A method and system for hardfacing a roll body

Abstract

The application discloses a roller body hard surface welding cladding method and system, and relates to the technical field of roller body cladding. The method comprises the following steps: position calibration analysis is performed on a to-be-cladded roller body after surface pretreatment, and base state distribution data of the corresponding outer surface of the to-be-cladded roller body is obtained; the outer surface of the to-be-cladded roller body is divided into cladding control zones and matched and verified according to the base state distribution data; in the process of welding cladding based on initial process parameters, molten pool image data and / or molten pool contour data of a current cladding control zone are obtained, and molten pool geometric feature parameters are extracted therefrom; the molten pool geometric feature parameters are compared with a target molten pool geometric range corresponding to the current cladding control zone, and a deposition quality change trend is determined; a process parameter feedback adjustment amount is generated based on the deposition quality change trend, so as to perform feedback adjustment on the initial process parameters of the current cladding control zone, and the problem of uneven hardness distribution of a roller surface in the prior art is solved.

Description

Technical Field

[0001] This invention relates to the field of roller cladding technology, and in particular to a method and system for hard surface welding cladding of rollers. Background Technology

[0002] In equipment such as mining roller presses, building material grinding rollers, steel rolling mill rolls, and conveyor rollers, the outer cylindrical surface of the roller is subjected to material extrusion, friction, and impact over a long period of time. This can easily lead to wear grooves, localized spalling, fatigue cracks, and thinning of the wear-resistant layer. To restore the roller's external dimensions and improve its wear resistance, current technologies typically employ hard-face welding cladding to form a wear-resistant alloy layer on the outer cylindrical surface of the roller.

[0003] Existing techniques typically involve the following process: First, determine the wear depth, crack location, and repair area on the roller surface. Then, remove the fatigue layer, spalling layer, oxide layer, and contaminants from the roller surface through turning, grinding, polishing, or sandblasting to create a suitable metal surface for welding cladding on the outer cylindrical surface of the roller. Next, preheat the roller according to its material, dimensions, and cladding material. During cladding, the roller is usually mounted on a rotating device or roller frame, rotating around its own axis while the welding torch or cladding head moves axially along the roller, forming a continuous weld bead on the outer cylindrical surface of the roller. Depending on the repair thickness and wear-resistant layer structure requirements, an underlayer or transition layer can be formed first, followed by the wear-resistant working layer, ensuring a certain overlap between adjacent weld beads. After each layer or part of the weld bead is completed, slag, spatter, and surface defects are typically cleaned, and the interpass temperature is monitored.

[0004] After the cladding is completed, the roller body is usually subjected to heat preservation, slow cooling or stress relief treatment to reduce the residual stress between the cladding layer and the roller body substrate. Then, the roller body is processed to the required size by turning, grinding and other methods, and hardness testing, dimensional inspection, flaw detection and dynamic balance testing are carried out to make the roller body meet the requirements for reuse.

[0005] In actual welding and cladding processes, the roller surface is usually not in a uniform wear state, but rather the old hardened layer has inconsistent thickness. Existing technologies typically involve inspecting and machining the roller before welding, then performing cladding according to predetermined welding current, welding voltage, wire feed speed, roller rotation speed, and welding torch movement speed, and restoring the roller dimensions by machining or grinding after cladding. The main focus is on pre-welding treatment, selection of cladding materials, cladding layers, and post-welding processing and inspection. During the cladding process, each weld seam is usually formed continuously according to set parameters and a set trajectory.

[0006] However, during the welding cladding repair process after the worn surface of the roller is removed by machining, although a continuous geometric contour is formed on the outer cylindrical surface of the roller, there may still be differences in the residual thickness of the old hardened layer, the degree of exposure of the base metal, and the local heat-affected microstructure along the axial and circumferential directions. Because the old hardened layer, base metal, and transition microstructure have different endothermic, conductive, and melting responses to welding heat input, different substrate areas will form different melting ranges and solidification conditions under the same heat input. This can easily lead to differences in local melt depth, molten pool spread, dilution ratio, and cooling shrinkage, resulting in inconsistent fusion line morphology between the cladding layer and the substrate, and fluctuations in the hardened layer microstructure and hardness distribution in adjacent areas. This, in turn, increases the risk of localized spalling due to uneven hardness distribution on the roller surface. Summary of the Invention

[0007] To address the aforementioned technical problems in the existing technology, embodiments of the present invention provide a method and system for hard surface welding and cladding of roller bodies. The technical solution is as follows: On the one hand, a method for hard-surface welding and cladding of a roller body is provided, the method comprising: S1. Perform position calibration analysis on the roller body to be clad after surface pretreatment, and obtain the substrate state distribution data of the corresponding outer circular surface of the roller body to be clad. The substrate state distribution data is used to characterize the differences in substrate morphology, surface transition state and hard surface welding cladding adaptation state at different positions on the outer circular surface.

[0008] S2. Based on the substrate state distribution data, the outer circular surface of the roller to be clad is divided into cladding control zones, and the corresponding initial process parameters are determined for each cladding control zone. At the same time, the matching of the initial process parameters is verified based on the substrate state distribution data of the corresponding cladding control zone. The initial process parameters include one or more of the following: hard surface welding parameters, cladding material supply parameters, relative motion parameters, and weld overlap parameters.

[0009] S3, during the welding cladding process based on the initial process parameters, acquire the molten pool image data and / or molten pool contour data of the current cladding control zone, and extract the molten pool geometric feature parameters from them. The molten pool geometric feature parameters include at least one or more of the following: molten pool width, molten pool length, molten pool area, molten pool height, and weld bead forming width.

[0010] S4. Compare the geometric characteristic parameters of the molten pool with the geometric range of the target molten pool corresponding to the current cladding control area to generate molten pool geometric deviation parameters, and determine the deposition quality change trend of the current cladding control area based on the molten pool geometric deviation parameters.

[0011] S5 generates process parameter feedback adjustment amounts based on the deposition quality change trend to adjust the initial process parameters of the current cladding control area, obtaining updated cladding process parameters, which serve as the basis for updating the initial process parameters of the next cladding control area.

[0012] On the other hand, a roller hardface welding cladding system is provided, the system comprising: The position calibration analysis module is used to perform position calibration analysis on the roller body to be melted and coated after surface pretreatment, and to obtain the substrate state distribution data of the corresponding outer circular surface of the roller body to be melted and coated.

[0013] The initial cladding preprocessing module is used to divide the outer circular surface of the roller to be clad into cladding control zones based on the substrate state distribution data, determine the corresponding initial process parameters for each cladding control zone, and perform matching verification of the initial process parameters based on the substrate state distribution data of the corresponding cladding control zone.

[0014] The molten pool geometric feature analysis module is used to acquire molten pool image data and / or molten pool contour data of the current cladding control zone during the welding cladding process based on initial process parameters, and extract molten pool geometric feature parameters from them.

[0015] The sedimentation quality change analysis module is used to compare the geometric characteristic parameters of the melt pool with the target melt pool geometric range corresponding to the current cladding control area, generate melt pool geometric deviation parameters, and determine the sedimentation quality change trend of the current cladding control area based on the melt pool geometric deviation parameters.

[0016] The cladding feedback adjustment module is used to generate process parameter feedback adjustment amounts based on the deposition quality change trend, so as to adjust the initial process parameters of the current cladding control area, obtain the updated cladding process parameters, and use them as the basis for updating the initial process parameters of the next cladding control area.

[0017] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: 1. By continuously linking the position calibration analysis of the outer cylindrical surface of the roller, the modeling of the substrate state distribution, the division of the cladding control zone, the matching and verification of the initial process parameters, the extraction of the geometric features of the molten pool, the judgment of the deposition quality change trend, and the feedback adjustment of the process parameters, the welding cladding process can form a front-end identification, mid-end adaptation, and back-end correction process around the differences in the substrate state at different positions of the roller. The differences in the substrate, such as the residual state of the old hard surface layer, the exposed state of the substrate metal, and the surface transition state, are transformed into distribution data that can participate in process control. In the cladding process, the geometric features of the molten pool are further used to reflect the actual deposition state, and the deviation of the molten pool is converted into the deposition quality change trend such as spreading, accumulation, and coverage. Finally, feedback adjustment amounts are formed for the type, direction, and magnitude of process parameters. It enables a dynamic correlation between parameter settings for different substrate regions and the actual cladding response, which helps reduce fluctuations in the molten pool morphology caused by different thermal responses of the old hard surface layer, base metal, and transition structure. This makes it easier to maintain coordination between the cladding layer's forming state, fusion state, and weld bead coverage state, thereby improving the problem of mismatch in local fusion depth, spreading state, accumulation thickness, and overlap coverage during the hard surface welding cladding repair of rollers.

[0018] 2. By establishing axial position coordinates, circumferential angle coordinates, and outer surface development coordinates, and performing unitized position calibration on the outer surface of the roller body on the development coordinates, the detection information such as surface contours, local grinding, old hard surface layer residues, exposed base metal, and surface transitions, which were originally scattered along the circumference and axis of the roller body, can be uniformly associated with specific locations. The substrate morphology and surface transition states are normalized and identified, and then the position state combination data is matched with the cladding adaptation relationship to form a state label that reflects the differences in cladding adaptation at different locations. Based on this, locations with similar adaptation states are classified into the same cladding control zone. Through this combination of position data and adaptation state data, the division of the cladding control zone is made more consistent with the actual substrate differences after roller body pretreatment. This allows subsequent process parameters to be configured not only according to axial stroke or weld sequence, but also regionalized based on differences in substrate heat absorption, heat conduction, and melting response. This treatment method can better identify the local differences caused by the alternating distribution of old hard surface layer residue and exposed substrate, which is conducive to providing a more matched initial process basis for different cladding areas and reducing the sudden changes in forming state caused by large differences in substrate state between adjacent areas.

[0019] 3. By defining the molten pool data acquisition window within the current cladding control area and acquiring molten pool image data or molten pool contour data, geometric features such as molten pool width, length, area, height, and weld bead forming width are extracted from the molten pool boundary. This allows the actual deposition state during the cladding process to be transformed into identifiable geometric parameters in real-time or near real-time. Initial process parameters are first decoupled and normalized into process parameter feature vectors. Then, the target molten pool geometric range is determined by combining the mapping relationship between process parameters and molten pool geometry. This allows molten pool deviation judgment to determine a reasonable geometric reference based on the parameters of the current control area. Molten pool geometric deviation parameters are generated by the out-of-bounds direction, out-of-bounds amount, and deviation level. These parameters are then aggregated into spreading change direction, accumulation change direction, and coverage change direction. These are then correlated with hardface welding parameters, cladding material supply parameters, relative motion parameters, and weld bead lap parameters, respectively, transforming the deviation identification results into process-oriented feedback adjustment quantities. This enables a clear correspondence between molten pool geometry changes, deposition quality trends, and process parameter categories, which helps improve the targeting of parameter adjustments. It allows the next cladding control zone to inherit the feedback results of the current zone and adapt and update accordingly, thereby reducing problems such as insufficient transfer of forming state between zones, lag in parameter adjustment, and fluctuations in weld quality during continuous cladding. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 A flowchart of a roller hard surface welding cladding method provided in an embodiment of the present invention; Figure 2 A flowchart for generating a corresponding location calibration and substrate state distribution is provided in the embodiments of the present invention; Figure 3 A flowchart corresponding to the control area division and initial parameter verification provided in an embodiment of the present invention; Figure 4 The flowcharts corresponding to the analysis and feedback adjustment of molten pool geometric deviations provided in the embodiments of the present invention are as follows; Figure 5 This is a schematic diagram of a roller hard surface welding cladding system provided in an embodiment of the present invention. Detailed Implementation

[0022] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0023] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0024] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0025] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0026] This invention provides a method for hard surface welding and cladding of a roller body, such as... Figure 1 The flowchart shown represents a method for hard surface welding and cladding of a roller body. The process of this method may include the following steps: S1. Perform position calibration analysis on the roller body to be clad after surface pretreatment, and obtain the substrate state distribution data of the corresponding outer circular surface of the roller body to be clad. The substrate state distribution data is used to characterize the differences in substrate morphology, surface transition state and hard surface welding cladding adaptation state at different positions on the outer circular surface.

[0027] S2. Based on the substrate state distribution data, the outer circular surface of the roller to be clad is divided into cladding control zones, and the corresponding initial process parameters are determined for each cladding control zone. At the same time, the matching of the initial process parameters is verified based on the substrate state distribution data of the corresponding cladding control zone. The initial process parameters include one or more of the following: hard surface welding parameters, cladding material supply parameters, relative motion parameters, and weld overlap parameters.

[0028] S3, during the welding cladding process based on the initial process parameters, acquire the molten pool image data and / or molten pool contour data of the current cladding control zone, and extract the molten pool geometric feature parameters from them. The molten pool geometric feature parameters include at least one or more of the following: molten pool width, molten pool length, molten pool area, molten pool height, and weld bead forming width.

[0029] S4. Compare the geometric characteristic parameters of the molten pool with the geometric range of the target molten pool corresponding to the current cladding control area to generate molten pool geometric deviation parameters, and determine the deposition quality change trend of the current cladding control area based on the molten pool geometric deviation parameters.

[0030] S5 generates process parameter feedback adjustment amounts based on the deposition quality change trend to adjust the initial process parameters of the current cladding control area, obtaining updated cladding process parameters, which serve as the basis for updating the initial process parameters of the next cladding control area.

[0031] In this embodiment, as Figure 2 The diagram shows a flowchart illustrating the corresponding generation of position calibration and substrate state distribution provided by an embodiment of the present invention. This method first calibrates the position of the roller body to be clad after surface pretreatment. Surface pretreatment includes one or more of turning, grinding, polishing, sandblasting, degreasing, and surface cleaning, used to remove loose layers, contaminants, oxide layers, and locally failed layers that affect welding cladding, thus forming a substrate surface on the outer cylindrical surface of the roller body that can be inspected and cladding controlled.

[0032] First, obtain the axial reference position, circumferential reference position, and outer diameter circumference of the roller to be clad. The axial reference position is the starting point for calculating the axial position of the roller. It can be selected from one end face of the roller, the end positioning step, the axial positioning surface of the fixture, or the starting position specified in the process document. During acquisition, move the detection head or welding torch to this starting position, and the axial displacement sensor reads the current position, recording it as the axial zero point. Subsequently, the distance in the axial direction from any position on the outer circumferential surface of the roller is calculated based on this axial zero point, thereby establishing the axial position coordinates.

[0033] The circumferential reference position serves as the starting point for calculating the circumferential angle of the roller. It can be selected from positioning marks on the roller surface, end keyways, circumferential positioning lines of the fixture, or manually calibrated starting lines. During acquisition, the circumferential reference position is aligned with the reference direction of the detection equipment. The current angle is read by the angle sensor and recorded as the zero circumferential angle. As the roller rotates, the change in circumferential angle is continuously read to establish the circumferential angle coordinates.

[0034] The outer diameter circumference is the length of the outer circular surface of the roller to be clad along the circumferential direction. It is obtained by using a laser diameter gauge, 3D contour scanner, or outer diameter measuring device to collect the outer diameter dimensions of the roller, and then generating the outer diameter circumference according to the circumferential length calculation rules. For rollers with different outer diameters after local grinding, this method can collect outer diameter data at different axial positions, then perform regional statistics on the collected results, and select the outer diameter circumference that matches the current cladding area for subsequent circumferential unfolding calculations.

[0035] After obtaining the circumferential angle coordinates and outer diameter circumference, the circumferential angle coordinates are converted into the circumferential unfolded position. The circumferential unfolded position refers to the linear position formed after unfolding along the circumferential direction of the outer cylindrical surface of the roller, starting from the circumferential reference position. Specifically, the unfolded length corresponding to the current circumferential angle is obtained based on the relative position of the current circumferential angle within the complete circumference and the outer diameter circumference. Subsequently, the axial position coordinates and the circumferential unfolded position are combined to generate the unfolded coordinates of the outer cylindrical surface of the roller to be clad. The unfolded coordinates of the outer cylindrical surface are used to convert the spatial position on the cylindrical surface of the roller into a planar position, enabling subsequent detection data, area division, and parameter matching to be performed under the same position reference.

[0036] On the outer circular surface, the coordinates are divided into units based on weld width and sampling interval, resulting in multiple position calibration units. Weld width is the expected coverage width formed by a single hard-surface weld cladding on the roller surface, which can be determined based on welding torch type, welding method, cladding material, wire feeding method, historical test welding data, and records of similar roller processes. Sampling interval is the distance between adjacent detection sampling positions. The sampling interval can be determined based on the weld width of the detection device, the density of changes in the roller surface condition, and the accuracy of subsequent cladding control. Specifically, the calibration parameters of the detection device are read, and combined with the distribution of residual old hard-surface layer, exposed substrate, and grinding marks on the pre-treated roller surface, to determine the sampling interval that can distinguish changes in the condition of adjacent surfaces. The sampling interval should not be too large to avoid mixing different substrate conditions into the same position unit; nor should it be too small to avoid increasing the amount of invalid data and affecting subsequent processing efficiency.

[0037] A position calibration unit is the smallest positional area in the unfolded coordinate system of the outer cylindrical surface used to carry the detection data. Each position calibration unit has a corresponding axial range and circumferential unfolded range. A position number is established for each position calibration unit, and the subsequently acquired detection data is associated with this position number, so that the substrate state of the outer cylindrical surface of the roller can be recorded, queried, and statistically analyzed according to position. After completing the unitization, the outer cylindrical surface detection data corresponding to each position calibration unit is obtained. The outer cylindrical surface detection data includes one or more of the following: surface contour data, local grinding state data, old hardened layer residual state data, exposed base metal state data, and surface transition state data.

[0038] Surface profile data is used to represent the geometry within the location calibration unit, including local depressions, protrusions, residual grooves, abrupt changes in profile, and variations in surface continuity. This data is acquired along the axial and circumferential directions by a laser rangefinder, a 3D profile scanning device, or a structured light detection device, and mapped to the unfolded coordinates of the outer circular surface.

[0039] Local grinding status data is used to indicate whether the position calibration unit has undergone grinding, polishing, or local finishing, and the degree of surface change after grinding. This data can be determined based on the working trajectory of the grinding equipment, changes in surface contour, and grinding texture in visual images.

[0040] Data on the residual state of the old hardened layer is used to indicate whether the old hardened layer remains within the location calibration unit and the distribution of the residual area. This data can be obtained by acquiring surface images using a vision acquisition device, and then identified based on the differences between the old hardened layer and the base metal in terms of color, reflectivity, texture, and boundary morphology. For areas where the image recognition results are unstable, hardness test results, historical cladding records, or manual verification results are introduced for confirmation.

[0041] The data on exposed base metal status is used to indicate whether the base metal is exposed within the location calibration unit and the extent of the exposed area. This data can be determined jointly based on visual images, material surface identification results, and identification results of residual old hardfacing layers. When no old hardfacing layer features are identified in a certain area, and the surface image or detection results match the characteristics of the base metal, that area is recorded as an exposed base metal area.

[0042] Surface transition state data is used to represent the boundary relationship between the residual area of ​​the old hardened layer, the exposed base metal area, and the locally ground area. This data is generated based on the changes in material state, profile state, and grounding state of adjacent calibration cells. When there is a change in material state, abrupt change in profile, or change in grounding marks between adjacent cells, the corresponding location is recorded as a surface transition region, and the transition position and direction are recorded.

[0043] Finally, the calibration units at each position are correlated with the corresponding outer cylindrical surface detection data to generate substrate state distribution data. This substrate state distribution data is a dataset indexed by the unfolded coordinates of the outer cylindrical surface, used to represent the substrate morphology, surface transition state, and cladding adaptation differences at different locations on the outer cylindrical surface of the roller. This data allows for the division of cladding control zones based on the actual substrate state at different locations in subsequent steps, and provides a location-based data foundation for initial process parameter determination and parameter matching verification.

[0044] After step S1, the substrate state distribution data of the outer cylindrical surface of the roller to be clad is obtained. Next, step S2 is executed, which further includes three core steps: cladding control zone division, initial process parameter determination, and matching verification. Figure 3 The diagram shown is a flowchart illustrating the control area division and initial parameter verification provided in an embodiment of the present invention. The specific implementation process is as follows: First, the process of dividing the cladding control zone is explained. The generated substrate state distribution data is read, where each location calibration unit is associated with corresponding outer circular surface detection data. Substrate detection features are extracted from the outer circular surface detection data of each location calibration unit. These features specifically include local unevenness depth values ​​in the surface contour data, boundary gradient values ​​between the ground and unground areas in the local grinding state data, residual layer thickness values ​​in the old hardened layer residual state data, exposed area ratio values ​​in the exposed base metal state data, and transition region width values ​​in the surface transition state data. These raw detection feature values ​​are normalized by dividing each feature value by its corresponding maximum possible range to obtain a dimensionless value between 0 and 1. After normalization, two types of state identifiers are generated. The first is the substrate morphology state identifier, which characterizes the differences in substrate morphology between each location calibration unit. Specifically, it consists of a numerical value representing surface smoothness and a code representing local unevenness features. The second type is the surface transition state identifier, which is used to characterize the difference in the transition area between the old hard surface layer and the base metal. Specifically, it consists of a numerical value representing the width of the transition area and a code representing the gradient of the microstructure change in the transition area.

[0045] After obtaining the substrate morphology status identifier and surface transition status identifier for each location calibration unit, these two identifiers are combined to form the location status combination data for that unit. The combination method involves sequentially concatenating the numerical codes of the substrate morphology status identifier and the surface transition status identifier into a string or a tuple. An internal pre-stored database of the correspondence between status combinations and cladding adaptation is used. This database was established based on extensive preliminary process test data and records the optimal welding cladding adaptation results corresponding to different location status combinations. The currently calculated location status combination data is matched against this pre-stored database to find the corresponding location adaptation category. The location adaptation category can be, for example, an easy-to-clad category, a low-heat-input category, a high-spreadability category, or a transition layer category.

[0046] Based on the matched positional adaptation categories, a corresponding hard-surface welding cladding adaptation status identifier is generated for each positional calibration unit. This adaptation status identifier is a code containing numbers and letters, such as a category code followed by a level number indicating the degree of adaptation. Preset adaptation similarity conditions are also included, specifically meaning that the hard-surface welding cladding adaptation status identifiers of two positional calibration units belong to the same main branch in a preset category hierarchy tree, and the difference in their adaptation level numbers does not exceed a preset threshold, such as a difference of no more than 2. All positional calibration units are traversed, and units that meet the above adaptation similarity conditions are grouped into the same cladding control area. After classifying all units, a summary analysis is performed, determining that the entire outer cylindrical surface of the roller to be clad is divided into several continuous or discontinuous cladding control areas. Each cladding control area covers several positional calibration units that are similar in substrate morphology and surface transition state.

[0047] For each defined cladding control zone, the following steps are performed to determine its corresponding initial process parameters: Specifically, a set of process parameters to be verified is generated. First, within the same cladding control zone, the frequency of occurrence of different hardface welding cladding adaptation status identifiers carried by all position calibration units is counted, and the percentage of each identifier relative to the total number of units in that control zone is calculated, generating regional adaptation status distribution data for that cladding control zone. This distribution data is presented in the form of a list or distribution map.

[0048] Based on the adaptation state distribution data of the region, the adaptation state identifier with the highest proportion is identified as the dominant adaptation state of the control area. Simultaneously, the variance or standard deviation of the percentage of each adaptation state identifier is calculated, and the result is used as the state dispersion of the control area. The dominant adaptation state reflects the most important base characteristics of the region, while the state dispersion reflects the consistency level of the base characteristics within the region. The dominant adaptation state and state dispersion are input into a pre-defined decision model, which outputs the process parameter requirement data for the cladding control area. The process parameter requirement data is a demand vector containing multiple dimensions, describing the heat input characteristics, material supply characteristics, action time characteristics, and weld bead coverage characteristics required to achieve high-quality cladding.

[0049] The pre-set decision model uses historical roller hard surface welding cladding process test data. Each set of test data includes the dominant adaptation state and state dispersion in the substrate state data measured before the test, as well as the process parameters used in the test and the final cladding quality evaluation results such as cladding layer hardness uniformity, dilution rate, and weld bead forming coefficient to form a training sample set. The sample set is trained using a rule-based classification regression tree algorithm. The output process parameter requirement data can make the corresponding cladding quality evaluation index reach the optimal. After training, the parameters and structure of the model are solidified, which is the pre-set decision model.

[0050] To transform the qualitative demand vector into quantifiable levels, the process parameter demand data is further quantified. The quantification process is as follows: Based on the expected values ​​of the thermal input characteristics in the demand vector, it is divided into three levels: low, medium, and high; based on the expected values ​​of the cladding material supply characteristics, it is divided into three levels: low, medium, and high; based on the expected values ​​of the relative interaction time characteristics, it is divided into three levels: fast, medium, and slow; and based on the expected values ​​of the weld bead coverage characteristics, it is divided into three levels: sparse, standard, and dense. Ultimately, the combination of these four levels constitutes the process parameter demand levels for this cladding control area.

[0051] An internally pre-stored process parameter mapping table, compiled from extensive process experiments and expert knowledge, records the correspondence between a set of process parameter requirement levels and a set of candidate process parameter groups. Based on the four requirement levels just generated, a search is performed in this pre-stored table to select the row that perfectly matches or is closest to these three levels, thus obtaining one or more candidate process parameter groups. Each candidate process parameter group specifically includes hardface welding parameters such as the numerical range of welding current and welding voltage, cladding material supply parameters such as wire feed speed, relative motion parameters such as roller speed and welding torch movement speed, and weld overlap parameters such as the percentage value of the overlap rate.

[0052] After obtaining the candidate process parameter set, its applicability is screened. The purpose of the screening is to ensure that the parameters not only meet the substrate condition requirements of this control area, but also guarantee the continuity of welding cladding between the preceding and following control areas, as well as the smoothness of the transition of the cladding control area. The specific screening rules are as follows: First, check whether each parameter in the candidate parameter set is within the equipment's capability range; second, simulate the weld bead shape generated under this parameter set to determine whether it can form a smooth transition with the weld bead generated by the pre-selected parameters in the adjacent control area, i.e., whether the height difference and width difference of the adjacent weld bead are less than a preset allowable value; finally, evaluate whether the heat-affected zone of the molten pool will excessively extend into the adjacent control area when cladding is performed under this parameter set. The candidate process parameter set that meets all the above screening conditions is used as the candidate process parameter set for this cladding control area. When only one candidate process parameter set is obtained after screening, this candidate process parameter set is determined as the process parameter set to be verified for this cladding control area; when multiple candidate process parameter sets are obtained after screening, the parameter change range of each candidate process parameter set relative to the determined process parameter set in the adjacent cladding control area is calculated.

[0053] The parameter variation amplitude is used to represent the degree of change between the current candidate process parameter group and the process parameter groups of adjacent cladding control zones. Specifically, the hardface welding parameters, cladding material supply parameters, relative motion parameters, and weld bead lap parameters in the candidate process parameter group are normalized to obtain the corresponding normalized parameter values. The difference between each normalized parameter value and the normalized parameter value of the corresponding process parameter in the adjacent cladding control zone is calculated to obtain the parameter variation amplitude of the candidate process parameter group. The lower the parameter variation amplitude, the smoother the changes in welding heat input, material supply, relative motion, and weld bead lap between the candidate process parameter group and the adjacent cladding control zone. The candidate process parameter group with the lowest parameter variation amplitude is selected as the process parameter group to be verified for the cladding control zone.

[0054] The generated parameter set to be verified is subjected to matching verification: the substrate state distribution data corresponding to the cladding control area is extracted again, and regional statistical processing is performed on the substrate detection features of all position calibration units. The regional statistical processing includes calculating the mean, median, standard deviation, and range between the maximum and minimum values. The processing results form the adaptation statistics of the cladding control area, specifically including substrate morphology statistics reflecting the overall level of substrate morphology, surface transition statistics reflecting the overall characteristics of the surface transition zone, and hardface welding cladding adaptation statistics reflecting the overall distribution of hardface welding cladding adaptation status within the area.

[0055] The constraint level of the cladding control zone is determined based on the aforementioned adaptive statistical data. The constraint level is determined by the dispersion and extreme value deviation of the adaptive statistical data: if both the standard deviation and range are less than the corresponding first preset threshold, the constraint level is relaxed; if they are between the first and second preset thresholds, it is moderate; if both are greater than the second preset threshold, it is strict. Different constraint levels correspond to different process parameter adaptation conditions. Process parameter adaptation conditions are a set of mathematical inequalities used to limit the range and combination relationships that each parameter in the process parameter set to be verified must satisfy. For example, for strict constraints, the adaptation conditions require the welding current to be within a narrow range of A1 to A2, and the ratio of wire feed speed to welding current to be between B1 and B2. Specific preset thresholds are set by the personnel in charge.

[0056] The specific values ​​of each parameter in the process parameter group to be verified are substituted into the above process parameter adaptation conditions one by one for matching verification. If all parameters meet the preset matching requirements, that is, all fall within the corresponding adaptation range, the matching verification passes, and the process parameter group to be verified is officially determined as the initial process parameters of the cladding control area. If any parameter does not meet the adaptation conditions, the verification fails. At this time, according to the specific item that does not meet the requirements and the degree of deviation, the next set of parameters will be reselected from the previously generated candidate process parameter group, or the current parameter group will be fine-tuned according to the preset step size to form a reselected process parameter group to be verified, and the previous step will be returned to perform matching verification again. This process is repeated until a parameter group that passes the matching verification is obtained or the maximum number of reselections is met, and it is determined as the initial process parameters of the cladding control area. The maximum number of reselections is preset by the preset personnel according to the current performance. If a parameter group that passes the matching verification is not obtained when the maximum number of reselections is reached, a matching verification failure alarm message is sent.

[0057] After completing the matching verification of the initial process parameters of the current cladding control zone in step S2, the welding cladding process begins, and step S3 is executed, as follows. Figure 4 The diagram shown is a flowchart corresponding to the molten pool geometric deviation analysis and feedback adjustment provided in an embodiment of the present invention. The specific implementation process of step S3 is as follows: The outer circumferential surface coordinates of the roller to be clad and the position range of each cladding control zone on the outer circumferential surface coordinates are obtained. At the start of welding cladding, the axial position of the welding torch relative to the roller and the rotational angle of the roller are read in real time using an encoder mounted on the welding torch or the roller's rotation axis. This real-time axial position and rotational angle are converted into real-time cladding position data on the outer circumferential surface coordinates. The difference between the axial encoder reading and the axial zero point is calculated to obtain the axial position coordinates; the rotational angle encoder reading is converted into radians or angles, and then unfolded into a circumferential unfolded position based on the roller's outer diameter circumference. This real-time cladding position data is compared with the position range of each cladding control zone to determine which cladding control zone the real-time position falls within, thus identifying the current cladding control zone to which the current cladding position belongs.

[0058] After determining the current cladding control zone, the corresponding molten pool data acquisition window is determined based on the location range of this control zone. The molten pool data acquisition window is a temporally and spatially defined set used to specify which locations and when to acquire molten pool data within the current cladding control zone. Specifically, it is determined as follows: axial start coordinates, axial end coordinates, circumferential start angle, and circumferential end angle are extracted from the location range of the current cladding control zone. The spatial range of the molten pool data acquisition window is set to a region slightly smaller than the current cladding control zone. For example, axially, it is recessed inward from the control zone start coordinates by a safe distance equal to a preset weld bead length, such as 5 mm; circumferentially, it is recessed inward from the start angle by a safe angle equal to the angle corresponding to a preset molten pool diameter. This setting aims to avoid acquiring mixed data influenced by adjacent control zones at the control zone boundaries. The acquisition timing of the molten pool data acquisition window is set as follows: when the real-time cladding position coordinates first enter the above spatial range, an acquisition start signal is triggered; when the real-time cladding position coordinates leave the above spatial range, an acquisition end signal is triggered. If the real-time cladding position pauses or regresses within the window, the acquisition timing will remain paused until the position continues to move forward.

[0059] During the active period of the molten pool data acquisition window, image data and contour data of the molten pool in the current cladding control area are acquired using at least one high-speed industrial camera. The high-speed industrial camera is mounted behind and fixed to the welding torch, and the molten pool contour data is also acquired using a laser contour scanner, thereby generating two-dimensional contour data of the molten pool area in real time.

[0060] The acquired raw molten pool image data and molten pool contour data are transmitted in real time to the image processing unit for preprocessing to obtain the basic molten pool data corresponding to the current cladding control area. The preprocessing process performs the following operations in sequence: First, the image data is converted to grayscale, transforming the color image into a single-channel grayscale image; second, a medium-level filtering algorithm is applied to remove salt-and-pepper noise and arc spatter interference from the image; third, histogram equalization is used to enhance the contrast of the molten pool area; fourth, for data from the laser contour scanner, outliers exceeding a preset height threshold, such as ±10 mm, are removed. After the above preprocessing, clear and noise-reduced basic molten pool data is obtained, which includes the enhanced grayscale image sequence and the filtered contour point cloud.

[0061] Molten pool region identification is performed on the basic data of the molten pool to determine the molten pool boundary data corresponding to the current cladding control area. For grayscale image sequences, a threshold-based segmentation method is used. First, the grayscale gradient of each pixel in the image is calculated, and an adaptive threshold is set, which is dynamically determined by the Otsu algorithm. Continuous pixel regions with grayscale gradients exceeding the threshold are marked as candidate molten pool regions. Then, based on the geometric prior knowledge of the molten pool region, i.e., the molten pool is usually comet-shaped or elliptical and located directly below the end of the welding wire, the region that best fits this shape is selected as the molten pool region. For contour data, the obvious bending or breakage locations of structured light stripes in the molten pool region are identified, and the point cloud lines at these locations are used as the molten pool boundary. The boundary point sets identified by the above two methods are fused, i.e., based on the main boundary identified by the image, the boundary identified by the contour data is corrected and supplemented, finally generating a closed and continuous molten pool boundary curve, which is the molten pool boundary data. This boundary data is stored in the form of an ordered point set, where each point contains the planar coordinates under the unfolded coordinates of the outer circular surface and the corresponding height value.

[0062] Finally, based on the molten pool boundary data, the geometric characteristic parameters of the molten pool in the current cladding control zone are calculated and determined: The molten pool width is defined as the maximum straight-line distance between two points on the molten pool boundary perpendicular to the welding direction (i.e., the weld width direction). The extreme point pairs in the welding direction normal are searched in the molten pool boundary point set, and their Euclidean distance is calculated to obtain the molten pool width value. The molten pool length is defined as the straight-line distance between the farthest point at the front end and the farthest point at the rear end of the molten pool boundary along the welding direction (i.e., the weld length direction). The molten pool length value is obtained by projecting the molten pool boundary point set along the welding direction vector and calculating the absolute value of the maximum difference in projection. The molten pool area is obtained by integrating the pixel region enclosed by the molten pool boundary. Specifically, the number of pixels within the closed boundary is multiplied by the actual area calibration value represented by each pixel. The molten pool height is defined as the difference between the maximum height of all points within the molten pool boundary and the average height of the surrounding base material region, where the surrounding base material region is selected as a ring-shaped area at a preset distance from the outer edge of the molten pool boundary, for example, 2 mm. The weld bead width is defined as the actual width of the weld bead after solidification. This value is obtained by analyzing the contour data collected by a laser profile scanner in the solidified region behind the molten pool. Specifically, it involves measuring the horizontal distance between the two inflection points on the top of the weld bead on a cross-section perpendicular to the welding direction in the solidified region. One or more values ​​among the calculated molten pool width, molten pool length, molten pool area, molten pool height, and weld bead width are then combined to form the geometric characteristic parameters of the molten pool in the current cladding control zone.

[0063] After obtaining the geometric characteristic parameters of the molten pool in the current cladding control zone in step S3, step S4 is executed. Step S4 first processes the initial process parameters determined in step S2 to generate the target molten pool geometric range, then compares the actual molten pool geometric characteristic parameters with it to generate molten pool geometric deviation parameters, and determines the deposition quality change trend accordingly. The specific implementation process is as follows.

[0064] The initial process parameters of the current cladding control zone are read. These parameters include one or more of the following: hardface welding parameters, cladding material supply parameters, relative motion parameters, and weld bead lap parameters. These parameters are then decoupled and normalized to generate a process parameter feature vector for the current cladding control zone. Decoupling involves separating interrelated parameters into independent components. For example, welding current and welding voltage jointly determine heat input, but after decoupling, they are treated as independent variables. Normalization converts the actual physical value of each parameter into a dimensionless relative value. The conversion formula is: subtract the minimum value of the parameter in the entire cladding control zone from its actual value as the numerator, and add 1 to the difference between the maximum and minimum values ​​as the denominator. After scaling, the normalized value is obtained. After decoupling and normalization, all parameters are mapped to the interval between 0 and 1 and arranged in a fixed order into a one-dimensional array, which is the process parameter feature vector.

[0065] A pre-stored database of process parameters and molten pool geometry mappings is established as follows: During the development phase, process engineers conduct numerous standard welding cladding tests for different substrate conditions and combinations of process parameters. In each test, process parameters and corresponding molten pool geometric feature parameters, including molten pool width, length, area, height, and weld bead width, are collected simultaneously. Using the process parameter feature vectors from each set of test data as input and the corresponding molten pool geometric feature parameters as output, a multivariate linear regression or neural network algorithm is used for training to obtain a mathematical model capable of predicting molten pool geometric feature parameters based on the process parameter feature vectors. Once the parameters and structure of this model are solidified, they constitute the pre-defined process parameter and molten pool geometry mapping relationship.

[0066] The process parameter feature vector of the current cladding control zone is input into the above mapping model. The model outputs a set of predicted values, which are the geometric reference values ​​of the molten pool under ideal conditions. The geometric reference values ​​of the molten pool include the reference molten pool width, reference molten pool length, reference molten pool area, reference molten pool height, and reference weld bead width. Simultaneously, based on the degree of dispersion in the substrate state distribution data of the current cladding control zone, the geometric allowable fluctuation amount corresponding to each molten pool geometric reference value is determined. The rule for determining the geometric allowable fluctuation amount is: the state dispersion value is multiplied by a preset fluctuation coefficient. This fluctuation coefficient varies for different geometric features, and the specific value is set by the pre-set personnel according to the current application scenario. The lower limit value is obtained by subtracting the corresponding allowable fluctuation amount from each reference value, and the upper limit value is obtained by adding the allowable fluctuation amount, thus forming a closed interval, which is the target molten pool geometric range of the current cladding control zone. This range contains five independent intervals, corresponding to five different molten pool geometric feature parameters.

[0067] After obtaining the target molten pool geometric range, the molten pool geometric characteristic parameters actually measured in step S3 are compared with this range item by item: For each molten pool geometric characteristic parameter, it is determined whether the actual measured value falls within the corresponding target interval. If the actual measured value is greater than the upper limit of the interval, the out-of-bounds direction is determined to be a positive out-of-bounds direction, and the difference between the actual measured value and the upper limit of the interval is taken as the molten pool out-of-bounds amount. If the actual measured value is less than the lower limit of the interval, the out-of-bounds direction is determined to be a negative out-of-bounds direction, and the difference between the lower limit of the interval and the actual measured value is taken as the molten pool out-of-bounds amount. If the actual measured value is within the interval, the out-of-bounds direction is marked as a no-out-of-bounds direction, and the corresponding molten pool out-of-bounds amount is recorded as zero.

[0068] Next, the aforementioned out-of-bounds direction and the amount of molten pool out-of-bounds are classified into deviation levels to obtain the corresponding molten pool geometric deviation levels. The specific method for classifying deviation levels is as follows: a deviation level threshold table is preset, which divides the numerical range of the molten pool out-of-bounds amount into multiple levels. For example, the percentage of the molten pool out-of-bounds amount relative to the corresponding geometrically permissible float is used as the classification basis. When the percentage is between 0% and 30%, it is classified as Level 1 deviation; between 30% and 70%, it is classified as Level 2 deviation; between 70% and 100%, it is classified as Level 3 deviation; and exceeding 100%, it is classified as Level 4 deviation. The out-of-bounds direction and the deviation level together form a binary pair, such as a positive Level 2 deviation or a negative Level 1 deviation.

[0069] The cross-boundary direction, cross-boundary amount, and geometric deviation level corresponding to each molten pool geometric feature parameter are summarized to generate the molten pool geometric deviation parameter for the current cladding control area. This deviation parameter is a data structure that records the above three items for each of the five geometric features. For items with no cross-boundary direction, the molten pool cross-boundary amount and deviation level are zero.

[0070] After generating the molten pool geometric deviation parameters, the process of determining the deposition quality variation trend in step S4 continues. First, based on the correlation between each molten pool geometric characteristic parameter and its corresponding boundary direction, boundary crossing amount, and molten pool geometric deviation level, a deposition impact category is assigned to each molten pool geometric characteristic parameter. The deposition impact categories are pre-divided into three categories: spreading impact category, accumulation impact category, and coverage impact category. The spreading impact category is related to the molten pool width, molten pool length, and molten pool area, as these parameters reflect the molten pool's spreading ability on the base material surface. The accumulation impact category is related to the molten pool height, as the molten pool height directly determines the accumulation thickness of a single cladding layer. The coverage impact category is related to the weld bead width, as the weld bead width determines the overlap coverage state between adjacent weld beads. Based on this association rule, each melt pool geometric feature parameter is assigned a corresponding depositional impact category label, and a consistency check is performed on the outbound directions of all melt pool geometric feature parameters under that category: the number of parameters in an outbound state under that category is counted, and the number of parameters with positive and negative outbound directions is counted separately. If all outbound parameters have the same outbound direction, i.e., all are positive outbound or all are negative outbound, the consistency check passes, and this unified direction is directly used as the basis for subsequent direction aggregation under that category. If there are two opposite outbound directions, positive and negative, under that category, the consistency check fails, and a unique depositional change direction needs to be determined according to the preset outbound direction conflict resolution rules.

[0071] The specific process for determining the pre-defined boundary crossing conflict resolution rules is as follows: Pre-designed personnel design orthogonal experiments for each type of deposition impact. Taking the spreading impact category as an example, this category includes three geometric characteristic parameters: weld pool width, weld pool length, and weld pool area. Multiple parameter combinations are set in the experiments, causing some geometric features to exhibit positive boundary crossings and others to exhibit negative boundary crossings. The final cladding layer quality after welding continues under each conflict condition is recorded. Quality evaluation indicators include weld surface smoothness, cladding layer hardness uniformity, and cross-sectional metallographic fusion state, which are verified and determined by pre-designed personnel. By comparing and analyzing the quality results under different conflict resolution strategies, the optimal conflict resolution rule is determined.

[0072] Using the same experimental method, conflict resolution rules for the buildup effect category and the coverage effect category were determined. The buildup effect category only includes one geometric feature parameter—the weld pool height—there is no possibility of directional conflict, and therefore no conflict resolution rules need to be set. The coverage effect category only includes one geometric feature parameter—the weld bead width—and similarly, there is no directional conflict.

[0073] When opposite directions of boundary crossing occur within a certain depositional impact category, the unique depositional change direction is determined according to the following priority order: For the spreading impact category, first check the boundary crossing direction of the melt pool width. If the melt pool width crosses the boundary positively, the depositional change direction for this category is determined as the overspreading direction within the spreading change directions; if the melt pool width crosses the boundary negatively, it is determined as the underspreading direction; if the melt pool width does not cross the boundary, continue checking the boundary crossing direction of the melt pool length, and determine the depositional change direction according to the same logic; if the melt pool length also does not cross the boundary, the actual boundary crossing direction of the melt pool area is taken as the standard. For the depositional impact category and the cover impact category, since each contains only one geometric characteristic parameter, the actual boundary crossing direction of that parameter is directly taken as the depositional change direction.

[0074] After resolving the aforementioned conflicts, a uniquely determined cross-boundary direction is obtained for each sedimentary impact category. If all melt pool geometric parameters within a category do not cross the boundary, the sedimentary change direction for that category is marked as unchanged. If a cross-boundary occurs, the sedimentary change direction is determined based on this unique cross-boundary direction: For the spreading impact category, when the cross-boundary direction is positive, the sedimentary change direction is an overspreading direction; when it is negative, it is an underspreading direction. For the deposition impact category, a positive cross-boundary corresponds to an overdeposition direction, and a negative cross-boundary corresponds to an underdeposition direction. For the cover impact category, a positive cross-boundary corresponds to an overcover direction, and a negative cross-boundary corresponds to an undercover direction.

[0075] Simultaneously, based on the melt pool incursion amount and melt pool geometric deviation level, the intensity of the melt pool geometric characteristic parameters under each sedimentary impact category is calculated to obtain the corresponding change intensity. The specific method for intensity calculation is as follows: For the same sedimentary impact category, firstly, the melt pool incursion amount and deviation level of all parameters in the incursion state under that category are collected. Then, the contribution intensity value of each parameter is calculated. The contribution intensity value is equal to the melt pool incursion amount of that parameter multiplied by its deviation level coefficient, which is determined by the deviation level. Finally, the contribution intensity values ​​of all parameters under that category are weighted and summed. After normalization, the summation result yields a value between 0 and 1, which is the change intensity of the corresponding sedimentary impact category. The change intensity is specifically divided into spreading change intensity, depositional change intensity, and cover change intensity, corresponding to the calculation results for the spreading impact category, depositional impact category, and cover impact category, respectively.

[0076] Finally, the intensity of change for each sedimentary influence category is combined with its corresponding sedimentary change direction to generate the current sedimentary quality change trend in the cladding control area. The combination method is as follows: the sedimentary change direction and intensity are both represented as a directional intensity value; for categories with no change in sedimentary change direction, the change intensity is zero. The information from these three categories is packaged into a data structure, namely the sedimentary quality change trend. This trend accurately describes the degree and direction of deviation between the actual and desired sedimentary states in the current cladding control area in terms of spread, accumulation, and cover, and is transmitted as output to step S5 for feedback adjustment.

[0077] After generating the deposition quality change trend of the current cladding control area in step S4, step S5 is executed. Based on the deposition quality change trend, a process parameter feedback adjustment amount is generated to adjust the initial process parameters of the current cladding control area. The updated parameters are then used as the basis for updating the initial process parameters of the next cladding control area. The specific implementation process is as follows.

[0078] The analysis step S4 outputs the deposition quality change trend, which includes the spreading change direction and its corresponding spreading change intensity, the accumulation change direction and its corresponding accumulation change intensity, and the cover change direction and its corresponding cover change intensity. Each change direction has three possible states: overspreading or underspreading, over-accumulation or under-accumulation, over-covering or under-covering, or no change.

[0079] Based on the specific direction of change within the trend of sediment quality variation, determine the categories of process parameters that need to be adjusted. The determination rules are as follows: When the deposition quality variation trend includes a spreading direction, whether it's overspreading or underspreading, the hard-face welding parameters and relative motion parameters are identified as the categories of process parameters to be adjusted. Hard-face welding parameters include welding current and welding voltage, which directly affect the heat input and thus control the spreading ability of the weld pool. Relative motion parameters include roller rotation speed and the welding torch's movement speed along the roller axis. The roller rotation speed determines the circumferential linear velocity of the weld, and the welding torch's movement speed determines the axial advance speed of the weld bead. Both together determine the duration of the welding heat source's effect on the roller surface and the heat input distribution. In actual control, there is a one-to-one correspondence between the welding torch's movement speed and the axial feed per revolution of the roller; either can be selected based on the system configuration. The level of adjustment detail is determined by the intensity of the spreading variation: if the intensity of the spreading variation is greater than a preset intensity linkage threshold, both the hard-face welding parameters and the relative motion parameters are adjusted simultaneously; if the intensity of the spreading variation is less than or equal to this threshold, the hard-face welding parameters are adjusted first. The method for determining the strength linkage threshold involves pre-selecting samples with different spread variation in strength for welding tests. The tests assess the corrective effects on molten pool geometry deviations by adjusting only the hardfacing welding parameters and by adjusting both types of parameters simultaneously. By comparing the corrective effect with the adjustment cost, the Pareto optimality principle is used to find the strength threshold that maximizes the corrective efficiency.

[0080] When the deposition quality trend includes the direction of deposition change, whether it is over-deposition or under-deposition, the cladding material supply parameters and relative motion parameters are identified as the categories of process parameters to be adjusted. The cladding material supply parameters include wire feed speed or powder delivery rate, which directly determines the amount of filler metal entering the molten pool per unit time, thus affecting the molten pool height and deposition thickness. The welding torch movement speed, a relative motion parameter, also affects the amount of cladding material deposited per unit length of weld bead.

[0081] When the deposition quality variation trend includes a change in coverage direction, whether it's over-coverage or under-coverage, the weld bead overlap parameter and relative motion parameter are identified as the categories of process parameters to be adjusted. The weld bead overlap parameter includes the percentage overlap between adjacent weld beads, which directly determines the overlap width and coverage state between weld beads. The welding torch speed or equivalent roller axial feed in the relative motion parameter determines the distance the welding torch offsets after each revolution, thus affecting the actual overlap rate.

[0082] After determining the category of process parameters to be adjusted, based on the specific direction of change in the deposition quality trend, and according to the preset correspondence between deposition change direction and parameter adjustment direction, the parameter adjustment direction for the corresponding process parameter category is determined. The method for determining this preset correspondence is as follows: Single-factor process experiments are conducted for each deposition change direction. For example, with other parameters fixed, the welding current is increased or decreased sequentially, and the change pattern of the molten pool spreading state is observed. When the welding current increases, both the width and length of the molten pool show an increasing trend. Therefore, the direction of excessive spreading corresponds to decreasing the welding current, and the direction of insufficient spreading corresponds to increasing the welding current. Similarly, single-factor experiments are conducted on parameters such as welding voltage, wire feed speed, welding torch movement speed, and overlap rate, recording the influence trend of each parameter change on the spreading, accumulation, and coverage states. All experimental results are summarized to form a complete correspondence table, which is then solidified. The specific correspondence is as follows: For the over-spreading direction, where the actual weld pool width, length, or area exceeds the upper limit of the target range, it indicates excessive heat input or excessive action time. In this case, parameter adjustment should focus on reducing the welding current and welding voltage in hardfacing welding parameters, or increasing the welding torch speed or roller speed in relative motion parameters to reduce heat input per unit length. For the under-spreading direction, where the actual weld pool geometry is less than the lower limit of the target range, it indicates insufficient heat input. Parameter adjustment should be done in the opposite direction: increase the welding current and welding voltage, or decrease the welding torch speed and roller speed.

[0083] For the over-accumulation direction, where the actual molten pool height exceeds the upper limit of the target range, it indicates that too much cladding material is supplied per unit length. The parameter adjustment direction is to reduce the wire feed speed or powder delivery rate in the cladding material supply parameters, or to increase the welding torch movement speed in the relative motion parameters to reduce the deposition per unit length. For the under-accumulation direction, the parameter adjustment direction is the opposite: increase the wire feed speed or slow down the welding torch movement speed.

[0084] For the over-coverage direction, where the actual weld bead width exceeds the upper limit of the target range, it indicates excessive overlap between adjacent weld beads, potentially resulting in excessively high weld beads or an uneven surface. The parameter adjustment direction is to reduce the overlap rate in the weld bead overlap parameters, or increase the axial feed rate or welding torch traverse step size in the relative motion parameters. For the under-coverage direction, where the actual weld bead width is less than the lower limit of the target range, it indicates gaps between adjacent weld beads or insufficient overlap. The parameter adjustment direction is the opposite: increase the overlap rate or decrease the axial feed rate.

[0085] While determining the direction of parameter adjustment, the adjustment amplitude level for the corresponding process parameter category is determined based on the intensity of change in the deposition quality trend. The boundary values ​​of the intervals used for classifying the adjustment amplitude levels are determined as follows: Multiple groups of samples with different intensities of change are selected by pre-selected personnel. For each group of samples, correction experiments are conducted using three different adjustment amplitudes: fine-tuning, medium-tuning, and emphasis. Fine-tuning is defined as an adjustment amount of 2% to 5% of the parameter's rated adjustment range; medium-tuning is defined as 5% to 10%; and emphasis is defined as 10% to 15%. The overall effectiveness of each adjustment amplitude is calculated by measuring the time required for the geometric characteristic parameters of the melt pool to recover to the target range under each adjustment amplitude, as well as quality indicators such as the hardness uniformity and dilution rate of the final cladding layer. The overall effectiveness is defined as the weighted sum of the cladding quality pass rate and the adjustment response speed. For the intensity of spread change, accumulation change, and overlay change, the corresponding adjustment amplitude level is determined according to the interval into which their values ​​fall. If a certain direction of change is unchanged, its intensity of change is zero, and the corresponding adjustment amplitude level is zero, i.e., no adjustment is made.

[0086] After obtaining the adjustment range level, the corresponding parameter adjustment range is further determined based on the adjustment range level. The parameter adjustment range refers to the specific amount of adjustment made to a specific process parameter value. An internally pre-defined table mapping adjustment range levels to adjustment step sizes is used. This table is determined as follows: based on previous single-factor process test data, for each process parameter, the correction effect of different absolute adjustment amounts or relative adjustment percentages on the geometric characteristic parameters of the molten pool is tested. Taking welding current as an example, adjustment step sizes are set to 1%, 2%, 3%...15% of the current value. The number of adjustment cycles and overshoot required for the molten pool width to recover to the target range under each step size are measured. The maximum step size that can complete the correction within two adjustment cycles without overshoot is selected as the adjustment range corresponding to the fine-tuning level; the maximum step size that can complete the correction within a single adjustment cycle without overshoot is selected as the adjustment range corresponding to the medium-tuning level; and the minimum step size that can complete the correction within half an adjustment cycle with an allowable overshoot of no more than 5% is selected as the adjustment range corresponding to the emphasis level. Through this experimental method, the specific adjustment step size for each process parameter at the three adjustment range levels is determined.

[0087] Based on the category of the process parameter to be adjusted and the corresponding adjustment range level, the specific parameter adjustment range is retrieved from the preset correspondence table. The sign of the adjustment range, i.e., increase or decrease, is determined by the aforementioned parameter adjustment direction: a negative adjustment range is taken when the direction is decrease, and a positive adjustment range is taken when the direction is increase.

[0088] When multiple depositional change directions simultaneously require adjustment of the same relative motion parameter, the final feedback adjustment amount is determined according to the following priority: spreading change direction takes precedence over depositional change direction, and depositional change direction takes precedence over overburden change direction. First, check if a spreading change direction exists. If it does, only the adjustment direction and magnitude corresponding to the spreading change direction are used, ignoring the requirements of depositional and overburden change directions. If no spreading change direction exists but a depositional change direction exists, the adjustment direction and magnitude corresponding to the depositional change direction are used. If only an overburden change direction exists, the adjustment direction and magnitude corresponding to the overburden change direction are used.

[0089] For other parameters that may be adjusted simultaneously by multiple changing directions, the same priority applies: hardfacing welding parameters respond first to spreading changes; cladding material supply parameters respond first to deposition changes. This priority rule is based on the following: spreading changes directly reflect heat input and molten pool fluidity, having the greatest impact on the basic quality of weld bead formation; deposition changes affect cladding layer thickness and are adjusted only if spreading is satisfactory; and overlay changes affect weld overlap and have the lowest priority.

[0090] The process parameter categories, adjustment directions, and adjustment ranges identified above are correlated to generate the process parameter feedback adjustment amounts for the current cladding control zone. The process parameter feedback adjustment amount is a data structure that records the name of each specific process parameter requiring adjustment, its adjustment direction indicator, and the absolute or relative value of its adjustment range. For parameter categories that do not require adjustment, their feedback adjustment amount is zero.

[0091] The initial process parameters of the current cladding control zone are combined with the process parameter feedback adjustment amount generated in step S5 to obtain the updated cladding process parameters. The combination method is as follows: for each specific process parameter to be adjusted, its initial value is added to or subtracted from the adjustment range specified in the feedback adjustment amount to calculate the new parameter value; for parameters not appearing in the feedback adjustment amount, their initial values ​​remain unchanged. The new values ​​of all parameters together constitute the updated cladding process parameters. The updated cladding process parameters are associated with the location range of the current cladding control zone, i.e., a mapping record is established, with the key being the identifier of the current cladding control zone or its location range, and the value being the updated cladding process parameter. In subsequent welding cladding processes, when entering the next cladding control zone, it is first checked whether there is an updated parameter associated with the location range of that control zone. If there is, the updated cladding process parameter is used as the basis for updating the initial process parameters of the next cladding control zone, i.e., the updated parameter is directly used as the initial process parameter of the next control zone, without repeating the process of selecting candidate parameter groups from the preset correspondence in step S2. If there is no parameter, the initial process parameters for the control zone are determined according to the normal process of step S2.

[0092] like Figure 5 The diagram shown is a structural schematic of a roller hard surface welding and cladding system provided in an embodiment of the present invention. The roller hard surface welding and cladding system provided in this embodiment of the present invention includes the following modules: The position calibration analysis module is used to perform position calibration analysis on the pre-treated surface of the roll body to be fused and clad, and to obtain the substrate state distribution data of the corresponding outer circular surface of the roll body to be fused and clad.

[0093] The initial cladding preprocessing module is used to divide the outer circular surface of the roller to be clad into cladding control zones based on the substrate state distribution data, determine the corresponding initial process parameters for each cladding control zone, and perform matching verification of the initial process parameters based on the substrate state distribution data of the corresponding cladding control zone.

[0094] The molten pool geometric feature analysis module is used to acquire molten pool image data and / or molten pool contour data of the current cladding control zone during the welding cladding process based on initial process parameters, and extract molten pool geometric feature parameters from them.

[0095] The sedimentation quality change analysis module is used to compare the geometric characteristic parameters of the melt pool with the target melt pool geometric range corresponding to the current cladding control area, generate melt pool geometric deviation parameters, and determine the sedimentation quality change trend of the current cladding control area based on the melt pool geometric deviation parameters.

[0096] The cladding feedback adjustment module is used to generate process parameter feedback adjustment amounts based on the deposition quality change trend, so as to adjust the initial process parameters of the current cladding control area, obtain the updated cladding process parameters, and use them as the basis for updating the initial process parameters of the next cladding control area.

[0097] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0098] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0099] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0100] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0101] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for hard-surface welding and cladding of a roller body, characterized in that, Includes the following steps: S1, perform position calibration analysis on the roller body to be clad after surface pretreatment, and obtain the substrate state distribution data of the outer circular surface of the roller body to be clad. The substrate state distribution data is used to characterize the differences in substrate morphology, surface transition state and hard surface welding cladding adaptation state at different positions on the outer circular surface. S2, based on the substrate state distribution data, divide the outer circular surface of the roller body to be clad into cladding control zones, determine the corresponding initial process parameters for each cladding control zone, and perform matching verification based on the substrate state distribution data of the corresponding cladding control zone. The initial process parameters include one or more of the following: hardface welding parameters, cladding material supply parameters, relative motion parameters, and weld bead lap parameters; S3, during the welding cladding process based on the initial process parameters, acquire the molten pool image data and / or molten pool contour data of the current cladding control zone, and extract the molten pool geometric feature parameters from them. The molten pool geometric feature parameters include at least one or more of the following: molten pool width, molten pool length, molten pool area, molten pool height, and weld bead forming width. S4, compare the geometric characteristic parameters of the molten pool with the geometric range of the target molten pool corresponding to the current cladding control area to generate molten pool geometric deviation parameters, and determine the deposition quality change trend of the current cladding control area based on the molten pool geometric deviation parameters; S5. Based on the deposition quality change trend, a process parameter feedback adjustment amount is generated to adjust the initial process parameters of the current cladding control area, thereby obtaining the updated cladding process parameters, which serve as the basis for updating the initial process parameters of the next cladding control area.

2. The roller body hard surface welding cladding method as described in claim 1, characterized in that, S1 includes: Obtain the axial reference position, circumferential reference position, and outer diameter circumference of the roll body to be melted and coated. Establish axial position coordinates using the axial reference position as the axial zero point and establish circumferential angle coordinates using the circumferential reference position as the circumferential zero angle. Based on the outer diameter circumference of the roll body to be clad, the circumferential angle coordinates are converted into the corresponding circumferential unfolded position, and the outer circular surface unfolded coordinates of the roll body to be clad are generated based on the axial position coordinates and the circumferential unfolded position. On the outer circular surface unfolded coordinate system, the surface is divided into units using a preset weld width and sampling interval to obtain position calibration units; The outer circular surface detection data corresponding to each position calibration unit is obtained respectively. The outer circular surface detection data includes one or more of the following: surface contour data, local grinding state data, old hard surface layer residual state data, base metal exposed state data, and surface transition state data. The calibration units at each location are associated with the corresponding outer circular surface detection data to generate substrate state distribution data.

3. The roller body hard surface welding cladding method as described in claim 2, characterized in that, The specific process for dividing the cladding control zone is as follows: The substrate state distribution data is analyzed to obtain the substrate detection features of each location calibration unit and normalize them to obtain the substrate morphology state identifier used to characterize the substrate morphology difference, and the surface transition state identifier used to characterize the transition difference between the old hard surface layer and the substrate metal. The substrate morphology status identifier and surface transition status identifier corresponding to each position calibration unit are combined to obtain the corresponding position status combination data. The position status combination data is then matched with the preset state combination-cladding adaptation correspondence to determine the corresponding position adaptation category. Generate a hard surface welding cladding adaptation status identifier for each position calibration unit based on the position adaptation category; When the hard surface welding cladding adaptation status indicator of the position calibration unit meets the preset adaptation similarity conditions, the corresponding position calibration unit is assigned to the same cladding control area. By summarizing all position calibration units, the cladding control zone on the outer circular surface of the roller to be clad is determined, and the corresponding initial process parameters are determined.

4. The roller body hard surface welding cladding method as described in claim 3, characterized in that, The determination of the corresponding initial process parameters includes generating a set of process parameters to be verified for the corresponding cladding control area. The specific process is as follows: The distribution ratio of different hardface welding cladding adaptation status identifiers within the same cladding control area is statistically analyzed to generate regional adaptation status distribution data for the corresponding cladding control area. Based on the regional adaptation state distribution data, the dominant adaptation state and state dispersion of the corresponding cladding control area are determined, and process parameter requirement data for the corresponding cladding control area are generated based on the dominant adaptation state and state dispersion. Based on the process parameter requirement data, the heat input requirement, cladding material supply requirement, relative interaction time requirement and weld bead coverage requirement of the corresponding cladding control zone are quantified to generate the process parameter requirement level of the corresponding cladding control zone. Based on the process parameter requirement level, a candidate process parameter group matching the process parameter requirement level is selected from the preset process parameter correspondence. The candidate process parameter groups are screened for applicability to obtain one or more process parameter groups to be verified that meet the requirements of substrate state, welding cladding continuity and cladding control zone transition in the corresponding cladding control zone. When multiple sets of process parameters to be verified are obtained, the parameter change range corresponding to each set of process parameters to be verified is determined, and the set of process parameters to be verified with the lowest parameter change range is selected as the set of process parameters to be verified for the corresponding cladding control area.

5. The roller body hard surface welding cladding method as described in claim 4, characterized in that, The matching verification process is as follows: Based on the substrate state distribution data of the cladding control area, the corresponding substrate detection features are extracted and regional statistical processing is performed to obtain the corresponding cladding control area adaptation statistics. The adaptation statistics include substrate morphology statistics, surface transition statistics and hard surface welding cladding adaptation statistics. Based on the adaptation statistics, the constraint level of the corresponding cladding control area is determined, and the process parameter adaptation conditions of the corresponding cladding control area are generated according to the constraint level. The process parameter adaptation conditions are used to limit the parameter range and parameter combination relationship that the process parameter group to be verified should meet. The process parameter group to be verified is matched with the process parameter adaptation conditions. When the preset matching requirements are met, the process parameter group to be verified is determined as the initial process parameters of the corresponding cladding control area. When the preset matching requirements are not met, the parameter group to be verified is reselected, and the reselected parameter group to be verified is re-verified until the preset matching requirements are met or the maximum number of reselections is reached.

6. The roller body hard surface welding cladding method as described in claim 5, characterized in that, The S3 includes: Based on the outer circular surface development coordinates of the roller to be clad and the position range of each cladding control zone, real-time cladding position data during the welding cladding process is obtained, and the current cladding control zone to which the current cladding position belongs is determined according to the real-time cladding position data. Based on the location range of the current cladding control area, the corresponding molten pool data acquisition window is determined. The molten pool data acquisition window is used to limit the acquisition location and acquisition sequence of molten pool image data and / or molten pool contour data within the current cladding control area. Within the molten pool data acquisition window, acquire molten pool image data and / or molten pool contour data of the current cladding control area, and preprocess the molten pool image data and / or molten pool contour data to obtain the basic molten pool data corresponding to the current cladding control area; The molten pool basic data is used to identify the molten pool region and determine the molten pool boundary data corresponding to the current cladding control area; Based on the molten pool boundary data, determine one or more of the following: molten pool width, molten pool length, molten pool area, molten pool height, and weld bead forming width, and use them as the geometric characteristic parameters of the molten pool in the current cladding control zone.

7. The roller body hard surface welding cladding method as described in claim 6, characterized in that, The S4 includes: The initial process parameters are decoupled and normalized to generate the process parameter feature vector of the current cladding control area; The process parameter feature vector is matched with the preset process parameter-molten pool geometry mapping relationship to determine the molten pool geometry reference value corresponding to the current cladding control area and its corresponding geometric allowable fluctuation amount, thereby generating the target molten pool geometry range of the current cladding control area. The geometric characteristic parameters of the molten pool are compared with the target geometric range item by item to determine the boundary direction and boundary crossing amount of the geometric characteristic parameters of the molten pool relative to the corresponding target geometric range. When the geometric characteristic parameters of the molten pool are within the corresponding target molten pool geometric range, the corresponding out-of-bounds direction is marked as a non-out-of-bounds direction, and the corresponding molten pool out-of-bounds amount is recorded as zero; The deviation levels of the out-of-bounds direction and the out-of-bounds amount of the molten pool are divided to obtain the corresponding geometric deviation levels of the molten pool. Based on the molten pool geometric deviation level, the direction of the boundary crossing, and the amount of the molten pool boundary crossing, the molten pool geometric deviation parameters of the current cladding control zone are generated.

8. The roller body hard surface welding cladding method as described in claim 7, characterized in that, The S4 further includes: Based on the correlation between each molten pool geometric feature parameter in the molten pool geometric deviation parameters and the corresponding cross-boundary direction, molten pool cross-boundary amount and molten pool geometric deviation level, the molten pool geometric feature parameters are labeled with deposition impact categories; Under the same depositional impact category, the geometric characteristic parameters of the molten pool are oriented according to the boundary crossing direction to obtain the corresponding depositional change direction. Based on the boundary crossing amount and the geometric deviation level of the molten pool, the intensity of the geometric characteristic parameters of the molten pool is calculated to obtain the corresponding change intensity. The intensity of change includes spreading intensity, deposition intensity, and overlay intensity, and the direction of deposition change includes spreading direction, deposition direction, and overlay direction; The intensity of the change is combined with its corresponding direction of sedimentary change to generate the sedimentary quality change trend of the current cladding control area.

9. The roller body hard surface welding cladding method as described in claim 8, characterized in that, The S5 includes: When the deposition quality change trend includes the spreading change direction, the hard surface welding parameters and / or relative motion parameters are determined as the process parameter categories to be adjusted; When the deposition quality change trend includes the direction of accumulation change, the cladding material supply parameters and / or relative motion parameters are determined as the process parameter categories to be adjusted. When the deposition quality change trend includes the coverage change direction, the weld overlap parameter and / or relative motion parameter are determined as the process parameter category to be adjusted; Based on the spreading change direction, the deposition change direction, or the coverage change direction, and based on the preset correspondence between the deposition change direction and the parameter adjustment direction, the parameter adjustment direction for the corresponding process parameter category is determined. Based on the corresponding change intensity, the adjustment magnitude level for the corresponding process parameter category is determined, and the corresponding parameter adjustment magnitude is determined based on the adjustment magnitude level. The process parameter category, the parameter adjustment direction, and the parameter adjustment range are associated to generate the process parameter feedback adjustment amount for the current cladding control zone; The initial process parameters and process parameter feedback adjustment values ​​of the current cladding control area are combined to obtain updated cladding process parameters. These updated cladding process parameters are then associated with the location range of the current cladding control area, serving as the basis for updating the initial process parameters of the next cladding control area.

10. A roller hardface welding and cladding system, employing the roller hardface welding and cladding method as described in any one of claims 1-9, comprising the following modules: The position calibration analysis module is used to perform position calibration analysis on the roller body to be melted and coated after surface pretreatment, and to obtain the substrate state distribution data of the outer circular surface of the roller body to be melted and coated. The initial cladding preprocessing module is used to divide the outer circular surface of the roller to be clad into cladding control zones according to the substrate state distribution data, and to determine the corresponding initial process parameters for each cladding control zone. At the same time, the initial process parameters are checked for matching according to the substrate state distribution data of the corresponding cladding control zone. The molten pool geometric feature analysis module is used to acquire molten pool image data and / or molten pool contour data of the current cladding control area during the welding cladding process based on the initial process parameters, and extract molten pool geometric feature parameters from them; The deposition quality change analysis module is used to compare the geometric characteristic parameters of the melt pool with the target melt pool geometric range corresponding to the current cladding control area, generate melt pool geometric deviation parameters, and determine the deposition quality change trend of the current cladding control area based on the melt pool geometric deviation parameters. The cladding feedback adjustment module is used to generate process parameter feedback adjustment amounts based on the deposition quality change trend, so as to adjust the initial process parameters of the current cladding control area, obtain updated cladding process parameters, and use them as the basis for updating the initial process parameters of the next cladding control area.

Images