A method, device, equipment and medium for monitoring quality of a metal surface treatment process

Abstract

The application discloses a kind of quality monitoring methods, devices, equipment and medium of metal surface treatment process, and the application belongs to computer technical field.The method comprises: obtaining the surface three-dimensional data of metal object;According to the surface three-dimensional data and cladding parameter, determine the target position information after cladding;In the process of laser cladding to the metal object, obtain actual position information after cladding and surface regularity information after cladding;Identify whether the actual position information and the target position information are consistent to carry out the quality monitoring of cladding thickness, and identify whether the surface regularity information after cladding meets target regularity to carry out the quality monitoring of cladding regularity.This technical solution can be in real-time monitoring and multidimensional quality evaluation during cladding process, reduce rework, improve the timeliness and comprehensiveness of cladding quality monitoring.

Description

Technical Field

[0001] This application belongs to the field of computer technology, and specifically relates to a method, device, equipment and medium for quality monitoring of metal surface treatment processes. Background Technology

[0002] With the gradual development of computer technology, more and more technologies are being applied to solve problems in industrial production, which has greatly improved the efficiency of industrial production. For example, in the scenario of laser cladding of metal components, quality monitoring is a core link because the cladding quality directly affects the performance of the components. With the development of computer processing technology, existing monitoring has shifted from manual sampling to an automated mode, judging the quality of the cladding results such as thickness and flatness after the cladding is completed.

[0003] However, the existing technology has obvious shortcomings: First, the detection of cladding quality is somewhat delayed. Even if quality problems are found in the cladding results, rework is required to complete the cladding process. Second, the evaluation of monitoring results is somewhat singular, focusing only on surface characteristics and not on the impact of the cladding quality on the actual function of the cladding layer during the cladding process. Summary of the Invention

[0004] The purpose of this application is to provide a method, apparatus, equipment, and medium for quality monitoring of metal surface treatment processes. The aim is to address the problems in existing technologies where delayed cladding quality detection leads to rework, and monitoring and evaluation focus only on surface features while ignoring process impacts. This technical solution acquires three-dimensional data of the metal surface and combines it with cladding parameters to determine target location information. During the cladding process, it acquires the actual location and surface regularity information in real time and compares it with the target. It sets filling fields and important monitoring areas for pits and cracks and densifies the monitoring grid. By introducing molten pool temperature distribution information and a trained cladding quality assessment model, it achieves real-time monitoring and multi-dimensional quality assessment of the cladding process, reducing rework and improving the timeliness and comprehensiveness of cladding quality monitoring.

[0005] In a first aspect, embodiments of this application provide a quality monitoring method for a metal surface treatment process, the method comprising:

[0006] Acquire three-dimensional surface data of a metallic object;

[0007] Based on the surface three-dimensional data and cladding parameters, determine the target location information after cladding;

[0008] During the laser cladding process on the metal object, the actual position information and surface regularity information after cladding are obtained;

[0009] The system identifies whether the actual location information matches the target location information to monitor the cladding thickness, and identifies whether the surface regularity information after cladding meets the target regularity to monitor the cladding regularity.

[0010] Furthermore, based on the surface three-dimensional data and cladding parameters, the target location information after cladding is determined, including:

[0011] Based on the three-dimensional surface data, determine whether there are pits on the surface of the metal object;

[0012] If a pit exists, a fill field is added to the spatial parameters of the three-dimensional point at the pit to write the fill value from the spatial position of the three-dimensional point to the original spatial parameters.

[0013] Based on the spatial location of the three-dimensional points, the filling value, and the cladding parameters, the target location information after cladding is determined;

[0014] or,

[0015] Based on the three-dimensional surface data, determine whether there are cracks on the surface of the metal object;

[0016] If a crack exists, a fill field is added to the spatial parameters of the three-dimensional point at the crack to write the fill value from the spatial position of the three-dimensional point to the original spatial parameters, as well as the crack width information.

[0017] Based on the spatial location of the three-dimensional point, the filling value, the crack width information, and the cladding parameters, the target location information after cladding is determined.

[0018] Furthermore, after acquiring the three-dimensional surface data of the metal object, the method further includes:

[0019] If a pit or crack is detected in the metal object, the location of the defect is recorded.

[0020] The laser cladding process for the metal object also includes:

[0021] Based on the location of the defect, an important monitoring area is determined, and the flatness of the important monitoring area is tested.

[0022] Furthermore, based on the location of the defect, important monitoring areas are determined, including:

[0023] The location and shape of the defect are input into the impact range assessment algorithm, and the important monitoring area is determined based on the output of the impact range assessment algorithm.

[0024] Furthermore, after determining the important monitoring area based on the defect location, the method further includes:

[0025] Obtain the monitoring grid accuracy for a typical area;

[0026] Based on the monitoring grid accuracy of the regular area, the grid is encrypted according to a preset method to obtain the monitoring grid accuracy of the important monitoring area.

[0027] Furthermore, the method also includes:

[0028] During the laser cladding process on the metal object, information on the temperature distribution of the molten pool is obtained;

[0029] The molten pool temperature distribution information is input into the cladding quality assessment model. Based on the output of the cladding quality assessment model, the quality monitoring result is determined. The cladding quality assessment model is trained based on historical sample data, which consists of data pairs composed of molten pool temperature distribution information and cladding results.

[0030] Furthermore, the historical sample data includes positive samples and negative samples;

[0031] The positive sample data, which is the cladding result corresponding to the molten pool temperature distribution information, meets the quality requirements.

[0032] The negative sample data refers to the cladding results corresponding to the molten pool temperature distribution information, which do not meet the quality requirements.

[0033] Secondly, embodiments of this application provide a quality monitoring device for a metal surface treatment process, the device comprising:

[0034] The surface 3D data acquisition module is used to acquire the surface 3D data of metal objects;

[0035] The target location information determination module is used to determine the target location information after cladding based on the surface three-dimensional data and cladding parameters.

[0036] The actual information acquisition module is used to acquire the actual position information and surface regularity information after laser cladding of the metal object during the laser cladding process.

[0037] The quality monitoring module is used to identify whether the actual location information is consistent with the target location information in order to monitor the quality of the cladding thickness, and to identify whether the surface regularity information after cladding meets the target regularity in order to monitor the quality of the cladding regularity.

[0038] Thirdly, embodiments of this application provide an electronic device including a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the method described in the first aspect.

[0039] Fourthly, embodiments of this application provide a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the first aspect.

[0040] Fifthly, embodiments of this application provide a chip, the chip including a processor and a communication interface, the communication interface being coupled to the processor, the processor being used to run programs or instructions to implement the method as described in the first aspect.

[0041] The technical solution provided in this application acquires three-dimensional surface data of a metal object; determines the target position information after cladding based on the three-dimensional surface data and cladding parameters; during the laser cladding process on the metal object, acquires the actual position information and surface regularity information after cladding; identifies whether the actual position information is consistent with the target position information for quality monitoring of cladding thickness, and identifies whether the surface regularity information after cladding conforms to the target regularity for quality monitoring of cladding regularity. This technical solution, by acquiring three-dimensional surface data of the metal and combining it with cladding parameters to determine the target position information, acquires the actual position and surface regularity information in real time during the cladding process and compares it with the target, sets filling fields and important monitoring areas for pits and cracks and densifies the monitoring grid, introduces molten pool temperature distribution information and a trained cladding quality assessment model, realizing real-time monitoring and multi-dimensional quality assessment of the cladding process, reducing rework, and improving the timeliness and comprehensiveness of cladding quality monitoring. Attached Figure Description

[0042] Figure 1 This is a flowchart illustrating the quality monitoring method for the metal surface treatment process provided in Embodiment 1 of this application;

[0043] Figure 2 This is a flowchart illustrating the quality monitoring method for the metal surface treatment process provided in Embodiment 2 of this application;

[0044] Figure 3 This is a schematic diagram of the structure of the quality monitoring device for the metal surface treatment process provided in Embodiment 3 of this application;

[0045] Figure 4 This is a schematic diagram of the structure of the electronic device provided in Embodiment 4 of this application. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this application clearer, specific embodiments of this application will be described in further detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely for explaining this application and not for limiting it. It should also be noted that, for ease of description, only the parts relevant to this application are shown in the drawings, not all of them. Before discussing exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe operations (or steps) as sequential processes, many of these operations can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but may also have additional steps not included in the drawings. The process can correspond to a method, function, procedure, subroutine, subroutine, etc.

[0047] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0048] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0049] The following description, in conjunction with the accompanying drawings, details the quality monitoring method, apparatus, equipment, and medium for metal surface treatment processes provided in this application through specific embodiments and application scenarios.

[0050] Example 1

[0051] Figure 1 This is a schematic flowchart of the quality monitoring method for the metal surface treatment process provided in Embodiment 1 of this application. Figure 1 As shown, the specific steps include the following:

[0052] S101: Acquire the three-dimensional surface data of a metal object;

[0053] Understandably, the implementing entity of this solution can be a device used to evaluate the quality of surface treatment of metal objects, such as the surface treatment equipment itself, which may have a quality evaluation function, or other electronic devices, such as control devices connected to the surface treatment equipment, such as tablet computers, desktop computers, smartphones and servers connected to it.

[0054] Among them, metal objects refer to metal components that need to undergo laser cladding treatment, such as automobile engine crankshafts and wind power equipment main shafts.

[0055] Surface three-dimensional data refers to the three-dimensional coordinate information that can characterize the spatial morphology of the surface of a metal object, including the X-axis, Y-axis, and Z-axis coordinate values ​​of each point on the surface, which can reflect whether there are features such as protrusions, depressions, and cracks on the surface.

[0056] In this solution, data acquisition can be completed by industrial 3D scanning equipment, such as a laser line scanner or a structured light 3D scanner. The scanning equipment performs a full-coverage scan of the surface of the metal object according to a preset path, with a scanning accuracy of up to 0.01mm. After the scan is completed, the raw data is transmitted to the computer data processing module to form a complete 3D point cloud model of the surface.

[0057] S102: Determine the target location information after cladding based on the surface three-dimensional data and cladding parameters;

[0058] Among them, cladding parameters refer to the key process parameters that affect the cladding effect during laser cladding, including laser power (e.g., 1000-3000W), laser scanning speed (e.g., 5-20mm / s), powder feeding rate (e.g., 5-20g / min), protective gas flow rate (e.g., 5-15L / min), etc.

[0059] Target location information refers to the ideal spatial coordinate data that each point on the surface of the cladding layer should reach after the cladding is completed, which can be used as a benchmark for subsequent quality monitoring.

[0060] In this solution, the surface 3D data can be preprocessed first to remove noise points and optimize the point cloud model. Then, the preset cladding position calculation algorithm is called to couple the surface 3D data with the cladding parameters. For example, based on the powder feeding amount and scanning speed in the cladding parameters, the theoretical thickness of the cladding layer per unit length is calculated. Combined with the original 3D coordinates of the metal object surface, the theoretical thickness is superimposed to obtain the target spatial coordinates of each point. Finally, the target position information file after cladding is formed and stored.

[0061] The preset cladding position calculation algorithm can be based on usage requirements. For example, if the current radius of a shaft is 68mm, and the shaft is installed in a sleeve with an optimal radius of 70cm, the preset cladding position calculation algorithm can determine the target position information of the shaft after cladding.

[0062] S103: During the laser cladding process on the metal object, obtain the actual position information and the surface regularity information after cladding;

[0063] The laser cladding process refers to the process in which laser cladding equipment delivers cladding powder, such as stainless steel powder or nickel-based alloy powder, to the surface of a metal object according to a preset program, and uses laser energy to melt and solidify the powder and the substrate surface to form a cladding layer.

[0064] Actual location information refers to the spatial coordinates of each point on the surface of the cladding layer during the cladding process.

[0065] Surface regularity information refers to parameter information that characterizes the smoothness and continuity of the cladding layer surface, such as the fluctuation range of the surface profile and the height difference between adjacent areas.

[0066] In this solution, a real-time detection device is installed next to the cladding head of the laser cladding equipment. For example, a line laser profile sensor is used to obtain the actual position information. The sensor moves synchronously with the cladding head, lagging behind the cladding head by 50-100mm to avoid interference from the high-temperature molten pool on the detection. The sensor collects the three-dimensional profile data of the cladding layer surface in real time and converts it into actual position information.

[0067] Surface regularity information can be calculated from data collected by the same sensor. For example, the difference between the highest and lowest points in the contour data and the straightness deviation of the contour line can be extracted. Alternatively, a high-resolution industrial camera, such as a 20-megapixel camera, can be used to capture surface images and analyze the surface regularity characteristics through image grayscale values. Data obtained in both ways is transmitted to the computer processing module in real time.

[0068] S104: Identify whether the actual position information is consistent with the target position information in order to perform quality monitoring of the cladding thickness, and identify whether the surface regularity information after cladding meets the target regularity in order to perform quality monitoring of the cladding regularity.

[0069] Among them, target regularity refers to the preset standard for the regularity of the cladding layer surface, such as surface contour fluctuation amplitude ≤0.1mm, height difference between adjacent areas ≤0.05mm, etc., which is set according to the usage scenario of the metal component. For example, the target regularity requirement for precision mold cladding is higher than that for ordinary structural parts. Quality monitoring of cladding thickness refers to judging whether the actual thickness of the cladding layer meets the design requirements, and quality monitoring of cladding regularity refers to judging whether the surface morphology of the cladding layer meets the performance requirements.

[0070] In this solution, a data comparison algorithm can be invoked to compare the actual location information with the target location information point by point, calculating the difference between the two in the Z-axis direction (thickness direction). If the difference is within a preset allowable range (e.g., ±0.05mm), the cladding thickness is deemed acceptable; otherwise, it is deemed unacceptable and the abnormal area is marked. For surface regularity information, the calculated surface regularity parameters are matched with the target regularity. For example, if the surface contour fluctuation amplitude is ≤0.1mm, it meets the target regularity; otherwise, it does not. Simultaneously, a quality monitoring report is generated, displaying the monitoring results of thickness and regularity in real time.

[0071] The technical solution provided in this embodiment obtains three-dimensional data of the metal surface and determines the target location information by combining cladding parameters. During the cladding process, the actual location and surface regularity information are obtained in real time and compared with the target, realizing real-time monitoring and multi-dimensional quality assessment of the cladding process, reducing rework, and improving the timeliness and comprehensiveness of cladding quality monitoring.

[0072] In one embodiment, optionally, determining the target location information after cladding based on the surface three-dimensional data and cladding parameters includes:

[0073] Based on the three-dimensional surface data, determine whether there are pits on the surface of the metal object;

[0074] If a pit exists, a fill field is added to the spatial parameters of the three-dimensional point at the pit to write the fill value from the spatial position of the three-dimensional point to the original spatial parameters.

[0075] Based on the spatial location of the three-dimensional points, the filling value, and the cladding parameters, the target location information after cladding is determined;

[0076] or,

[0077] Based on the three-dimensional surface data, determine whether there are cracks on the surface of the metal object;

[0078] If a crack exists, a fill field is added to the spatial parameters of the three-dimensional point at the crack to write the fill value from the spatial position of the three-dimensional point to the original spatial parameters, as well as the crack width information.

[0079] Based on the spatial location of the three-dimensional point, the filling value, the crack width information, and the cladding parameters, the target location information after cladding is determined.

[0080] Among them, pits refer to the recessed areas formed on the surface of metal objects due to corrosion, collisions, etc., such as circular or irregular pits with a diameter of 2-10mm and a depth of 0.5-3mm.

[0081] The spatial parameters of a three-dimensional point refer to the X-axis, Y-axis, and Z-axis coordinate parameters of each point in the pit region of the surface three-dimensional point cloud.

[0082] The filling field refers to a newly added field in the spatial parameter data structure used to record information related to pit filling. The filling value refers to the additional cladding thickness value that needs to be added to the original surface height of the pit in order to make the pit area reach the target plane after cladding. For example, when the pit depth is 1mm, the filling value can be set to 1.2mm (leaving a machining allowance of 0.2mm).

[0083] In this solution, the computer data processing module first performs feature recognition on the surface 3D data, calls a pit detection algorithm, such as an algorithm based on point cloud curvature analysis, to calculate the curvature value of each region. Regions with abnormally large curvature values ​​are identified as potentially containing pits. This is then confirmed through manual verification or with the assistance of high-resolution images. If a pit is found, a filling field is added to the spatial parameter data structure of each 3D point in the pit region. Based on parameters such as the pit's depth and area, a filling value calculation model is used, for example, depth × 1.2 + area coefficient × 0.1, to calculate the filling value for each 3D point and write it into the filling field.

[0084] Finally, by combining the spatial location of the three-dimensional points, the filling value, and the cladding parameters, such as the thickness coefficient corresponding to the powder feeding amount, the target location information of each point is recalculated to ensure that the pitted area can be smoothly covered after cladding.

[0085] In another embodiment, a crack refers to a linear defect on the surface of a metal object caused by stress, fatigue, etc., such as a slender gap with a length of 5-50 mm and a width of 0.1-1 mm; crack width information refers to the maximum width value of the crack in the direction perpendicular to the crack direction, for example, the width information of a certain crack is 0.3 mm.

[0086] In this solution, a crack detection algorithm, such as one based on point cloud distance transformation, can be used to analyze the 3D surface data. Linear, continuous low-height areas are identified as potentially containing cracks. Ultrasonic testing is then used to confirm the actual length and width of the cracks. If a crack is found, a fill field is added to the spatial parameter data structure of each 3D point in the crack area. In addition to writing the fill value, the crack width information is calculated based on the crack depth (e.g., depth × 1.5). Combining the spatial location of the 3D points, the fill value, the crack width information, and the cladding parameters, such as increasing the laser power by 10% when the crack width is greater than 0.5mm, the calculation logic for the target location information is adjusted. For example, the target thickness is appropriately increased in areas with larger crack widths to ensure the crack is fully filled. Finally, the target location information after cladding is determined.

[0087] The technical solution provided in this embodiment addresses pits and cracks on the surface of metal objects. By adding a filling field to the three-dimensional point space parameters and introducing filling values ​​and crack width information, it achieves precise customization of the target location information of the defect area. This solves the problem in the prior art of lacking targeted target planning for the defect area, resulting in the continued existence of defects or uneven thickness of the cladding layer after cladding. It can ensure that the pits and cracks achieve the ideal surface morphology after cladding, thus improving the reliability of the cladding quality in the defect area.

[0088] In one embodiment, optionally, after acquiring the three-dimensional surface data of the metal object, the method further includes:

[0089] If a pit or crack is detected in the metal object, the location of the defect is recorded.

[0090] The laser cladding process for the metal object also includes:

[0091] Based on the location of the defect, an important monitoring area is determined, and the flatness of the important monitoring area is tested.

[0092] Among them, the defect location refers to the specific spatial coordinate range of the pit or crack on the surface of the metal object, such as the coordinate range of a circular area with a radius of 5mm with the center of the pit as the origin, or the coordinate range of a linear area formed by the coordinates of the two ends of the crack and the direction of the crack.

[0093] In this solution, after the defect identification of the surface 3D data is completed, the boundary coordinates of the defect area can be automatically extracted. For example, the defect area can be separated from the normal area by a point cloud segmentation algorithm, and the coordinate range of the minimum bounding rectangle or circle of the defect area can be determined. This coordinate range is stored in the database as the defect location information. At the same time, a defect location marker map is generated, and the defect location is marked in the 3D model with a special color to facilitate the location during subsequent monitoring.

[0094] Important monitoring areas refer to areas determined based on the location of defects that require special attention to the cladding quality. These areas are typically the defect location and a certain range around it, such as a 10mm annular area around the defect location.

[0095] Flatness testing refers to monitoring the flatness of the surface after cladding in a certain area, which is achieved by calculating parameters such as the height difference and flatness deviation of various points on the surface.

[0096] During laser cladding, important monitoring areas can be automatically defined based on stored defect location information. For example, the boundary of the defect location can be extended outward by 10mm to form an important monitoring area. A real-time detection device, such as a line laser profile sensor, is controlled to perform high-frequency detection on this area, with a detection frequency 1-2 times higher than that of conventional areas (e.g., 10Hz for conventional areas and 20Hz for important monitoring areas). Three-dimensional surface data of this area after cladding is collected, and flatness parameters are calculated. For example, if the flatness deviation is ≤0.08mm, an early warning signal is immediately triggered if the parameter exceeds the acceptable range, prompting the operator to adjust the cladding parameters.

[0097] The technical solution provided in this embodiment solves the problems of insufficient monitoring of defect areas and easy neglect of the cladding quality around defects in the prior art by recording the defect location and delineating important monitoring areas for special flatness detection. It achieves key control of the cladding quality of defect areas, can detect flatness abnormalities in the area around defects in a timely manner, avoids the impact of poor cladding quality around defects on the overall performance of the component, and improves the pertinence of cladding quality monitoring.

[0098] In one embodiment, optionally, determining the important monitoring area based on the defect location includes:

[0099] The location and shape of the defect are input into the impact range assessment algorithm, and the important monitoring area is determined based on the output of the impact range assessment algorithm.

[0100] Among them, the morphology of the defect refers to the specific shape and size parameters of the pit or crack, such as the diameter, depth, and whether it is a regular circle of the pit, and the length, width, and whether it has branches of the crack. The influence range assessment algorithm refers to the algorithm that can calculate the range of the surrounding area that the defect may affect during the cladding process based on the location and morphology of the defect, such as the influence range model algorithm based on the coupling of defect size and cladding parameters.

[0101] In this scheme, the morphology of the defect can be extracted first. For example, the diameter D and depth H of the pit, and the length L and width W of the crack can be obtained through image recognition algorithms. The coordinate data of the defect location and the extracted morphological parameters are then input into the influence range evaluation algorithm. The algorithm has preset influence coefficients for different defect types. For example, the influence range radius of the pit is D×1.5+H×2, and the influence range width of the crack is W×10 and the length is L+20mm. Combined with the laser action range in the cladding parameters, such as a laser spot diameter of 5mm, the coordinate range of the surrounding area that the defect may affect is calculated. This range is the important monitoring area. After the algorithm completes the calculation, it outputs the boundary coordinates of the important monitoring area for subsequent detection device positioning.

[0102] The technical solution provided in this embodiment introduces an impact range assessment algorithm to transform defect location and morphological parameters into quantifiable important monitoring area ranges. This solves the problem in the prior art where the delineation of important monitoring areas relies on manual experience and the range is inaccurate. It achieves accurate calculation of important monitoring areas, avoiding both the waste of resources caused by excessively large monitoring ranges and the omissions caused by excessively small monitoring ranges, thereby improving the scientificity and accuracy of the delineation of important monitoring areas.

[0103] In one embodiment, optionally, after determining the important monitoring area based on the defect location, the method further includes:

[0104] Obtain the monitoring grid accuracy for a typical area;

[0105] Based on the monitoring grid accuracy of the regular area, the grid is encrypted according to a preset method to obtain the monitoring grid accuracy of the important monitoring area.

[0106] Among them, the monitoring grid accuracy refers to the size of the grid unit used to collect data within the monitoring area. The smaller the grid unit, the higher the accuracy. For example, the monitoring grid accuracy of a regular area is 1mm×1mm, that is, the side length of each grid unit is 1mm, and the grid accuracy of an important monitoring area after densification is 0.5mm×0.5mm.

[0107] The preset method refers to the pre-set mesh precision encryption rules, such as reducing the side length by half on the basis of the normal precision, or setting the encryption multiplier according to the severity of the defect, such as 2 times encryption when the defect depth is greater than 2mm, 3 times encryption when the defect depth is greater than 3mm, and so on.

[0108] In this solution, the monitoring grid precision for the regular area can be obtained from the system parameter configuration, for example, if the regular area grid precision is read as 1mm×1mm. This precision is then densified according to a preset method. If the preset method is "regular precision side length reduced by half", the monitoring grid precision for the important monitoring area is calculated to be 0.5mm×0.5mm. If the preset method is "densification based on defect depth, doubling the density for every 1mm increase in depth", when the defect depth is 2mm, the density is doubled, and the regular precision of 1mm×1mm is densified to 0.5mm×0.5mm. The densified monitoring grid precision parameters are sent to the real-time detection device. The detection device divides the important monitoring area into grid units according to this precision and collects data from each grid unit, ensuring that the data collection density in the important monitoring area is higher than that in the regular area.

[0109] The technical solution provided in this embodiment solves the problem in the prior art that the important monitoring area and the regular area use the same grid precision, and the data collection in the important area is not detailed enough. It realizes high-density data collection in the important monitoring area, which can more accurately capture surface morphology changes in the important area and discover minute flatness anomalies that are difficult to detect under the conventional precision, thereby improving the accuracy of cladding quality monitoring in the important monitoring area.

[0110] Example 2

[0111] Figure 2 This is a schematic flowchart of the quality monitoring method for the metal surface treatment process provided in Embodiment 2 of this application. Figure 2 As shown, the specific steps include the following:

[0112] S201, Obtain the three-dimensional surface data of the metal object;

[0113] S202, Based on the surface three-dimensional data and cladding parameters, determine the target location information after cladding;

[0114] S203, during the laser cladding process on the metal object, the temperature distribution information of the molten pool is obtained;

[0115] The molten pool temperature distribution information refers to the real-time temperature values ​​and distribution characteristics at different spatial locations within the molten pool area during laser cladding. The molten pool area is a liquid metal region formed by the co-melting of the metal substrate and cladding powder under the action of laser energy. It is typically elliptical, with a major axis of 3-8 mm and a minor axis of 2-5 mm. The molten pool temperature distribution information includes not only the temperature values ​​of the molten pool center, edge, and heat-affected zone, but also the temperature gradient of each region, such as the temperature difference between the molten pool center and edge, and the temperature difference between the heat-affected zone and the substrate. For example, under certain operating conditions, the molten pool center temperature is 1750-1850℃, the molten pool edge temperature is 1200-1300℃, and the highest temperature in the heat-affected zone is 800-850℃, with the temperature gradient controlled at 450-550℃. This information can be stored and transmitted as a color temperature field cloud map (e.g., red for high-temperature areas, blue for low-temperature areas) or as a temperature data matrix. For example, a temperature matrix could be a 50×50 pixel array of temperature values, with each pixel corresponding to an actual area of ​​0.1 mm × 0.1 mm.

[0116] In this solution, the temperature distribution information of the molten pool can be acquired by fixing a high-precision dual-color infrared thermal imager, such as the FLIRA655sc model, to the side of the cladding head of the laser cladding equipment, at an angle of 30-45° to the laser beam, to avoid direct laser damage to the equipment. The sampling frequency of the thermal imager can be set to 50-100Hz, and the field of view should cover the molten pool and the surrounding heat-affected zone within a 5-10mm range. The thermal imager wirelessly transmits the captured temperature field cloud image to the computer's processing module for preprocessing.

[0117] Specifically, the processing method may include first removing irrelevant background areas by image cropping, then removing temperature noise points by median filtering algorithm, and finally converting the color cloud map into a 16-bit grayscale temperature data matrix to obtain standardized molten pool temperature distribution information.

[0118] S204, the molten pool temperature distribution information is input into the cladding quality assessment model, and the quality monitoring result is determined based on the output of the cladding quality assessment model. The cladding quality assessment model is trained based on historical sample data, which consists of data pairs composed of molten pool temperature distribution information and cladding results.

[0119] The cladding quality assessment model is a data-driven mathematical or machine learning model that maps the relationship between molten pool temperature distribution and cladding quality. Common types include Convolutional Neural Network (CNN) models, Support Vector Machine (SVM) models, or Gradient Boosting Tree (XGBoost) models. Among these, CNN models are preferred because they excel at processing image-based data, such as temperature field cloud maps. The core output of the model is a "qualified" or "unqualified" prediction result, or a multi-dimensional assessment report including "porosity risk," "crack risk," and "lack of fusion risk," such as outputting "qualified, low porosity risk (<5%)" or "unqualified, high crack risk," etc.

[0120] Historical sample data refers to paired data accumulated through numerous experiments or actual production under the same or similar cladding scenarios, such as the same substrate material and the same cladding powder type. This data consists of "melt pool temperature distribution information" and "corresponding cladding results." Each sample must include complete temperature field data, such as temperature field cloud maps and temperature gradient curves, corresponding cladding process parameters, such as laser power and scanning speed, and final cladding quality inspection results, such as ultrasonic testing for porosity, penetrant testing for cracks, and tensile testing for bonding strength. For example, in a set of historical samples, the melt pool temperature distribution is "center 1800℃, edge 1250℃, temperature difference 550℃," the cladding parameters are "laser power 2000W, scanning speed 10mm / s," and the final cladding result is "qualified, no porosity, no cracks, bonding strength 420MPa."

[0121] In this scheme, during the model training phase, at least 1000 sets of historical sample data can be collected and the samples are divided into training and test sets in an 8:2 ratio. CNN is selected as the basic model architecture. The input layer receives the preprocessed temperature data matrix, such as 640×512 pixels. Through 3 convolutional layers with a kernel size of 3×3 and a stride of 1, temperature distribution features such as the shape of the high-temperature zone and the trend of temperature gradient change are extracted. Two pooling layers are used for max pooling with a kernel size of 2×2 to compress the feature dimension. One fully connected layer maps the features to feature vectors. The output layer outputs the judgment result of "qualified" (output value > 0.5) or "unqualified" (output value ≤ 0.5) through the sigmoid activation function.

[0122] The model was trained using the training set. The error between the predicted value and the actual cladding result was calculated using the cross-entropy loss function. The model parameters were adjusted using the Adam optimizer with a learning rate of 0.001 and 100 iterations. The model accuracy was verified using the test set after each training round. Training was stopped when the accuracy stabilized above 95%, and the final model parameters were saved.

[0123] During the real-time quality assessment phase, the pre-processed molten pool temperature distribution information can be input into the trained cladding quality assessment model. The model completes the calculation and outputs the assessment result within 0.1-0.5 seconds. If the output is "qualified," normal cladding continues. If the output is "unqualified," the risk type, such as "crack risk," is further output, and other linkage controls are immediately triggered. On the one hand, a pause signal is sent to the laser cladding equipment, and on the other hand, a warning message is displayed on the operation interface. After confirmation by the operator, the equipment restarts cladding according to the adjusted parameters. At the same time, the module records the anomaly and the handling process, adding it to the historical sample database for subsequent model iteration and optimization.

[0124] The technical solution provided in this embodiment addresses the problem of existing technologies that only focus on surface features after cladding and ignore core influencing factors in the cladding process by introducing molten pool temperature distribution information and a machine learning model. On one hand, molten pool temperature is a key factor determining the microstructure and macroscopic defects of the cladding layer. Real-time monitoring of temperature distribution allows for early prediction of quality risks, avoiding rework required after problems are discovered. On the other hand, the evaluation model, trained on a large number of historical samples, enables intelligent and precise judgment of cladding quality, with an accuracy rate far exceeding that of human experience-based judgment. Furthermore, the model can be continuously optimized through sample supplementation to adapt to different cladding scenarios, further improving the adaptability and reliability of quality monitoring.

[0125] In one embodiment, optionally, the historical sample data includes positive samples and negative samples;

[0126] The positive sample data, which is the cladding result corresponding to the molten pool temperature distribution information, meets the quality requirements.

[0127] The negative sample data refers to the cladding results corresponding to the molten pool temperature distribution information, which do not meet the quality requirements.

[0128] Positive sample data refers to samples from historical sample data whose final cladding results, corresponding to the molten pool temperature distribution information, fully meet the preset quality standards. These samples must simultaneously meet multiple quality indicators. Specifically, macroscopically, they can be free of cracks (length ≤ 0.2 mm), peeling (area ≤ 0.5 mm²), and obvious depressions / protrusions (height difference ≤ 0.05 mm); microscopically, they can be free of pores (diameter ≤ 0.1 mm) and inclusions (area ≤ 0.3 mm²); and mechanically, the bond strength must be greater than 80% of the tensile strength of the substrate, and the hardness fluctuation range must be less than ±15%. For example, in a positive sample, the molten pool temperature distribution is "center 1780℃, edge 1280℃, temperature difference 500℃", and the cladding result is "no cracks, no pores, bond strength 430 MPa, hardness HV580-620".

[0129] Negative sample data refers to samples from historical sample data where the final cladding result corresponding to the molten pool temperature distribution information does not meet at least one preset quality standard. Specific non-compliance indicators and their severity must be clearly marked: for example, "excessive porosity (diameter 0.8mm, quantity 3), crack defect (length 2mm, penetrating the cladding layer), insufficient bonding strength (350MPa < 400MPa), excessive hardness fluctuation (HV500-680, fluctuation range > 15%)," etc. Each negative sample must correspond to a unique or primary reason for non-compliance. For example, in a negative sample, the molten pool temperature distribution is "center 1950℃, edge 1300℃, temperature difference 650℃," and the cladding result is: longitudinal cracks exist because the excessive molten pool temperature difference causes thermal stress to exceed the material's tensile strength.

[0130] This solution for collecting and labeling positive sample data can be implemented by selecting production batches or test groups with stable cladding processes and qualified quality. The data collection process follows these steps: First, record complete molten pool temperature distribution information using a dual-color infrared thermal imager, including temperature field cloud maps and temperature gradient data. Simultaneously, record corresponding cladding parameters such as laser power, scanning speed, and powder feed rate. After cladding, perform comprehensive quality inspections on the clad parts. Use a phased-array ultrasonic testing instrument, such as the Olympus OmniScanMX2, to detect internal porosity and inclusions. Use a dye penetrant testing agent, such as DPT-5, to detect surface cracks. Use a universal testing machine to test bond strength and a Vickers hardness tester to test hardness distribution. If all test indicators meet the quality standards, mark the temperature distribution information, cladding parameters, and test report as a positive sample. When entering the data into the database, add a "positive sample" tag and key qualification indicators.

[0131] The labeling of negative sample data follows a similar approach. For example, quality defects may be created by deliberately adjusting cladding parameters, or defective products from actual production may be collected. Specifically, the data can be collected in the following steps: Record the molten pool temperature distribution information and cladding parameters. After cladding, determine the defect type and severity using the aforementioned detection methods. For example, if ultrasonic testing reveals a pore diameter of 0.6 mm located in the middle of the cladding layer, and penetrant testing reveals a surface crack length of 3 mm distributed along the scanning direction, this set of temperature distribution information + cladding parameters + defect detection report is marked as a negative sample. When entering the data into the database, a "negative sample" label, the specific defect type, and a defect cause analysis must be added.

[0132] Since the data is used for model training, this embodiment also requires balancing and preprocessing the sample data. To avoid biased predictions due to an imbalance in the number of positive and negative samples (e.g., too many positive samples and too few negative samples), the samples need to be balanced. Specifically, when the number of positive samples far exceeds the number of negative samples, oversampling methods can be used for balancing, such as generating similar negative samples using a Generative Adversarial Network (GAN). When the number of negative samples is too large, undersampling methods can be used, such as randomly selecting a number of negative samples that are roughly equal to the number of positive samples. After balancing, the temperature distribution information of all samples is normalized, and the cladding parameters are standardized to ensure that the feature weights are balanced during model training.

[0133] This embodiment addresses the problems of ambiguous sample data definitions and inconsistent quality leading to poor model training performance in existing technologies by clearly defining the positive and negative sample division criteria and the collection and annotation methods for historical sample data. On one hand, clear positive and negative sample criteria ensure the validity and consistency of the sample data; positive samples provide acceptable temperature distribution patterns, while negative samples provide defective temperature distribution patterns, enabling the model to learn clear quality judgment boundaries. On the other hand, detailed defect annotations not only improve the model's accuracy in identifying defects but also provide a reference for subsequent cladding parameter adjustments, further strengthening the closed-loop control of monitoring-judgment-adjustment and laying a data foundation for the reliable operation of the cladding quality assessment model.

[0134] Example 3

[0135] Figure 3 This is a schematic diagram of the quality monitoring device for the metal surface treatment process provided in Embodiment 3 of this application. Figure 3 As shown, the device includes:

[0136] The surface three-dimensional data acquisition module 301 is used to acquire the surface three-dimensional data of a metal object;

[0137] The target location information determination module 302 is used to determine the target location information after cladding based on the surface three-dimensional data and cladding parameters.

[0138] The actual information acquisition module 303 is used to acquire the actual position information and the surface regularity information after laser cladding of the metal object during the laser cladding process.

[0139] The quality monitoring module 304 is used to identify whether the actual position information is consistent with the target position information in order to monitor the quality of the cladding thickness, and to identify whether the surface regularity information after cladding meets the target regularity in order to monitor the quality of the cladding regularity.

[0140] In this embodiment, for pits and cracks on the surface of metal objects, a filling field is added to the three-dimensional point space parameters and the filling value and crack width information are introduced. This enables precise customization of the target location information of the defect area, which solves the problem in the prior art that the defect area lacks targeted target planning, resulting in the defect still existing or the thickness of the cladding layer being uneven after cladding. This ensures that the pit and crack areas achieve the ideal surface morphology after cladding, and improves the reliability of the cladding quality of the defect area.

[0141] The quality monitoring device for the metal surface treatment process in this application embodiment can be a device, or a component, integrated circuit, or chip in a terminal. The device can be a mobile electronic device or a non-mobile electronic device. For example, mobile electronic devices can be mobile phones, tablets, laptops, PDAs, in-vehicle electronic devices, wearable devices, ultra-mobile personal computers (UMPCs), netbooks, or personal digital assistants (PDAs), etc., while non-mobile electronic devices can be servers, network-attached storage (NAS), personal computers (PCs), televisions (TVs), ATMs, or self-service machines, etc. This application embodiment does not impose specific limitations.

[0142] The quality monitoring device for the metal surface treatment process in this application embodiment can be a device with an operating system. This operating system can be Android, iOS, or other possible operating systems; this application embodiment does not specifically limit the specific operating system used.

[0143] The quality monitoring device for metal surface treatment provided in this application can realize the various processes implemented in the above embodiments. To avoid repetition, it will not be described again here.

[0144] Example 4

[0145] like Figure 4 As shown, this application embodiment also provides an electronic device 400, including a processor 401, a memory 402, and a program or instructions stored in the memory 402 and executable on the processor 401. When the program or instructions are executed by the processor 401, they implement the various processes of the above-described metal surface treatment quality monitoring method embodiment and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0146] It should be noted that the electronic devices in the embodiments of this application include the mobile electronic devices and non-mobile electronic devices described above.

[0147] Example 5

[0148] This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above-described quality monitoring method embodiment for metal surface treatment and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0149] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0150] Example 6

[0151] This application embodiment also provides a chip, which includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run programs or instructions to implement the various processes of the above-described quality monitoring method embodiment for metal surface treatment, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0152] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0153] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0154] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0155] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

[0156] The above description is merely a preferred embodiment and the technical principles employed in this application. This application is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions that can be made by those skilled in the art will not depart from the scope of protection of this application. Therefore, although this application has been described in detail through the above embodiments, this application is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of this application, the scope of which is determined by the scope of the claims.

Claims

1. A quality monitoring method for a metal surface treatment process, characterized in that, The method includes: Acquire three-dimensional surface data of a metallic object; Based on the surface three-dimensional data and cladding parameters, determine the target location information after cladding; During the laser cladding process on the metal object, the actual position information and surface regularity information after cladding are obtained; To monitor the quality of cladding thickness, it is necessary to identify whether the actual location information is consistent with the target location information, and to monitor the quality of cladding regularity by identifying whether the surface regularity information after cladding meets the target regularity. Specifically, determining the target location information after cladding based on the surface three-dimensional data and cladding parameters includes: Based on the surface three-dimensional data, it is determined whether there are pits on the surface of the metal object; among them, based on the point cloud curvature analysis algorithm, the curvature value of each region is calculated, and regions with abnormally large curvature values ​​are judged to be likely to have pits, and then confirmed by manual verification or high-resolution image assistance. If a pit exists, a fill field is added to the spatial parameters of the 3D point at the pit to write the fill value from the spatial position of the 3D point to the original spatial parameters; wherein, the spatial parameters of the 3D point refer to the X-axis, Y-axis, and Z-axis coordinate parameters of each point in the pit area in the surface 3D point cloud; based on the depth and area parameters of the pit, the fill value of each 3D point is calculated through the fill value calculation model and written into the fill field; Based on the spatial location of the three-dimensional points, the filling value, and the cladding parameters, the target location information after cladding is determined; or, Based on the surface three-dimensional data, it is determined whether there are cracks on the surface of the metal object; wherein, based on the point cloud distance transformation algorithm, the surface three-dimensional data is analyzed to identify linear continuous low-height areas as potentially containing cracks, and then ultrasonic flaw detection equipment is used to assist in confirming the actual length and width of the cracks. If a crack exists, a fill field is added to the spatial parameters of the three-dimensional point at the crack to write the fill value from the spatial position of the three-dimensional point to the original spatial parameters, as well as the crack width information; wherein, the spatial parameters of the three-dimensional point refer to the X-axis, Y-axis, and Z-axis coordinate parameters of each point in the pit area in the surface three-dimensional point cloud; the fill value is calculated based on the crack depth; Based on the spatial location of the three-dimensional point, the filling value, the crack width information, and the cladding parameters, the target location information after cladding is determined.

2. The quality monitoring method for metal surface treatment process according to claim 1, characterized in that, After acquiring the three-dimensional surface data of the metal object, the method further includes: If a pit or crack is detected in the metal object, the location of the defect is recorded. The laser cladding process for the metal object also includes: Based on the location of the defect, an important monitoring area is determined, and the flatness of the important monitoring area is tested.

3. The quality monitoring method for the metal surface treatment process according to claim 2, characterized in that, Based on the location of the defect, important monitoring areas are determined, including: The location and shape of the defect are input into the impact range assessment algorithm, and the important monitoring area is determined based on the output of the impact range assessment algorithm.

4. The quality monitoring method for metal surface treatment process according to claim 2, characterized in that, After determining the important monitoring area based on the defect location, the method further includes: Obtain the monitoring grid accuracy for a typical area; Based on the monitoring grid accuracy of the regular area, the grid is encrypted according to a preset method to obtain the monitoring grid accuracy of the important monitoring area.

5. The quality inspection method for the metal surface treatment process according to claim 1, characterized in that, The method further includes: During the laser cladding process on the metal object, information on the temperature distribution of the molten pool is obtained; The molten pool temperature distribution information is input into the cladding quality assessment model. Based on the output of the cladding quality assessment model, the quality monitoring result is determined. The cladding quality assessment model is trained based on historical sample data, which consists of data pairs composed of molten pool temperature distribution information and cladding results.

6. The quality monitoring method for metal surface treatment process according to claim 5, characterized in that, The historical sample data includes positive samples and negative samples; The positive sample data, which is the cladding result corresponding to the molten pool temperature distribution information, meets the quality requirements. The negative sample data refers to the cladding results corresponding to the molten pool temperature distribution information, which do not meet the quality requirements.

7. A quality monitoring device for a metal surface treatment process, characterized in that, The device includes: The surface 3D data acquisition module is used to acquire the surface 3D data of metal objects; The target location information determination module is used to determine the target location information after cladding based on the surface three-dimensional data and cladding parameters. The actual information acquisition module is used to acquire the actual position information and surface regularity information after laser cladding of the metal object during the laser cladding process. The quality monitoring module is used to identify whether the actual location information is consistent with the target location information in order to monitor the quality of the cladding thickness, and to identify whether the surface regularity information after cladding meets the target regularity in order to monitor the quality of the cladding regularity. Specifically, determining the target location information after cladding based on the surface three-dimensional data and cladding parameters includes: Based on the surface three-dimensional data, it is determined whether there are pits on the surface of the metal object; among them, based on the point cloud curvature analysis algorithm, the curvature value of each region is calculated, and regions with abnormally large curvature values ​​are judged to be likely to have pits, and then confirmed by manual verification or high-resolution image assistance. If a pit exists, a fill field is added to the spatial parameters of the 3D point at the pit to write the fill value from the spatial position of the 3D point to the original spatial parameters; wherein, the spatial parameters of the 3D point refer to the X-axis, Y-axis, and Z-axis coordinate parameters of each point in the pit area in the surface 3D point cloud; based on the depth and area parameters of the pit, the fill value of each 3D point is calculated through the fill value calculation model and written into the fill field; Based on the spatial location of the three-dimensional points, the filling value, and the cladding parameters, the target location information after cladding is determined; or, Based on the surface three-dimensional data, it is determined whether there are cracks on the surface of the metal object; wherein, based on the point cloud distance transformation algorithm, the surface three-dimensional data is analyzed to identify linear continuous low-height areas as potentially containing cracks, and then ultrasonic flaw detection equipment is used to assist in confirming the actual length and width of the cracks. If a crack exists, a fill field is added to the spatial parameters of the three-dimensional point at the crack to write the fill value from the spatial position of the three-dimensional point to the original spatial parameters, as well as the crack width information; wherein, the spatial parameters of the three-dimensional point refer to the X-axis, Y-axis, and Z-axis coordinate parameters of each point in the pit area in the surface three-dimensional point cloud; the fill value is calculated based on the crack depth; Based on the spatial location of the three-dimensional point, the filling value, the crack width information, and the cladding parameters, the target location information after cladding is determined.

8. An electronic device, characterized in that, It includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the quality monitoring method for the metal surface treatment process as described in any one of claims 1-6.

9. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the quality monitoring method for the metal surface treatment process as described in any one of claims 1-6.

Images