" caused by fixed threshold, thereby balancing the calculation efficiency and matching integrity, and reducing invalid search. For example, the associated search diameter D2=w3×D1+w4×((V max1 +V max2 ) / 2) 1 / 3 Wherein, D1 can dominate the short-distance search, ((V max1 +V max2 ) / 2) 1 / 3The step length can be set according to the distance requirement of the extreme recessed area, for example, a large volume recessed area can require a longer protruding area to fill. The specific values of the weights w3 and w4 can be set by the implementer according to the actual situation. For example, w3 = 0.4, w4 = 0.6.

[0086] The above, in combination with the maximum protruding volume in all protruding volumes, the maximum recessed volume in all recessed volumes, and the dynamic setting of the equivalent diameter corresponding to the current protruding area, balances the calculation efficiency and matching integrity, and reduces the workload of invalid search.

[0087] In a specific embodiment, S24 includes the following steps: S241, according to the preset distance attenuation coefficient, the second distance between the current protruding area and the current associated recessed area, obtaining the distance matching degree between the current protruding area and the current associated recessed area.

[0088] S242, according to the protruding volume corresponding to the current protruding area and the recessed volume corresponding to the current associated recessed area, obtaining the volume matching degree between the current protruding area and the current associated recessed area.

[0089] S243, according to the included angle between the first normal direction corresponding to the current protruding area and the second normal direction corresponding to the current associated recessed area, obtaining the direction consistency degree between the current protruding area and the current associated recessed area.

[0090] S244, according to the distance matching degree, the volume matching degree and the direction consistency degree between the current protruding area and the current associated recessed area, obtaining the target matching degree between the current protruding area and the current associated recessed area.

[0091] Wherein, the closer the second distance, the higher the distance matching degree between the corresponding regions. For example, the distance matching degree E ij =exp(-λ×d ij ), which is used to represent the exponential decay of material transfer efficiency with distance, avoiding energy loss caused by long-distance transfer. Wherein, d ij is the second distance between the ith protruding area and the jth associated recessed area, λ is the 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 decays, suitable for high viscosity materials such as ceramics, the smaller the value of λ, allowing longer distance matching, suitable for low viscosity materials such as aluminum alloy.

[0092] The closer the convex volume and the concave volume, the higher the material transfer efficiency, and the higher the volume matching degree between the corresponding convex region and the associated concave region. For example, the volume matching degree F ij = exp(-|(V i ij )-1| σ ), where V i is the convex volume corresponding to the ith convex region, V ij is the concave volume of the jth associated concave region corresponding to the ith convex region, and σ is an adjustment index for controlling the punishment strength of the volume mismatch, i.e., the imbalance of material supply and demand, and the greater the value of σ, the higher the sensitivity to volume difference. The specific value of λ can be set by the implementer according to the actual situation, for example, λ = 2.

[0093] When the first normal direction corresponding to the ith convex region and the second normal direction corresponding to the jth associated concave region are consistent, i.e., the included angle θ ij = 0° between the first normal direction corresponding to the ith convex region and the second normal direction corresponding to the jth associated concave region, it indicates that the filling effectiveness of the material melted by laser heating of the ith convex region to the jth associated concave region material is higher, and the corresponding G ij = 1. When the first normal direction corresponding to the ith convex region and the second normal direction corresponding to the jth associated concave region are opposite, i.e., θ ij = 180°, it indicates that the filling effectiveness of the material melted by laser heating of the ith convex region to the jth associated concave region material is lower, and the corresponding G ij = 0. Therefore, the degree of consistency G ij =(1+cosθ ij ) / 2 is obtained between the first normal direction corresponding to the ith convex region and the jth associated concave region.

[0094] According to the distance matching degree, the volume matching degree and the direction consistency degree between the current convex region and the current associated concave region, and combining the weighted sum of the distance matching degree, the volume matching degree and the direction consistency degree respectively corresponding to the weight, the target matching degree between the current convex region and the current associated concave region is obtained, which represents the material transfer adaptability between the convex region and the concave region. The specific values of the weights corresponding to the distance matching degree, the volume matching degree and the direction consistency degree can be set by the implementer according to the actual situation, for example, the weights corresponding to the distance matching degree, the volume matching degree and the direction consistency degree are 0.3, 0.4 and 0.3 respectively.

[0095] ​The above, the distance matching degree is obtained to analyze the physical law of the material transfer efficiency decay with distance, the volume matching degree is used to analyze the material supply and demand balance degree, the normal direction is analyzed by the direction consistency degree analysis method The influence of the laser energy transmission efficiency and effectiveness, and the target matching degree is obtained by comprehensively considering the three dimensions, which represents the material transfer adaptability between the convex region and the concave region, as the data basis for subsequent planning of more reasonable material transfer path and control parameter, to control the grinding process of laser grinding, thereby improving the grinding efficiency and grinding effect of the platform.

[0096] S3, according to the mutual matching degree between each convex region and each concave region, the grinding path of the laser beam and the corresponding time sequence control data are obtained, wherein the time sequence control data includes the corresponding state mode of the laser beam at each unit time point when the platform to be ground is ground according to the grinding path, and the state mode includes heating mode and cooling mode.

[0097] In the embodiment, the laser grinder works in a cycle according to the heating and cooling state model. When the laser beam acts on the convex region in the heating mode, the material temperature is rapidly raised by high energy density. When the temperature reaches the melting point or vaporization point of the material, the convex part is evaporated or melted and transferred to the concave region to fill the gap by the flow of the melted material. After heating, the cooling model is switched, for example, inert gas is introduced, so that the melted material can be quickly solidified, avoiding excessive melting or thermal deformation of the platform due to continuous heating. At the same time, thermal stress will be generated in the material during heating (expansion) and cooling (contraction). The alternating operation can make the stress distribution uniformized through periodic temperature change, avoiding material cracking or delamination caused by single intense heating / cooling, so as to balance the material removal efficiency and form control through dynamic thermal management, and further improve the grinding effect of the laser beam on the platform to be ground.

[0098] In a specific embodiment, S3 includes the following steps: S31, all the mutual matching degrees are sorted in descending order, and the priority matrix is obtained according to the sorting result, wherein the row priority in the priority matrix is arranged in descending order according to the maximum mutual matching degree of the corresponding convex region, the column priority is arranged in descending order according to the maximum mutual matching degree sum of the corresponding concave region, and the value of each element in the priority matrix is the mutual matching degree between the corresponding row convex region and the corresponding column concave region.

[0099] S32, according to the priority matrix, the optimal path algorithm is used to optimize the moving path of the laser beam between each convex region and each concave region, and the corresponding grinding path of the laser beam in the laser grinder is obtained.

[0100] S33, obtaining a total heating time of the laser beam for each protrusion region according to the protrusion volume corresponding to each protrusion region, the preset material properties, the preset laser power, and the preset efficiency of the laser grinder.

[0101] S34, obtaining a state mode of the laser beam at each unit time point according to the sequence of each protrusion region in the grinding path, the total heating time of each protrusion region, and the preset cooling time.

[0102] wherein the row priority in the priority matrix is in descending order of the maximum mutual matching degree corresponding to the protrusion region, the column priority is in descending order of the maximum sum of mutual matching degrees corresponding to the depression region, and the element value Y uv corresponds to the mutual matching degree between the protrusion region of the corresponding row and the depression region of the corresponding column. If the element value corresponds to a depression region that is not the associated depression region of the protrusion region of the corresponding row, the element value corresponds to the value 0. The number of rows in the priority matrix is consistent with the total number of protrusion regions, and the number of columns in the priority matrix is consistent with the total number of depression regions.

[0103] According to the priority matrix, the optimal path algorithm is used to optimize the movement path of the laser beam between each protrusion region and each depression region, for example, the improved Travelling Salesman Problem is used to generate the optimal movement path: min(Σ u,v (Z uv ×x uv )+Σ u,v (x uv / Y uv )), wherein Z uv is equal to the second distance between the protrusion region of the u-th row and the depression region of the v-th column, x uv is a path selection variable. By introducing the mutual matching degree as a heuristic factor, the path sequence is dynamically adjusted to obtain a movement path with the highest global mutual matching degree, which is used as the grinding path of the laser beam in the laser grinder. It is known to those skilled in the art that any optimal path algorithm in the prior art falls within the protection scope of the present application, and will not be described here.

[0104] In a specific embodiment, the preset material properties include material density ρ, specific heat capacity c, melting temperature difference ΔT, and latent heat of vaporization J, and the total heating time of the laser beam for the i-th protrusion region is obtained according to the protrusion volume V i, the preset material properties, the preset laser power P and the preset efficiency η of the laser grinder, and obtain the total heating time of the laser beam corresponding to the i-th raised area. By limiting the total heating time, the material of the raised area is effectively transferred to the associated recessed area while preventing the deformation of the platform substrate caused by excess energy and preventing the height of the hot area from being too high, thereby improving the effect of platform grinding. For example, the total heating time Ψ corresponding to the i-th raised area is 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 grinder are set by the implementer according to actual conditions. For example, η=0.5.

[0105] Arrange the raised areas in the order of the grinding path, and perform cyclic work according to the state model of heating and cooling, and obtain the state mode corresponding to the laser beam at each unit time point when grinding the grinding platform according to the grinding path. For example, the value 1 represents the heating mode, and the value 0 represents the cooling mode. If the total heating time Ψ1 corresponding to the first raised area 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 the grinding of the first raised area is {(t1, 1), (t2, 1), ..., (t N1 ,1),(t N1+1 ,0),……,(t N1+N0 ,0)}. The total heating time Ψ corresponding to the i-th raised area in the grinding path i Including Ni unit time points, the corresponding timing control sub-data SX of the i-th convex area is i is {(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 in the heating mode and the total number of unit time points in the cooling mode corresponding to the first i-1 convex regions.

[0106] As described above, a priority matrix is ​​formed by the degree of mutual matching, and the multi-region matching problem is converted into a computable structured model. The high-value areas are focused, invalid processing is reduced, and the overall grinding efficiency is improved. The path order is dynamically adjusted based on the priority matrix to obtain the grinding path with the highest global mutual matching degree. The total heating time corresponding to each raised area of ​​the laser beam is obtained according to the raised volume and preset parameters, so that each raised area is arranged in the order of the grinding path, and the heating and cooling state model is cyclically worked to obtain the state mode corresponding to the laser beam at each unit time point when the grinding platform to be ground is ground according to the grinding path. The problems of low efficiency of multi-region collaborative processing and difficulty in controlling thermal deformation in laser grinding are solved, thereby improving the grinding efficiency and grinding effect.

[0107] S4, grinding the platform to be ground according to the grinding path and control data.

[0108] Based on the grinding path and control data, the laser grinder's laser head is driven along the grinding path to ensure coverage of all raised areas to be heated. The control data also controls the switching between heating and cooling modes. Specifically, when in heating mode, the laser beam is controlled to heat the surface, while when in cooling mode, the laser grinder's laser head is controlled to move along the grinding path to the next raised area to be heated. This balances material removal with thermal deformation control, improving the flatness of the surface being ground.

[0109] In a specific embodiment, the control method further includes: S5, after the platform to be ground is ground according to the grinding path and the control data, an updated surface image and updated point cloud data corresponding to the platform to be ground are obtained.

[0110] S6, obtaining the smoothness corresponding to the platform to be ground according to the point cloud data. When the smoothness does not meet the preset stop condition, repeat step S1 until the new smoothness corresponding to the platform to be ground meets the preset stop condition, and obtain the target platform.

[0111] The smoothness of the surface to be ground is assessed based on the surface image and point cloud data. For example, the surface image and point cloud data are input into a preset roughness assessment model to obtain the smoothness of the surface to be ground. The preset roughness assessment model can be constructed based on a convolutional neural network model and trained based on the labeled surface image and point cloud data to improve the accuracy of the smoothness. Those skilled in the art will recognize that any roughness assessment model and corresponding model training method in the prior art fall within the scope of protection of the present invention and will not be further described here.

[0112] Through multiple rounds of iterative control and grinding, the raised areas and the material protrusion height caused by the previous heating can be removed during the next controlled grinding, and the recessed areas can be filled more finely, thereby gradually improving the smoothness of the platform to be ground and enhancing the grinding effect.

[0113] As mentioned above, by comparing the target depth value with the actual depth value in the point cloud data, combining the surface image to extract the raised area and the recessed area and calculate the volume, the spatial distribution and material quantity difference of each type of area 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, and by integrating multi-dimensional parameters such as the positional relationship and volume matching between the raised area and the recessed area, the mutual matching degree between the raised area and the recessed area is analyzed, thereby quantitatively characterizing the feasibility and efficiency of material transfer between the raised area and the recessed area when heated, facilitating the priority matching of high-value area pairs, improving the rationality and resource utilization of the overall processing planning, generating a grinding path based on the mutual matching degree, allocating heating / cooling timing 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, so that the laser beam movement path is rationalized, the energy output is precise, and the idle stroke and thermal deformation are reduced. At the same time, thermal stress is controlled through periodic cooling, while improving the grinding efficiency and the grinding effect of the platform.

[0114] Example 2 This embodiment 2 provides a laser grinding control system for a platform, such as Figure 2 As shown, the laser grinding control system for the platform includes: The area extraction module 21 is used to obtain a number of raised areas, a number of recessed areas, a raised volume corresponding to each raised area, and a recessed volume corresponding to each recessed area in the surface image according to the target depth value, surface image, and point cloud data corresponding to the platform to be ground, wherein the point cloud data includes the actual depth value corresponding to each pixel point in the surface image.

[0115] The mutual matching degree acquisition module 22 is used to obtain the mutual matching degree between each convex area and each concave area based on the position of each convex area and each concave area in the surface image, the convex volume corresponding to each convex area, and the concave volume corresponding to each concave area, wherein the mutual matching degree is a value greater than or equal to 0.

[0116] The control data acquisition module 23 is used to obtain the grinding path of the laser beam and the corresponding timing control data based on the degree of mutual matching between each raised area and each recessed area, wherein the timing control data includes the state mode corresponding to the laser beam at each unit time point when grinding the grinding platform according to the grinding path, and the state mode includes a heating mode and a cooling mode.

[0117] The grinding module 24 is controlled to grind the target platform according to the grinding path and the control data In an embodiment, the region extraction module 21 comprises: A first distance acquisition submodule is configured to acquire a target platform thickness corresponding to the target platform, a surface image, point cloud data, and a first distance between a bottom surface of the target platform and a point cloud data acquisition device, wherein an actual depth value of each pixel point in the point cloud data is a distance between the corresponding position of each pixel point on the target platform and the point cloud data acquisition device.

[0118] A target depth value acquisition submodule is configured to determine a difference between the first distance and the target platform thickness as a target depth value corresponding to the target platform.

[0119] A height difference value acquisition submodule is configured to determine a difference between the target depth value and the actual depth value of each pixel point as a height difference value corresponding to each pixel point.

[0120] A region extraction submodule is configured to acquire a plurality of protruding regions and a plurality of recessed regions in the surface image according to a comparison result between the height difference value corresponding to each pixel point and a preset height threshold.

[0121] A protruding volume acquisition submodule is configured to acquire a protruding volume corresponding to each protruding region according to the height difference value corresponding to each pixel point in each protruding region.

[0122] A recessed volume acquisition submodule is configured to acquire a recessed volume corresponding to each recessed region according to the height difference value corresponding to each pixel point in each recessed region.

[0123] In an embodiment, the region extraction submodule comprises: A first comparison unit is configured to, for any pixel point, determine a first mark value corresponding to the pixel point as a first category value if the height difference value corresponding to the current pixel point is greater than the preset height threshold, wherein the category of the pixel point corresponding to the first category value is a protruding pixel point.

[0124] A second comparison unit is configured to, if the height difference value corresponding to the current pixel point is less than the preset height threshold, determine the first mark value corresponding to the pixel point as a second category value, wherein the category of the pixel point corresponding to the second category value is a recessed pixel point.

[0125] A mark value updating unit is configured to update the first mark value corresponding to all pixel points according to a region growing algorithm to acquire a second mark value corresponding to each pixel point.

[0126] The pixel point classification unit is configured to determine the category of the current pixel point as a convex pixel point if the second mark value corresponding to the current pixel point is the first category value, or determine the category of the current pixel point as a concave pixel point if the second mark value corresponding to the current pixel point is the second category value.

[0127] The region extraction unit is configured to obtain a plurality of convex regions and a plurality of concave regions according to the category of each pixel point.

[0128] In an embodiment, the mutual matching degree acquisition module 22 comprises: The data analysis submodule is configured to obtain a second distance between each convex region and each concave region, a first normal direction corresponding to each convex region, and a second normal direction corresponding to each concave region according to the position of each convex region and each concave region in the surface image.

[0129] The fluid resistance matrix acquisition submodule is configured to obtain a fluid resistance matrix corresponding to the polishing platform according to the position and actual depth value of each pixel point in the surface image.

[0130] The associated concave region screening submodule is configured to, for any convex region, obtain a plurality of associated concave regions corresponding to the current convex region from all concave regions according to the second distance between the current convex region and each concave region.

[0131] The target matching degree acquisition submodule is configured to, for any associated concave region corresponding to the current convex region, obtain a target matching degree between the current convex region and the current associated concave region according to 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.

[0132] The minimum resistance path length acquisition submodule is configured to obtain a minimum resistance path length between the current convex region and the current associated concave region according to the fluid resistance matrix corresponding to the polishing platform.

[0133] The mutual matching degree acquisition submodule is configured to obtain a mutual matching degree between the current convex region and the current associated concave region according to the target matching degree and the minimum resistance path length between the current convex region and the current associated concave region.

[0134] In an embodiment, the data analysis submodule comprises: The covariance matrix construction unit is configured to, for any convex region, construct a covariance matrix corresponding to the current convex region according to the coordinates of all pixel points in the current convex region in the surface image.

[0135] The first normal direction determination unit is configured to perform eigenvalue decomposition on the covariance matrix corresponding to the current protrusion region, and determine a feature vector corresponding to the minimum eigenvalue as the first normal direction of the current protrusion region.

[0136] In an embodiment, the fluid resistance matrix acquisition submodule comprises: The curvature acquisition unit is configured to acquire gradient values, Gaussian curvatures and average curvatures corresponding to each pixel point according to the positions and actual depth values of the pixel points.

[0137] The terrain resistance acquisition unit is configured to acquire terrain resistances corresponding to each pixel point according to the gradient values of the pixel points and preset terrain resistance coefficients.

[0138] The curvature resistance acquisition unit is configured to acquire curvature resistances corresponding to each pixel point according to the Gaussian curvatures and average curvatures of the pixel points.

[0139] The comprehensive resistance acquisition unit is configured to acquire comprehensive resistances corresponding to each pixel point according to the terrain resistances and curvature resistances of the pixel points.

[0140] The smooth resistance acquisition unit is configured to smooth the comprehensive resistances corresponding to each pixel point according to a Gaussian kernel function, and acquire smooth resistances corresponding to each pixel point.

[0141] The fluid resistance matrix acquisition unit is configured to acquire a fluid resistance matrix according to the smooth resistances corresponding to all pixel points and the coordinates of each pixel point in the surface image.

[0142] In an embodiment, the associated recessed region screening submodule comprises: The equivalent diameter acquisition unit is configured to acquire an equivalent diameter corresponding to a current protrusion region according to an area of a connected domain corresponding to the current protrusion region.

[0143] The associated search diameter acquisition unit is configured to acquire an associated search diameter corresponding to a current protrusion region according to a maximum protrusion volume in all protrusion volumes, a maximum recessed volume in all recessed volumes and the equivalent diameter corresponding to the current protrusion region.

[0144] The associated recessed region screening unit is configured to, for any recessed region, determine a current recessed region as an associated recessed region corresponding to a current protrusion region if a second distance between the current protrusion region and the current recessed region is less than the associated search diameter corresponding to the current protrusion region.

[0145] The region traversal unit is configured to traverse all recessed regions and acquire all associated recessed regions corresponding to a current protrusion region.

[0146] In a specific embodiment, the target matching degree acquisition submodule includes: The distance matching degree acquisition unit is used to acquire the distance matching degree between the current convex area and the current associated concave area according to a preset distance attenuation coefficient and the second distance between the current convex area and the current associated concave area.

[0147] The volume matching degree obtaining unit is configured to obtain the volume matching degree between the current convex region and the current associated concave region according to the convex volume corresponding to the current convex region and the concave volume corresponding to the current associated concave region.

[0148] The directional consistency degree acquisition unit is used to acquire the directional consistency degree between the current convex area and the current associated concave area according to the angle between the first normal direction corresponding to the current convex area and the second normal direction corresponding to the current associated concave area.

[0149] The target matching degree acquisition unit is used to acquire the target matching degree between the current convex area and the current associated concave area according to the distance matching degree, volume matching degree and direction consistency degree between the current convex area and the current associated concave area.

[0150] In a specific embodiment, the control data acquisition module 23 includes: The priority matrix acquisition submodule is used to sort all mutual matching degrees in order from large to small, and obtain the priority matrix according to the sorting results, wherein the row priority in the priority matrix is ​​that the convex areas are arranged in descending order according to the corresponding maximum mutual matching degrees, and the column priority is that the concave areas are arranged in descending order according to the corresponding maximum mutual matching degree sum, and each element value in the priority matrix is ​​the mutual matching degree between the convex area of ​​the corresponding row and the concave area of ​​the corresponding column.

[0151] 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 according to the priority matrix, and obtain the grinding path corresponding to the laser beam in the laser grinder.

[0152] The total heating time acquisition submodule is used to obtain the total heating time corresponding to each raised area of ​​the laser beam 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 grinder.

[0153] The state mode acquisition submodule is used to obtain the state mode corresponding to the laser beam at each unit time point according to the corresponding order of each raised area in the grinding path, the total heating time corresponding to each raised area and the preset cooling time.

[0154] It should be noted that the information interaction, execution process and the like between the above modules are based on the same concept as the method embodiments of the present application, and the specific functions and the resulting technical effects can be referred to the method embodiments part, which will not be described here.

[0155] The above is only a preferred embodiment of the present application, and does not limit the present application in any form. Although the present application has been disclosed as above with a preferred embodiment, it is not intended to limit the present application. Any person skilled in the art can make some changes or modifications to the above disclosed technical content without departing from the scope of the technical solution of the present application, and any simple modification, equivalent change and modification of the above embodiments according to the technical essence of the present application are still within the scope of the technical solution of the present application.

Claims

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

2. The laser grinding control method for a platform according to claim 1, characterized in that: S1 includes the following steps: S11, obtaining a target platform thickness, a surface image, point cloud data corresponding to the platform to be ground, and a first distance between the bottom surface of the platform to be ground and a point cloud data acquisition device, wherein the actual depth value corresponding to each pixel point in the point cloud data is the distance between the corresponding position of each pixel point on the platform to be ground and the point cloud data acquisition device; S12, determining the difference between the first distance and the target platform thickness as the target depth value corresponding to the platform to be ground; S13, determining the difference between the target depth value and the actual depth value corresponding to each pixel as the height difference corresponding to each pixel; S14, obtaining a plurality of convex areas and a plurality of concave areas corresponding to the surface image based on a comparison result between a height difference value corresponding to each pixel point and a preset height threshold; S15, obtaining the convex volume corresponding to each convex area according to the height difference corresponding to each pixel point in each convex area; S16, obtaining the concave volume corresponding to each concave area according to the height difference corresponding to each pixel point in each concave area.

3. The laser grinding control method for a platform according to claim 2, characterized in that: S14 includes the following steps: S141: For any pixel point, if the height difference corresponding to the current pixel point is greater than the preset height threshold, determine the first label value corresponding to the pixel point as a first category value, wherein the category of the pixel point corresponding to the first category value is a convex pixel point; S142: If the height difference corresponding to the current pixel point is less than the preset height threshold, determining the first label value corresponding to the pixel point as a second category value, wherein the category of the pixel point corresponding to the second category value is a concave pixel point; S143, updating the first label values ​​corresponding to all pixels according to a region growing algorithm to obtain a second label value corresponding to each pixel; S144: If the second label value corresponding to the current pixel is the first category value, the category corresponding to the current pixel is determined to be a convex pixel; if the second label value corresponding to the current pixel is the second category value, the category corresponding to the current pixel is determined to be a concave pixel; S145 , obtaining a plurality of raised areas and a plurality of sunken areas according to the category corresponding to each pixel point.

4. The laser grinding control method for a platform according to claim 1, characterized in that: S2 includes the following steps: S21, acquiring, based on positions of each convex area and each concave area in the surface image, a second distance between each convex area and each concave area, a first normal direction corresponding to each convex area, and a second normal direction corresponding to each concave area; S22, obtaining a fluid resistance matrix corresponding to the platform to be polished according to the position and actual depth value corresponding to each pixel point in the surface image; S23, for any convex area, obtaining a plurality of associated concave areas corresponding to the current convex area from all concave areas according to the second distance between the current convex area and each concave area; S24, for any associated concave area corresponding to the current convex area, obtaining a target matching degree between the current convex area and the current associated concave area based on the second distance between the current convex area and the current associated concave area, the convex volume and the first normal direction corresponding to the current convex area, and the concave volume and the second normal direction corresponding to the current associated concave area; S25, obtaining the minimum resistance path length between the current raised area and the currently associated recessed area according to the fluid resistance matrix corresponding to the platform to be ground; S26 , obtaining a mutual matching degree between the current convex region and the current associated concave region according to the target matching degree and the minimum resistance path length between the current convex region and the current associated concave region.

5. The laser grinding control method for a platform according to claim 4, characterized in that: S21 includes the following steps: S211, for any convex area, constructing a covariance matrix corresponding to the current convex area according to the coordinates of all pixels in the current convex area in the surface image; S212 , performing eigendecomposition on the covariance matrix corresponding to the current convex region, and determining the eigenvector corresponding to the minimum eigenvalue as the first normal direction corresponding to the current convex region.

6. The laser grinding control method for a platform according to claim 4, characterized in that: S22 includes the following steps: S221, obtaining the gradient value, Gaussian curvature, and average curvature corresponding to each pixel point according to the position and actual depth value corresponding to each pixel point; S222, obtaining the terrain resistance corresponding to each pixel point based on the gradient value corresponding to each pixel point and a preset terrain resistance coefficient; S223, obtaining the curvature resistance corresponding to each pixel point based on the Gaussian curvature and the average curvature corresponding to each pixel point; S224, obtaining the comprehensive resistance corresponding to each pixel point based on the terrain resistance and curvature resistance corresponding to each pixel point; S225, smoothing the integrated resistance corresponding to each pixel point according to the Gaussian kernel function to obtain the smoothed resistance corresponding to each pixel point; S226 , obtaining a fluid resistance matrix according to the smooth resistance corresponding to all pixel points and the coordinates of each pixel point in the surface image.

7. The laser grinding control method for a platform according to claim 4, characterized in that: S23 includes the following steps: S231, for any convex region, obtaining the equivalent diameter corresponding to the current convex region according to the area of ​​the connected domain corresponding to the current convex region; S232, obtaining an associated search diameter corresponding to the current convex area according to 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 area; S233: For any concave area, if the second distance between the current convex area and the current concave area is smaller than the associated search diameter corresponding to the current convex area, the current concave area is determined as the associated concave area corresponding to the current convex area; S234: traverse all the concave areas to obtain all associated concave areas corresponding to the current convex area.

8. The laser grinding control method for a platform according to claim 4, characterized in that: S24 includes the following steps: S241, obtaining a distance matching degree between the current convex area and the current associated concave area according to a preset distance attenuation coefficient and a second distance between the current convex area and the current associated concave area; S242, obtaining a 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; S243, obtaining a degree of directional consistency between the current convex region and the current associated concave region based on an angle between a first normal direction corresponding to the current convex region and a second normal direction corresponding to the current associated concave region; S244 , obtaining a target matching degree between the current convex region and the current associated concave region according to the distance matching degree, volume matching degree, and direction consistency degree between the current convex region and the current associated concave region.

9. The laser grinding control method for a platform according to claim 1, characterized in that: S3 includes the following steps: S31, sorting all mutual matching degrees in descending order, and obtaining a priority matrix based on the sorting results, wherein the row priority in the priority matrix is ​​that the convex areas are arranged in descending order according to the corresponding maximum mutual matching degrees, and the column priority is that the concave areas are arranged in descending order according to the corresponding maximum mutual matching degree sum, and each element value in the priority matrix is ​​the mutual matching degree between the convex area of ​​the corresponding row and the concave area of ​​the corresponding column; S32, optimizing the moving path of the laser beam between each raised area and each recessed area using an optimal path algorithm according to the priority matrix, and obtaining a grinding path corresponding to the laser beam in the laser grinder; S33, obtaining a total heating time of the laser beam corresponding to each raised area according to the raised volume corresponding to each raised area, preset material properties, preset laser power, and preset efficiency of the laser grinder; S34 , obtaining a state mode of the laser beam at each unit time point according to the order of each raised area in the grinding path, the total heating time and the preset cooling time corresponding to each raised area.

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

Images