Laser-based surface treatment method and equipment

Abstract

The invention discloses a laser-based surface treatment method and equipment, and the method comprises the steps: obtaining drawing document data of a to-be-treated metal material, carrying out the graphic layering processing of the drawing document data, and generating a multi-layer drawing document model; generating a first laser processing parameter set by combining the type, thickness and surface characteristics of the metal material; the metal material is placed on a jig provided with an automatic adjusting device, and the position of the jig is sequentially adjusted; the metal material is marked through a first laser according to the multi-layer drawing and document model, and machining feedback information of the metal material is collected through a sensor in the marking process; according to the difference between the processing response data and the drawing file data, generating a single-layer drawing file for correcting the processing deviation through an image processing and neural network identification method; the laser marking process and data feedback analysis are deeply fused, so that the precision and consistency of pattern forming are effectively improved; and meanwhile, the machining deviation can be remarkably reduced, so that the overall surface treatment quality and the product yield are improved.

Description

Technical Field

[0001] The present invention relates to the field of laser processing technology, and in particular to a laser-based surface processing method and equipment. Background Art

[0002] As the surface quality requirements for metal parts continue to rise in industries such as precision manufacturing, consumer electronics, new energy vehicles, and aerospace, laser processing technology is increasingly being used in the field of material surface treatment. In particular, laser surface treatment offers significant advantages such as non-contact, high precision, and a minimal heat-affected zone for microstructure marking, functional pattern formation, and fine patterning of metal surfaces. It has become an important means of improving product surface performance, enhancing recognition, and functionality.

[0003] Existing laser surface treatment methods typically use a fixed pattern to drive the laser along a preset path. Control parameters are often pre-set based on the material's initial state, lacking feedback and dynamic adjustment to the actual response during processing. This is particularly true when processing metal materials with large reflectivity variations or uneven surface conditions, leading to issues such as uneven marking depth, blurred edges, or pattern distortion. These drawbacks stem primarily from the existing solutions' lack of real-time monitoring and analysis of processing feedback information. This inability to adjust laser parameters based on actual changes during processing makes it difficult to ensure pattern accuracy and surface consistency.

[0004] In view of this, it is necessary to improve the existing laser processing technology to solve the technical problem that it is difficult to adapt to the surface differences of metal materials during the process, resulting in insufficient pattern accuracy. Summary of the Invention

[0005] The purpose of the present invention is to provide a laser-based surface treatment method and equipment to solve the above technical problems.

[0006] To achieve this object, the present invention adopts the following technical solutions: A laser-based surface treatment method comprises the following steps: Acquiring drawing data of the metal material to be processed, and performing graphic layering processing on the drawing data to generate a multi-layer drawing model including spatial structure information; In combination with the type, thickness and surface characteristics of the metal material, a first laser processing parameter set matching the multi-layer pattern is calculated and generated based on a preset parameter library; Placing the metal material on a jig provided with an automatic adjustment device, and sequentially adjusting the position of the jig through a displacement control module so that the surface of the metal material is at the quasi-focal position of the first laser; Marking the metal material according to the multi-layer pattern using a first laser, and collecting processing feedback information of the metal material using a sensor during the marking process, and transmitting the feedback information to an edge processing module for processing response data analysis; the feedback information includes surface reflectivity, temperature change, and pattern profile; According to the difference between the processing response data and the drawing data, a single-layer drawing for correcting the processing deviation is generated through image processing and neural network recognition methods.

[0007] Optionally, according to the difference between the processing response data and the image file data, a single-layer image file for correcting the processing deviation is generated through image processing and neural network recognition methods, and then the method further includes: Optimizing the processing parameters of the second laser according to the processing feedback information to form a correction parameter set and a refined image file suitable for fine surface processing; A second laser is used to perform surface treatment on the metal material after the first treatment according to the refined image file and the correction parameter set, and combined with laser micro-control to perform refinement treatment to obtain a second treated material with high-precision pattern edges and surface consistency.

[0008] Optionally, the step of acquiring the drawing data of the metal material to be processed and performing graphic layering processing on the drawing data to generate a multi-layer drawing model including spatial structure information specifically includes the following steps: Collecting basic design drawings of the metal material to be processed, the basic design drawings include two-dimensional vector graphics or three-dimensional modeling data, and performing format standardization conversion on the structural contour information in the basic design drawings to generate standard drawing files; Based on the standard drawing file, the graphic analysis module is called to identify the geometric structure of the required processing area on the metal material surface, and preliminary zoning and marking are performed according to different functional areas; the zoning marking includes boundary areas, fill areas, and high-density pattern areas; A layered modeling algorithm is used to reconstruct the spatial structure of the annotated graphics according to the preset layer logic to form a multi-layered image model containing depth information. The multi-layered image model includes at least three layers: a path outline layer, a structure filling layer, and a function adjustment layer. The multi-layer drawing model is parameterized and integrity checked to ensure that each layer contains a clear definition of the processing area, spatial coordinate information, and layer stacking order.

[0009] Optionally, combining the type, thickness, and surface characteristics of the metal material, generating a first laser processing parameter set that matches the multi-layer pattern based on a preset parameter library, specifically comprising the following steps: Analyzing the type, thickness, and surface roughness data of the metal material, calculating the absorption coefficient, thermal conductivity, and melting threshold using a material property quantification model, and generating a material property vector; Extracting spatial structural information of the multi-layer drawing model, the spatial structural information including inter-layer correlation, minimum line width and aspect ratio features, and constructing a drawing feature vector; The material characteristic vector and the image feature vector are input into a preset parameter library, and a basic laser parameter group is mapped out through a multi-dimensional matching algorithm. The basic laser parameter group includes a pulse frequency range, a scanning speed range and a power density reference value.

[0010] Optionally, the material characteristic vector and the image feature vector are input into a preset parameter library, and a basic laser parameter group is mapped out through a multi-dimensional matching algorithm, and then the following steps are further included: Based on the basic laser parameter group, a thermal-mechanical coupling simulation model is called to dynamically predict the machining deformation and the depth of the heat-affected zone, and an anti-interference parameter sequence is generated through iterative correction; The anti-interference parameter sequence is input into the parameter optimization function, and the energy distribution is constrained by combining the historical data of the material surface reflectivity to output the first laser processing parameter set that fully matches the multi-layer drawing model.

[0011] Optionally, the marking process of the metal material by the first laser according to the multi-layer drawing model, collecting processing feedback information of the metal material by a sensor during the marking process, and transmitting the feedback information to the edge processing module for processing response data analysis specifically includes the following steps: Based on the multi-layer drawing model and the first laser processing parameter set, the first laser is driven to perform layered scanning and marking on the surface of the positioned metal material, and a sensing module is simultaneously activated; the sensing module includes a coaxial optical sensor, an infrared sensing module, and a three-dimensional profile sensor; The dynamic distribution of surface reflectivity in the laser action area is collected in real time through a coaxial optical sensor, while the material temperature gradient change data is monitored through an infrared sensor module. A three-dimensional profile sensor is used to obtain the topographic data of the marking pattern layer by layer, and the reflectivity, temperature and topographic data are integrated into a synchronous processing feedback matrix.

[0012] Optionally, the method further includes the following steps: acquiring the topographic data of the marking pattern layer by layer using a three-dimensional profile sensor, fusing the reflectivity, temperature and topographic data into a synchronous processing feedback matrix; and The processing feedback matrix is input into the edge processing module, and the basic physical feature set and the abnormal fluctuation feature set are separated by the feature decoupling engine; Performing regression analysis on the basic physical feature set to generate a process stability assessment, while running a predictive model on the abnormal fluctuation feature set to output a deviation evolution trend; By integrating the processing stability evaluation and deviation evolution trend, the laser parameter correction amount is calculated through a hybrid optimization function to generate a processing response data packet transmitted to the downstream module.

[0013] Optionally, based on the difference between the processing response data and the image file data, a single-layer image file for correcting the processing deviation is generated through image processing and neural network recognition methods, specifically comprising the following steps: Analyze the laser parameter correction amount and deviation evolution trend in the processing response data packet, extract the spatial coordinate difference characteristics with the image data, and generate a multi-dimensional difference map; Input the multidimensional difference map into a pre-trained graph neural network model, identify the deviation type and spatial transmission path through association mapping, and output a deviation classification matrix containing position weight coefficients; Based on the deviation classification matrix, a dynamic rule engine is called to match the preset correction strategy library to generate a set of drawing element correction rules for deviations such as contour distortion, size offset and thermal deformation; According to the drawing element correction rule set, a structural constraint reconstruction operation is performed on the drawing data to generate a single-layer drawing that retains the reference relationship and compensates for the accumulated deviation to correct the processing deviation.

[0014] Optionally, optimizing the processing parameters of the second laser according to the processing feedback information to form a correction parameter set and a refined drawing suitable for fine surface processing specifically includes the following steps: Fusing the geometric compensation information of the corrected single-layer drawing with the thermodynamic history data in the processing feedback information to construct a surface energy field coupling model; The energy field coupling model is analyzed by a micro-area feature migration algorithm to generate an optimization objective function set for edge sharpening, texture uniformity, and residual stress control; Inputting the optimization objective function set into a dynamic constraint solver to match the physical parameter boundary conditions of the second laser, and outputting a modified parameter set including a power gradient distribution and a scanning path sequence; Adaptive reconstruction is performed on the single-layer image file based on the correction parameter set to generate a refined image file that retains the compensated geometric features and meets the energy regulation requirements.

[0015] The present invention also provides a surface treatment system, which adopts the laser-based surface treatment method as described above, wherein the surface treatment system includes an automatic adjustment device, wherein a driving end of the automatic adjustment device is provided with a jig; A sensor module, a first laser and a second laser are arranged above the fixture; and a control module, wherein the control module is provided with a storage unit, and the storage unit is used for pre-storing image data.

[0016] Compared with the prior art, the present invention has the following beneficial effects: first, the image data of the metal material to be processed is obtained and layered to form a multi-layer image model containing spatial structure information; then, the matching first laser processing parameters are calculated through a preset parameter library in combination with the physical properties of the metal material; in the processing preparation stage, the metal material is fixed on the fixture of the automatic adjustment device and its position is adjusted to the quasi-focus of the first laser; then, the first laser is started to perform marking processing according to the image model, and at the same time, the sensor collects feedback information including reflectivity, temperature and pattern contour in real time and inputs it into the edge processing module for analysis; finally, based on the difference between the actual processing feedback and the original image, a single-layer image is generated through image processing and neural network algorithm to correct the processing error, providing a basis for subsequent fine processing; this solution realizes real-time monitoring and response control of the processing process by deeply integrating the laser marking process with data feedback analysis, effectively improving the accuracy and consistency of pattern formation; at the same time, the correction mechanism based on image recognition and neural network can significantly reduce the processing deviation caused by differences in metal surface properties, thereby improving the overall surface treatment quality and product yield. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. 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 paying any creative work.

[0018] The structures, proportions, sizes, etc. depicted in the drawings of this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with this technology. They are not intended to limit the conditions under which the present invention can be implemented and therefore have no substantive technical significance. Any structural modifications, changes in proportional relationships, or adjustments in size should still fall within the scope of the technical contents disclosed in the present invention without affecting the effects and objectives that can be achieved by the present invention.

[0019] Figure 1 This is one of the flow charts of the laser-based surface treatment method of the first embodiment; Figure 2 This is the second flow chart of the laser-based surface treatment method of the first embodiment; Figure 3 This is a schematic structural diagram of the surface treatment equipment of the second embodiment. DETAILED DESCRIPTION

[0020] In order to make the purpose, features, and advantages of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described below 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 ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0021] In the description of the present invention, it should be understood that the terms "upper," "lower," "top," "bottom," "inner," "outer," etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings and are intended solely to facilitate the description of the present invention and simplify the description. They are not intended to indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limitations on the present invention. It should be noted that when a component is considered to be "connected" to another component, it may be directly connected to the other component or there may be a centrally located component.

[0022] The technical solution of the present invention will be further described below with reference to the accompanying drawings and through specific implementation methods.

[0023] Example 1: Combine Figure 1 and Figure 2 As shown, an embodiment of the present invention provides a laser-based surface treatment method, comprising the following steps: S1 acquires the image data of the metal material to be processed and performs graphic layering processing on it to generate a multi-layered image model that includes spatial structure information. A data input interface receives 2D vector graphics or 3D modeling data and performs a structured analysis of the geometric features of the original image. Using a graphic layering algorithm (such as path contours and fill area zoning annotations), a multi-layered spatial model containing depth information is reconstructed. This layering process provides a hierarchical benchmark for subsequent laser processing parameter matching. For example, the path contour layer is used to define edge processing priorities, and the structure fill layer is used to optimize material removal strategies.

[0024] S2, based on the metal material type, thickness, and surface properties, calculates and generates the first set of laser processing parameters that match the multi-layer pattern based on a preset parameter library. Initial laser parameters are calculated based on the metal material type (such as stainless steel, aluminum alloy), thickness, and surface properties (such as roughness and reflectivity) using data such as thermal conductivity and absorption coefficient stored in the preset parameter library, combined with characteristics such as the minimum line width and aspect ratio of the pattern's spatial structure. This stage dynamically matches material properties with pattern complexity to ensure that parameters such as laser energy density and scanning speed are compatible with processing requirements, reducing processing defects caused by heat accumulation or insufficient energy.

[0025] S3, place the metal material on a fixture equipped with an automatic adjustment device, and adjust the position of the fixture in sequence through the displacement control module so that the surface of the metal material is in the quasi-focal position of the first laser; use the displacement control module of the automatic adjustment fixture to achieve positioning, ensure that the material surface is stably in the quasi-focal area of the laser, and compensate for the focus drift caused by mechanical vibration and thermal deformation through real-time feedback adjustment to ensure the stability of the laser energy density distribution and avoid fluctuations in processing quality caused by defocus.

[0026] S4, using the first laser to mark the metal material according to the multi-layer pattern model, and collecting the processing feedback information of the metal material through the sensor during the marking process, and transmitting the feedback information to the edge processing module for processing response data analysis; the feedback information includes surface reflectivity, temperature change and pattern contour morphology.

[0027] The first laser performs layer-by-layer scanning according to the multi-layered image model. Simultaneously, a coaxial optical sensor, infrared sensor module, and 3D profile sensor collect reflectivity, temperature variation, and topography data of the processing area. The edge processing module integrates and analyzes this feedback data, extracting dynamic features of the processing process (such as temperature fluctuation range and profile deviation trends) in real time. This provides a quantitative basis for subsequent corrections, thereby improving the response speed and anti-interference capabilities of process control.

[0028] S5, based on the difference between the processing response data and the drawing data, a single-layer drawing is generated for correcting the processing deviation through image processing and neural network recognition methods.

[0029] Image processing algorithms (such as edge contrast analysis and thermal deformation area location) and neural network models (such as deviation pattern recognition based on convolutional networks) are used to match the differences between processing response data and original drawings, identifying key deviation types (such as contour deformation and heat-affected zone offset). Based on the spatial distribution characteristics of the deviations, the geometric parameters of the single-layer drawing (such as compensation path offset and correction contour curvature) are dynamically adjusted to generate a corrected drawing with error compensation capabilities, thereby reducing the impact of accumulated errors on the final processing quality.

[0030] S6, optimizing the processing parameters of the second laser according to the processing feedback information to form a correction parameter set and a refined image file suitable for fine surface processing; Combining the feedback information from step S4 (such as temperature history data and residual stress distribution) with the corrected image from step S5, a parameter optimization function (such as gradient descent or constrained programming) is used to generate a corrected parameter set suitable for the second laser, including power gradient distribution and scanning path rearrangement strategy. Simultaneously, the geometric features of the corrected single-layer image are fine-tuned (such as enhancing edge sharpness and homogenizing texture density) to produce a refined image that meets the requirements of fine processing and improves the surface quality of the secondary processing stage.

[0031] S7, using a second laser to perform surface treatment on the metal material after the first treatment according to the refined image file and the correction parameter set, and combining laser micro-control to perform refinement treatment to obtain a second treated material with high-precision pattern edges and surface consistency.

[0032] By leveraging the high-frequency, short-pulse characteristics of a second laser (e.g., a picosecond / femtosecond laser), combined with refined image files and corrected parameter sets, localized energy adjustments are performed on the surface after the initial processing. For example, a low-energy overlapping scanning strategy is employed to smooth out burrs or residual heat-affected zones on pattern edges. Dynamic power regulation is used to enhance energy distribution uniformity in areas with insufficient texture consistency. This stage significantly improves surface topography visual consistency, edge sharpness, and fatigue resistance, achieving the comprehensive goal of high-precision laser surface treatment.

[0033] The working principle of the present invention is as follows: first, the image data of the metal material to be processed is obtained and layered to form a multi-layer image model containing spatial structure information; then, the matching first laser processing parameters are calculated through a preset parameter library in combination with the physical properties of the metal material; in the processing preparation stage, the metal material is fixed on the fixture of the automatic adjustment device and its position is adjusted to the quasi-focus of the first laser; then, the first laser is started to perform marking processing according to the image model, and at the same time, the sensor collects feedback information including reflectivity, temperature and pattern contour in real time, and inputs it into the edge processing module for analysis; finally, based on the difference between the actual processing feedback and the original image, a single-layer image is generated through image processing and neural network algorithm to correct the processing error, providing a basis for subsequent fine processing; this solution realizes real-time monitoring and response control of the processing process by deeply integrating the laser marking process with data feedback analysis, effectively improving the accuracy and consistency of pattern formation; at the same time, the correction mechanism based on image recognition and neural network can significantly reduce the processing deviation caused by differences in metal surface properties, thereby improving the overall surface treatment quality and product yield.

[0034] In this embodiment, it is specifically explained that step S1 specifically includes the following steps: S11, collecting basic design drawings of the metal material to be processed, the basic design drawings including two-dimensional vector graphics or three-dimensional modeling data, and converting the structural contour information in the basic design drawings into a standardized format to generate a standard drawing file; Basic design drawing data, including 2D vector graphics (e.g., DXF, SVG) or 3D modeling data (e.g., STEP, IGES), is received through a data input interface (e.g., USB, Ethernet, or dedicated modeling software interface). After parsing the original drawing, structural contour information (e.g., boundary lines, hole dimensions, texture density) is extracted. Standardized conversion algorithms (e.g., coordinate system alignment, unit unification, and redundant data removal) are then used to generate a standard drawing file.

[0035] S12, based on the standard drawing file, calls the graphic analysis module to identify the geometric structure of the required processing area on the metal material surface, and preliminarily divides and marks the different functional areas; the partition marking includes the boundary area, the filling area and the high-density pattern area.

[0036] A graphics parsing module (such as a geometric feature recognition engine) analyzes the processing areas in the standard drawing. Using edge detection algorithms (such as the Sobel operator) and relationship analysis, it identifies boundary regions (such as contour lines and corner turning points). Fill regions are divided according to fill density thresholds (such as grid fill and crosshatch fill). High-density pattern areas are annotated based on pattern distribution characteristics (such as dot density and line spacing). This zoning and annotation process clarifies the processing priorities and parameter adaptation requirements for different functional areas. For example, prioritizing processing accuracy in boundary areas and optimizing energy allocation strategies in high-density pattern areas provides structured input for subsequent hierarchical modeling.

[0037] S13, using a hierarchical modeling algorithm to reconstruct the spatial structure of the annotated graphics according to a preset layer logic to form a multi-layered graphics model containing depth information, the multi-layered graphics model including at least three layers: a path outline layer, a structure filling layer, and a function adjustment layer; Use a layered modeling algorithm (such as spatial overlay or depth priority) to reconstruct the annotated graphics into a multi-layered drawing model containing depth information according to the preset layer logic (such as processing order from outside to inside): Path contour layer: defines the machining path of the outer contour and key geometric features, including boundary coordinates and contour closure information; Structural filling layer: records the topology of the internal filling area (such as grid spacing, intersection angle) and material removal strategy; Function adjustment layer: Contains fine-tuning parameters for specific functional areas (such as the scanning speed compensation coefficient for high-density pattern areas).

[0038] This hierarchical model provides a geometric benchmark for staged control of subsequent laser parameter matching. For example, the path profile layer can be used to set the laser's baseline scanning path, while the function adjustment layer performs dynamic energy regulation for complex areas.

[0039] S14, parameter annotation and integrity verification are performed on the multi-layer drawing model to ensure that each layer contains a clear processing area definition, spatial coordinate information and layer stacking order; after the verification is passed, the multi-layer drawing model is used as the basic graphic data input module for subsequent laser marking parameter matching and path planning.

[0040] Perform parameter annotation (such as the coordinate range of the processing area and the aspect ratio threshold) and logic verification on each layer of the multi-layer drawing model: Parameter annotation: The embedded verification circuit verifies the rationality of the layer stacking order (for example, the path outline layer must be at the bottom layer) and the coordinate accuracy (for example, whether there are negative values or out-of-limit values); Integrity check: Check the association between layers (such as whether the fill layer completely covers the outline layer area) and ensure that there are no logical conflicts (such as overlapping areas are not repeatedly marked).

[0041] In this embodiment, it is specifically explained that step S2 specifically includes the following steps: S21, analyzes the type, thickness, and surface roughness data of the metal material, calculates the absorption coefficient, thermal conductivity, and melting threshold through the material property quantification model, and generates a material property vector; Data on the metal material type (e.g., aluminum alloy, titanium alloy), thickness, and surface roughness is obtained using a material testing device (e.g., a spectrum analyzer or surface roughness meter) and then fed into a material property quantification model. This model, based on a database of material physical properties (e.g., absorption coefficient comparison tables, empirical formulas for thermal conductivity) and a melting threshold calculation module (e.g., based on the theoretical relationship between laser power and material phase transition temperature), converts the material's physical properties into numerical vectors (e.g., the components of the vector dimensions represent absorptivity, thermal conductivity, and melting threshold, respectively).

[0042] S22, extracting spatial structural information of the multi-layer drawing model, the spatial structural information including inter-layer correlation, minimum line width and aspect ratio features, and constructing a drawing feature vector; Extract key structural features from the multi-layer image model generated in step S1: Inter-layer correlation: Analyze the spatial nesting relationship between the path outline layer, fill layer, and function adjustment layer (such as whether the fill area completely covers the outline boundary); Minimum line width: Identify the finest processing path width in each layer (such as the line width threshold of the micro-hole edge); Aspect Ratio Feature: Calculates the ratio of pattern depth to width (e.g., aspect ratio constraint for groove structures).

[0043] Based on the above features, the image feature vector is constructed to quantify the spatial complexity of the model (such as the sensitivity of high aspect ratio regions to heat accumulation), providing structured input for differentiated matching of laser parameters.

[0044] S23, inputting the material characteristic vector and the image feature vector into a preset parameter library, and mapping out a basic laser parameter group through a multi-dimensional matching algorithm. The basic laser parameter group includes a pulse frequency range, a scanning speed range, and a power density reference value.

[0045] The material characteristic vector and the image feature vector are input into the preset parameter library (which stores the laser parameters and result data of historical processing cases). A multi-dimensional matching algorithm (such as cosine similarity calculation or principal component analysis) is used to search the parameter library for the benchmark parameter group that is closest to the current processing requirements. The basic parameters output by the matching include: Pulse frequency range: Determine the reasonable range of pulse interval based on the material melting threshold and line width constraint; Scan speed range: Dynamically adjust the laser beam movement rate based on aspect ratio and thermal conductivity to prevent excessive ablation; Power density baseline: Set the initial energy density based on the absorption coefficient and surface roughness to balance melting efficiency and heat-affected zone control.

[0046] This stage provides an initial benchmark for parameter optimization, avoiding the inefficiency caused by relying entirely on experience-based parameter tuning.

[0047] S24, based on the basic laser parameter group, calls the thermal-mechanical coupling simulation model to dynamically predict the processing deformation and heat-affected zone depth, and generates an anti-interference parameter sequence through iterative correction.

[0048] Based on a basic set of laser parameters, a thermal-mechanical coupling simulation model (such as a finite element simulation engine) is invoked, incorporating material parameters such as thermal conductivity and elastic modulus to dynamically predict potential local deformation (such as geometric distortion caused by thermal expansion) and the depth of the heat-affected zone (such as the thickness of the material phase transition layer) that may occur during machining. By iteratively modifying these basic parameters (such as reducing the scanning speed or power density in the predicted deformation area), a deformation-reducing parameter sequence (such as a segmented pulse frequency adjustment strategy or a gradient power compensation scheme) is generated to proactively prevent quality defects caused by heat accumulation during actual machining.

[0049] S25, inputting the anti-interference parameter sequence into the parameter optimization function, combining the material surface reflectivity historical data to constrain the energy distribution, and outputting the first laser processing parameter set that fully matches the multi-layer drawing model.

[0050] The anti-interference parameter sequence is input into a parameter optimization function (such as a constrained optimization model based on energy conservation). Combined with historical material surface reflectivity data (obtained from a coaxial sensor or a pre-stored database), the energy distribution is dynamically constrained. For example, a power density compensation factor can be added to high-reflectivity areas, or peak energy input can be limited in low-reflectivity areas. The optimization function outputs a first set of laser processing parameters that fully matches the multi-layer image model. This ensures precise alignment of processing energy and scanning paths in different areas (such as boundaries and fill layers), thereby achieving the goal of collaboratively optimizing thermal stability and topographic accuracy.

[0051] In this embodiment, it is specifically explained that step S4 specifically includes the following steps: S41, based on the multi-layer drawing model and the first laser processing parameter set, drives the first laser to perform layered scanning and marking on the surface of the positioned metal material, and simultaneously activates the sensing module; the sensing module includes a coaxial optical sensor, an infrared sensing module and a three-dimensional profile sensor.

[0052] Based on the processing positioning calibrated in step S3, the first laser is driven to perform layered scanning and marking according to the preset layers of the multi-layered drawing model (e.g., path contour layer → structure filling layer → function adjustment layer). A coaxial optical sensor integrated in the laser optical path, an infrared sensor module fixed to the side of the fixture, and a 3D profile sensor above the processing area are simultaneously activated to form a multi-dimensional data acquisition system. A timing controller is used at this stage to ensure synchronization between laser scanning and sensor data acquisition, preventing feedback data inaccuracies caused by signal delays.

[0053] S42, the dynamic distribution of surface reflectivity in the laser action area is collected in real time through a coaxial optical sensor, and the material temperature gradient change data is monitored through an infrared sensing module.

[0054] A coaxial optical sensor (such as a CMOS sensor embedded in the laser galvanometer optical path) captures the reflectivity distribution of the laser-affected area in real time, such as the spatial difference between high-reflectivity regions (such as unmelted areas) and low-reflectivity regions (such as oxidized or melted areas). Simultaneously, an infrared sensing module (such as a multi-point temperature measurement array) monitors the material surface temperature gradient (such as the peak temperature and diffusion rate in heat accumulation areas), providing quantitative input for dynamic assessment of the heat-affected zone.

[0055] S43 uses a three-dimensional profile sensor to obtain the topographic data of the marking pattern layer by layer, and fuses the reflectivity, temperature and topographic data into a synchronized processing feedback matrix.

[0056] After each layer is marked, a 3D profile sensor (such as a white-light interferometer) acquires 3D topographic data (such as depth error and edge steepness) of the processed pattern through non-contact scanning. Reflectivity distribution data, temperature gradient change data, and topographic data are aligned according to spatial coordinates and integrated into a processing feedback matrix that incorporates temporal and spatial correlations. This matrix incorporates the coupled characteristics of physical fields (temperature, reflectivity) and geometric fields (topographic deviation), providing a multi-dimensional data foundation for subsequent deviation analysis and correction.

[0057] S44, inputting the processing feedback matrix into the edge processing module, and separating the basic physical feature set and the abnormal fluctuation feature set through the feature decoupling engine; The edge processing module decomposes the processing feedback matrix using a feature decoupling engine (such as a principal component analysis algorithm or an independent component analysis module), extracting a set of basic physical features consistent with the preset processing objectives (such as expected reflectivity mean values and temperature fluctuation ranges) and a set of abnormal fluctuation features that deviate from expectations (such as reflectivity mutation points and abnormal temperature peaks). This decoupling operation breaks down complex data into independently analyzable feature units, reducing the computational complexity of subsequent processing.

[0058] S45, performing regression analysis on the basic physical feature set to generate a processing stability assessment, while running a prediction model on the abnormal fluctuation feature set to output a deviation evolution trend; A multivariate linear regression analysis is performed on the basic physical feature set to generate a processing stability score (e.g., whether the reflectivity variance is within a threshold or the temperature rise rate is stable) to determine processing consistency under the current parameters. Simultaneously, a timing prediction model (e.g., a long-short-term memory network model) is run on the abnormal fluctuation feature set (e.g., localized topography collapse trends) to output the evolution trend of the deviation (e.g., the degree of thermal deformation accumulation and the contour error diffusion path). This stage uses quantitative evaluation and trend prediction to clarify the priority of the correction strategy.

[0059] S46, integrates the processing stability evaluation and the deviation evolution trend, calculates the laser parameter correction through the hybrid optimization function, and generates a processing response data packet transmitted to the downstream module.

[0060] The processing stability evaluation results and the deviation evolution trend are input into the hybrid optimization function (such as a combination of a linear weighted objective function and constraint conditions) to calculate the laser parameter correction: Parameter correction types: including pulse frequency compensation (to suppress temperature overshoot), scanning speed adjustment (to compensate for morphology collapse) and power density gradient setting (to balance reflectivity differences); Dynamic allocation of correction parameters: Generate a spatial distribution scheme for correction parameters based on the deviation type and evolution trend weight (e.g., morphology deviation takes priority over temperature fluctuation).

[0061] Finally, a processing response data packet containing correction parameters and execution timing is generated and transmitted to the downstream control module to drive real-time parameter adjustment.

[0062] In this embodiment, it is specifically explained that step S5 specifically includes the following steps: S51: Analyze the laser parameter corrections and deviation evolution trends in the processing response data packet, extract the spatial coordinate difference characteristics from the image data, and generate a multi-dimensional difference map. By analyzing the laser parameter corrections (such as power compensation coefficients and scanning speed adjustment values) and deviation evolution trends (such as thermal deformation diffusion direction and dimensional offset accumulation) in the processing response data packet, the spatial coordinate difference characteristics (such as path contour offset coordinate points and fill area error) compared to the original image data are extracted. The difference characteristics are multi-dimensionally encoded according to geometric position, deviation type (such as lateral / longitudinal offset, energy excess / deficit), and time evolution stage, generating a difference map that includes position correlation and deviation magnitude.

[0063] S52: Input the multidimensional difference map into a pre-trained graph neural network model. The model uses association mapping to identify deviation types and spatial transmission paths, and outputs a deviation classification matrix containing position weight coefficients. The model also uses a pre-trained graph neural network model (a network model trained based on historical processing data, where nodes represent spatial locations and edge weights represent deviation correlations) to extract local and global deviation correlation features (such as the chain transmission path of contour distortion and the diffusion correlation of thermal deformation) through graph convolution operations. The network outputs a deviation classification matrix containing deviation classification labels (such as type A for dimensional offset and type B for contour distortion) and spatial weight coefficients (such as the correction priority of critical path nodes).

[0064] S53, based on the deviation classification matrix, calls the dynamic rule engine to match the preset correction strategy library to generate a set of drawing element correction rules for deviations such as contour distortion, size offset and thermal deformation.

[0065] Based on the deviation type labels and weight coefficients in the deviation classification matrix, a dynamic rule engine (such as a knowledge graph-driven reasoning module) is called to query the preset correction strategy library and match the rule set by priority: Contour distortion: Curvature compensation and boundary resampling strategies of the path contour layer are used; Size offset: adjust the scanning spacing and power density gradient of the filling layer; Thermal deformation: superimpose reverse compensation paths or heat sink areas to suppress energy input.

[0066] The generated correction rule set contains geometric adjustment parameters and energy control thresholds for different deviation types, ensuring precise adaptation of the strategy to the deviation characteristics.

[0067] S54, performing a structural constraint reconstruction operation on the drawing data according to the drawing element correction rule set, and generating a single-layer drawing that retains the reference relationship and compensates for the accumulated deviation while correcting the processing deviation.

[0068] Based on the correction rule set, perform spatial constraint reconstruction on the original image data: Datum relationship maintenance: Maintain key geometric relationships of the processing area (such as boundary closure and adjacent area connectivity) through topological consistency detection algorithms.

[0069] Deviation compensation embedding: On the basis of retaining the reference relationship, the parameters of the drawing elements are adjusted according to the correction rules (such as coordinate offset compensation of the path contour layer and line spacing shrinkage compensation of the fill layer).

[0070] The final output is a corrected single-layer drawing, the geometric structure of which can not only eliminate the cumulative impact of the current deviation, but also ensure compatibility with the undeformed area, providing optimized benchmark data for subsequent secondary processing.

[0071] In this embodiment, it is specifically explained that step S6 specifically includes the following steps: S61 integrates the geometric compensation information of the corrected single-layer image with the thermodynamic history data from the processing feedback to construct a surface energy field coupling model. This model is based on the material heat conduction equation and the laser energy distribution function. This model uses the compensated geometric features as a spatial basis and overlays thermodynamic history data (such as temperature field decay curves) as dynamic constraints to quantify the correlation between energy input, thermal deformation, and residual stress.

[0072] S62, analyzing the energy field coupling model through the micro-area feature migration algorithm to generate a set of optimization objective functions for edge sharpening, texture uniformity, and residual stress control; Feature migration algorithm: Using the micro-area feature migration algorithm (such as regional feature correlation matching based on graph theory), the energy field coupling model is decomposed into several micro-area units, and the energy control target of each unit is analyzed: Edge sharpening: defining energy gradient constraints in high-steepness topography areas (such as pattern edges) to suppress edge blurring caused by melt diffusion; Texture uniformity: constructing a uniform energy distribution function in repetitive texture areas to reduce density differences caused by scanning speed fluctuations; Residual stress control, introduce heat sink compensation parameters in high stress sensitive areas, limit laser energy peak and action time.

[0073] Objective function construction, generating a set of multi-objective optimization functions based on the analytical results, such as minimizing local stress gradients and maximizing texture uniformity indicators as optimization directions.

[0074] S63, inputting the optimization objective function set into the dynamic constraint solver, matching the physical parameter boundary conditions of the second laser, and outputting a modified parameter set including the power gradient distribution and the scanning path sequence; Constraint matching and solving: The optimization objective function set is input into a dynamic constraint solver (such as a nonlinear programming solver). Combined with the physical boundary conditions of the second laser (such as the maximum power threshold, the minimum spot size limit, and the maximum acceleration constraint of the scanning galvanometer), the parameter feasible domain is solved using the Lagrange multiplier method or the penalty function method.

[0075] Generate a modified parameter set based on the solution results: Power gradient distribution, defining a power density grading strategy based on micro-area energy requirements (e.g., high power in edge areas, medium power in texture areas); Scan path sequence and optimize path planning order to balance heat accumulation (e.g., using block jump scanning strategy to reduce local temperature rise).

[0076] S64, performing adaptive reconstruction on the single-layer image file based on the correction parameter set to generate a refined image file that retains the compensated geometric features and meets the energy control requirements.

[0077] Geometric feature preservation: based on the energy control requirements of the correction parameter set (such as power gradient distribution), local adjustments are made to the geometric structure of the single-layer correction image (such as extending the path overlap rate corresponding to the residence time in the high-power area).

[0078] Energy-adaptive reconstruction precisely matches geometric features (such as compensated edge curvature) with laser parameters (such as spot diameter) in space through coordinate scaling or path spacing adjustment. The resulting refined image ensures the simultaneous adaptation of geometric compensation features and energy input strategies.

[0079] Example 2: Combine Figure 3 As shown, the present invention also provides a surface treatment system, which adopts the laser-based surface treatment method as in Example 1. The surface treatment system includes an automatic adjustment device 10, a driving end of the automatic adjustment device 10 is provided with a fixture 20; a sensor module 30, a first laser 40 and a second laser 50 are provided above the fixture 20; and a control module, the control module is provided with a storage unit, and the storage unit is used to pre-store image data.

[0080] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A surface treatment method based on laser, characterized in that, The following steps are involved: Acquiring drawing data of the metal material to be processed, and performing graphic layering processing on the drawing data to generate a multi-layer drawing model including spatial structure information; In combination with the type, thickness and surface characteristics of the metal material, a first laser processing parameter set matching the multi-layer pattern is calculated and generated based on a preset parameter library; Placing the metal material on a jig provided with an automatic adjustment device, and sequentially adjusting the position of the jig through a displacement control module so that the surface of the metal material is at the quasi-focal position of the first laser; Marking the metal material according to the multi-layer pattern using a first laser, and collecting processing feedback information of the metal material using a sensor during the marking process, and transmitting the feedback information to an edge processing module for processing response data analysis; the feedback information includes surface reflectivity, temperature change, and pattern profile; According to the difference between the processing response data and the drawing data, a single-layer drawing for correcting the processing deviation is generated through image processing and neural network recognition methods.

2. The laser-based surface treatment method according to claim 1, characterized in that According to the difference between the processing response data and the drawing data, a single-layer drawing is generated for correcting the processing deviation through image processing and neural network recognition methods, and then the method further includes: Optimizing the processing parameters of the second laser according to the processing feedback information to form a correction parameter set and a refined image file suitable for fine surface processing; A second laser is used to perform surface treatment on the metal material after the first treatment according to the refined image file and the correction parameter set, and combined with laser micro-control to perform refinement treatment to obtain a second treated material with high-precision pattern edges and surface consistency.

3. The laser-based surface treatment method according to claim 1, characterized in that: The method of obtaining the drawing data of the metal material to be processed and performing graphic layering processing on the drawing data to generate a multi-layer drawing model including spatial structure information specifically includes the following steps: Collecting basic design drawings of the metal material to be processed, the basic design drawings include two-dimensional vector graphics or three-dimensional modeling data, and performing format standardization conversion on the structural contour information in the basic design drawings to generate standard drawing files; Based on the standard drawing file, the graphic analysis module is called to identify the geometric structure of the required processing area on the metal material surface, and preliminary zoning and marking are performed according to different functional areas; the zoning marking includes boundary areas, fill areas, and high-density pattern areas; A layered modeling algorithm is used to reconstruct the spatial structure of the annotated graphics according to the preset layer logic to form a multi-layered image model containing depth information. The multi-layered image model includes at least three layers: a path outline layer, a structure filling layer, and a function adjustment layer. The multi-layer drawing model is parameterized and integrity checked to ensure that each layer contains a clear definition of the processing area, spatial coordinate information, and layer stacking order.

4. The laser-based surface treatment method according to claim 1, characterized in that In combination with the type, thickness and surface characteristics of the metal material, a first laser processing parameter set matching the multi-layer pattern is calculated and generated based on a preset parameter library, specifically comprising the following steps: Analyzing the type, thickness, and surface roughness data of the metal material, calculating the absorption coefficient, thermal conductivity, and melting threshold using a material property quantification model, and generating a material property vector; Extracting spatial structural information of the multi-layer drawing model, the spatial structural information including inter-layer correlation, minimum line width and aspect ratio features, and constructing a drawing feature vector; The material characteristic vector and the image feature vector are input into a preset parameter library, and a basic laser parameter group is mapped out through a multi-dimensional matching algorithm. The basic laser parameter group includes a pulse frequency range, a scanning speed range and a power density reference value.

5. The laser-based surface treatment method according to claim 4, characterized in that: The material characteristic vector and the image feature vector are input into a preset parameter library, and a basic laser parameter group is mapped out through a multi-dimensional matching algorithm, and then the following steps are included: Based on the basic laser parameter group, a thermal-mechanical coupling simulation model is called to dynamically predict the machining deformation and the depth of the heat-affected zone, and an anti-interference parameter sequence is generated through iterative correction; The anti-interference parameter sequence is input into the parameter optimization function, and the energy distribution is constrained by combining the historical data of the material surface reflectivity to output the first laser processing parameter set that fully matches the multi-layer drawing model.

6. The laser-based surface treatment method according to claim 1, characterized in that The method of marking the metal material according to the multi-layer drawing model by the first laser, collecting processing feedback information of the metal material by the sensor during the marking process, and transmitting the feedback information to the edge processing module for processing response data analysis specifically includes the following steps: Based on the multi-layer drawing model and the first laser processing parameter set, the first laser is driven to perform layered scanning and marking on the surface of the positioned metal material, and a sensing module is simultaneously activated; the sensing module includes a coaxial optical sensor, an infrared sensing module, and a three-dimensional profile sensor; The dynamic distribution of surface reflectivity in the laser action area is collected in real time through a coaxial optical sensor, while the material temperature gradient change data is monitored through an infrared sensor module. A three-dimensional profile sensor is used to obtain the topographic data of the marking pattern layer by layer, and the reflectivity, temperature and topographic data are integrated into a synchronous processing feedback matrix.

7. The laser-based surface treatment method according to claim 6, characterized in that: The method uses a three-dimensional profile sensor to obtain the topographic data of the marking pattern layer by layer, and fuses the reflectivity, temperature and topographic data into a synchronous processing feedback matrix, and then further includes the following steps: The processing feedback matrix is input into the edge processing module, and the basic physical feature set and the abnormal fluctuation feature set are separated by the feature decoupling engine; Performing regression analysis on the basic physical feature set to generate a process stability assessment, while running a predictive model on the abnormal fluctuation feature set to output a deviation evolution trend; By integrating the processing stability evaluation and deviation evolution trend, the laser parameter correction amount is calculated through a hybrid optimization function to generate a processing response data packet transmitted to the downstream module.

8. The laser-based surface treatment method according to claim 1, characterized in that: According to the difference between the processing response data and the drawing data, a single-layer drawing for correcting the processing deviation is generated through image processing and neural network recognition methods, which specifically includes the following steps: Analyze the laser parameter correction amount and deviation evolution trend in the processing response data packet, extract the spatial coordinate difference characteristics with the image data, and generate a multi-dimensional difference map; Input the multidimensional difference map into a pre-trained graph neural network model, identify the deviation type and spatial transmission path through association mapping, and output a deviation classification matrix containing position weight coefficients; Based on the deviation classification matrix, a dynamic rule engine is called to match the preset correction strategy library to generate a set of drawing element correction rules for deviations such as contour distortion, size offset and thermal deformation; According to the drawing element correction rule set, a structural constraint reconstruction operation is performed on the drawing data to generate a single-layer drawing that retains the reference relationship and compensates for the accumulated deviation to correct the processing deviation.

9. The laser-based surface treatment method according to claim 2, characterized in that: Optimizing the processing parameters of the second laser according to the processing feedback information to form a correction parameter set and a refined drawing file suitable for fine surface processing specifically includes the following steps: Fusing the geometric compensation information of the corrected single-layer drawing with the thermodynamic history data in the processing feedback information to construct a surface energy field coupling model; The energy field coupling model is analyzed by a micro-area feature migration algorithm to generate an optimization objective function set for edge sharpening, texture uniformity, and residual stress control; Inputting the optimization objective function set into a dynamic constraint solver to match the physical parameter boundary conditions of the second laser, and outputting a modified parameter set including a power gradient distribution and a scanning path sequence; Adaptive reconstruction is performed on the single-layer image file based on the correction parameter set to generate a refined image file that retains the compensated geometric features and meets the energy regulation requirements.

10. A surface treatment system, characterized in that: The laser-based surface treatment method according to any one of claims 1 to 9 is adopted, wherein the surface treatment system includes an automatic adjustment device, and a jig is provided at a driving end of the automatic adjustment device; A sensor module, a first laser and a second laser are provided above the fixture; and a control module, wherein the control module is provided with a storage unit, and the storage unit is used for pre-storing image data.

Images