A method for generating and executing control instructions of a slitting device

Abstract

The application provides a slitting equipment control instruction generation and execution scheduling method, comprising: collecting and fusing tension values, warping height values, thickness distribution data and surface finish in the steel strip conveying process, identifying the bending stress residual state of the steel strip caused by uneven thickness or winding storage, and generating a steel strip comprehensive abnormal state evaluation result; according to the center line offset distance and position distribution, combining the regional thickness difference identified by the thickness distribution data, marking the deformation region in the width direction of the steel strip, and extracting the stable section position with a warping height deviation value less than a preset threshold and relatively uniform thickness; determining the dynamic reference line position suitable for the real-time shape change of the steel strip based on the stable section position data; generating the transverse movement distance control instruction of the slitting cutter and the corresponding feed depth preset value according to the dynamic reference line position and the thickness distribution data, and obtaining the difference value of the control instruction and the actual position of the cutter.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a method for generating and scheduling control instructions for striping equipment. Background Technology

[0002] In the metal processing industry, slitting equipment control technology is crucial for the processing accuracy and product quality of steel strips, especially in industries such as automobiles and home appliances, where high-precision steel strip slitting directly affects the performance and reliability of the final product. During the slitting process, the physical properties of the steel strip, such as thickness, surface finish, and tension distribution, significantly affect the processing results. However, existing methods often rely on single sensor data or fixed control parameters, making it difficult to adapt to the dynamic changes in the steel strip caused by its inherent characteristics during transport. For example, when the equipment relies solely on position sensors to monitor the edge position of the steel strip, it cannot simultaneously detect changes in the tension distribution within the steel strip. This leads to situations where, even when thickness fluctuations occur in the middle of the steel strip, the fixed tool positioning is still executed based on the edge position data, resulting in cutting deviations. Furthermore, control methods based solely on tension sensors cannot identify the interference of local friction coefficient differences on tension readings when the surface finish of the steel strip is uneven, causing the system to misjudge the actual material stress state. These methods lack flexibility when handling complex deformations. Specifically, a single data source cannot provide complete information on the multi-dimensional deformation of the steel strip. When the steel strip exhibits both longitudinal tension fluctuations and lateral warping simultaneously, a control system based on single tension data cannot distinguish between these two different types of deformation. It can only respond according to a preset fixed logic and cannot dynamically adjust the tool position and cutting depth based on the real-time morphological changes of the steel strip, leading to decreased slitting accuracy or machining defects. The core technical challenge lies in accurately identifying the steel strip's baseline to determine the tool's cutting path and positioning reference. Uneven thickness of the steel strip or residual bending stress generated during winding and storage can cause the steel strip's centerline to shift. In particular, when tension fluctuations exceed the elastic deformation range of the steel strip material, local warping further exacerbates the difficulty of baseline identification. This complex deformation renders traditional fixed-reference-point identification methods ineffective, making it difficult to accurately determine the tool's feed depth and lateral movement distance for precise cutting positioning. Summary of the Invention

[0003] This invention provides a method for generating and scheduling control instructions for striping equipment, mainly including:

[0004] By integrating tension values, warpage height values, thickness distribution data, and surface finish data during the steel strip conveying process, a comprehensive abnormal state assessment result for the steel strip is generated. Based on this comprehensive abnormal state assessment result, the longitudinal and transverse local deformation distributions in the warpage height values ​​and thickness distribution data are extracted. Combined with the friction coefficient variation areas determined by the surface finish data, the distance and position distribution of the steel strip centerline offset are determined. According to the distance and position distribution of the centerline offset, combined with the regional thickness differences in the thickness distribution data, the deformation areas in the width direction of the steel strip are marked, and the warpage height deviation value is extracted. The system identifies stable sections where thickness uniformity meets preset conditions; determines dynamic baseline positions based on the distance and positional distribution of the centerline offset of the stable sections; generates lateral movement distance control commands and preset feed depth values ​​for the slitting tool according to the dynamic baseline positions and thickness deviation areas in the thickness distribution data; iteratively updates the lateral movement distance control commands and preset feed depth values ​​based on the stable sections and the centerline offset; and schedules the slitting equipment actuators and adjusts the tool trajectory using the lateral movement distance control commands and preset feed depth values.

[0005] Furthermore, the integrated data on tension, warpage height, thickness distribution, and surface finish during the steel strip conveying process are used to generate a comprehensive assessment result of the steel strip's abnormal state, including:

[0006] Tension sensor readings at multiple locations during steel strip conveying are acquired, and the corresponding warpage height measurements are recorded. A thickness gradient map of the steel strip's transverse thickness distribution is generated using ultrasonic scanning, and a surface roughness distribution map is generated by detecting the steel strip's surface roughness. Based on the yield strength and elastic modulus of the steel strip material, a residual stress distribution matrix is ​​calculated. For regions in the residual stress distribution matrix that exceed the yield strength, the warpage height deviation and thickness variation rate at the corresponding locations are extracted to determine the mapping relationship between stress concentration and deformation degree. Based on this mapping relationship, sections with abnormalities in continuous sampling points are identified, and the standard deviation of the warpage height and the coefficient of variation of the thickness distribution within these sections are calculated. An abnormal region is grouped using a clustering algorithm to generate a section anomaly distribution map containing the anomaly type and its impact range. Based on the section anomaly distribution map, the proportion of the anomaly type in the width and length directions of the steel strip is statistically analyzed, and the area ratio of the abnormal region to the normal region is calculated to generate a comprehensive anomaly status assessment result for the steel strip.

[0007] Furthermore, based on the comprehensive abnormal state assessment results, the longitudinal and transverse local deformation distributions in the warping height value and the thickness distribution data are extracted, and the friction coefficient variation area determined by the surface finish data is combined to determine the distance and position distribution of the steel strip centerline offset, including:

[0008] Based on the comprehensive abnormal state assessment results, longitudinal and transverse deformation distribution data are extracted from the warping height values ​​to generate a deformation distribution matrix. For the deformation distribution matrix, combined with the thickness distribution data, thickness abrupt change boundaries in the steel strip width direction are identified. Based on the surface finish data, friction coefficient variation regions are extracted, and the contribution of friction coefficient variation to centerline offset is calculated. Based on the deformation distribution matrix and the friction coefficient variation regions, a linear regression method is used to calculate the centerline offset distance. Based on the centerline offset distance and the thickness abrupt change boundaries, a centerline offset position distribution map is generated, and the offset position coordinates are recorded.

[0009] Furthermore, the step of dividing and marking the deformation areas in the width direction of the steel strip according to the distance and position distribution of the centerline offset, combined with the regional thickness differences in the thickness distribution data, and extracting the warping height deviation value and the location of stable sections where the thickness uniformity meets preset conditions, includes:

[0010] Based on the distance of the centerline offset, the steel strip width direction is divided into regions using a gridding method to generate a region distribution matrix; for each grid in the region distribution matrix, the degree of deviation between the warping height value and a preset threshold is calculated; based on the thickness distribution data, the thickness variation coefficient of adjacent grids is calculated; based on the degree of deviation and the variation coefficient, grids that meet the preset conditions are marked as stable segments.

[0011] Furthermore, determining the dynamic baseline position based on the distance and positional distribution of the centerline offset of the stable segment position includes:

[0012] Extract the center coordinates of the stable section location, and calculate the theoretical offset value by combining the distance of the centerline offset; extract the roughness distribution based on the surface finish data, calculate the contribution value of frictional resistance to the offset, and generate an initial correction vector; adjust the initial correction vector based on the comprehensive abnormal state evaluation results to generate an offset correction amount sequence; based on the offset correction amount sequence, use an interpolation method to generate a baseline correction curve and determine the dynamic baseline position coordinate sequence.

[0013] Furthermore, the step of generating a lateral movement distance control command and a preset feed depth value for the slitting tool based on the position of the dynamic baseline and the thickness deviation area in the thickness distribution data includes:

[0014] Extract the dynamic baseline position coordinate sequence and identify the thickness deviation region boundary in the thickness distribution data; calculate the distance from the thickness deviation region boundary to the dynamic baseline position and generate a lateral movement distance value; calculate the feed depth compensation coefficient according to the degree of deviation of the thickness deviation region and generate a preset feed depth value.

[0015] Furthermore, the iterative update of the lateral movement distance control command and the feed depth preset value based on the stable segment position and the centerline offset distance includes:

[0016] Extract the reliability score of the stable section position to generate a comprehensive reliability index; based on the comprehensive reliability index, adjust the centerline offset distance to generate an updated lateral movement distance value; calculate the offset compensation amount according to the updated lateral movement distance value to generate a corrected feed depth control command sequence; use the feed depth control command sequence to generate a control command format recognizable by the slitting equipment.

[0017] Furthermore, the step of using the lateral movement distance control command and the preset feed depth value to schedule the slitting equipment actuator and adjust the tool trajectory includes:

[0018] Using the lateral movement distance control command and the preset feed depth value, the slitting equipment actuator is driven, and real-time feedback signals from the actuator are obtained. Based on the thickness distribution data and the surface finish data, the tool feed speed and contact pressure are adjusted to generate real-time optimized tool trajectory data. Based on the real-time optimized tool trajectory data, the coordinates of the slitting edge position are collected, the width and straightness parameters of the slitting gap are calculated, and a slitting accuracy matching degree evaluation value is generated. Based on the slitting accuracy matching degree evaluation value, multi-sensor data is integrated to trigger an abnormal state re-evaluation process and update the control command.

[0019] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0020] This invention discloses a method for generating and scheduling control commands for slitting equipment, aiming to solve the problem of insufficient slitting accuracy in business scenarios caused by uneven thickness, residual bending stress, and centerline offset of steel strips. This invention collects multi-source data such as tension, warpage height, thickness distribution, and surface finish, integrates and evaluates the comprehensive abnormal state of the steel strip, accurately identifies centerline offset and local deformation distribution, and then determines the location of stable sections and generates dynamic baselines. Based on this, this invention combines thickness deviation and stress effects to generate control commands for tool lateral movement and feed depth, and ensures that the slitting trajectory matches the preset accuracy through real-time data iterative optimization. If the match is insufficient, the anomaly assessment and correction process is iteratively optimized to ultimately achieve high-precision and adaptive control of steel strip slitting. This invention significantly improves the processing quality of steel strips and the operational stability of equipment, providing reliable technical support for precision manufacturing under complex working conditions. Attached Figure Description

[0021] Figure 1 This is a flowchart of a method for generating and executing control instructions for a striping device according to the present invention.

[0022] Figure 2 This is a schematic diagram of a method for generating and scheduling control instructions for a striping device according to the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0024] like Figures 1-2 This embodiment of a method for generating and scheduling control instructions for striping equipment may specifically include:

[0025] Step S101: Collect and integrate the tension value, warpage height value, thickness distribution data and surface finish during the steel strip conveying process, identify the residual bending stress caused by uneven thickness or winding storage of the steel strip, and generate a comprehensive abnormal state assessment result of the steel strip.

[0026] Real-time readings from tension sensors at multiple locations during the steel strip conveying process are acquired. For each sampling point, the corresponding warpage height is recorded using laser measurement. An ultrasonic thickness gauge scans the transverse thickness distribution of the steel strip to form a thickness gradient map. A roughness meter is used to detect the microscopic undulations of the steel strip surface, generating a surface finish distribution map. The elastic deformation safety threshold is calculated based on the yield strength and elastic modulus of the steel strip material. The storage time and roll diameter parameters of the steel strip are extracted from the winding history. A bending stress residual distribution model is constructed using finite element analysis, obtaining the residual stress numerical matrix for each region of the steel strip. For high-stress regions in the residual stress numerical matrix that exceed a preset proportion of the material's yield strength, the warpage height deviation and thickness change rate at the corresponding locations are extracted. Correlation analysis is used to determine the mapping relationship between stress concentration and deformation degree. If the tension fluctuation amplitude exceeds a preset multiple of the elastic deformation safety threshold, the region is marked as a plastic deformation risk zone. Based on the regions in the surface finish distribution map where the roughness exceeds a preset threshold, abnormal friction coefficient segments are defined, obtaining an abnormal feature vector containing the location of stress exceeding the standard, the deformation risk level, and surface condition indicators. Based on the aforementioned abnormal feature vectors, segments where multiple consecutive sampling points exhibit abnormality markers are identified. The standard deviation of the warp height and the coefficient of variation of the thickness distribution within these segments are calculated. K-means clustering is then used to group areas with similar levels of abnormality. The direction of abnormality propagation is determined based on the temporal variation of tension values ​​within each group. The duration of the abnormality is calculated by combining the steel strip conveying speed and the length of the abnormal segment, thus determining a segment abnormality distribution map that includes the abnormality type, severity, and impact range. Based on the spatial distribution and severity level of each abnormality type in the segment abnormality distribution map, the proportion of different abnormality types in the width and length directions of the steel strip is statistically analyzed. The area ratio of the abnormal area to the normal area is calculated to determine the overall quality status level of the steel strip, resulting in a comprehensive abnormality status assessment of the steel strip.

[0027] Specifically, in one embodiment, multi-source data acquisition during the steel strip conveying process is achieved through a sensor array arranged at different locations on the slitting equipment. Tension sensors employ piezoelectric sensors, with a measuring point arranged every 200 millimeters along the width of the steel strip to monitor the tension distribution of the steel strip in real time during conveying. Warpage height measurement uses a laser displacement sensor, which emits a laser beam perpendicularly to the surface of the steel strip and calculates the distance from the steel strip surface to the sensor based on the time difference of the reflected light, thereby obtaining the warpage deformation data of the steel strip. An ultrasonic thickness gauge moves laterally across the steel strip, using the time difference of ultrasonic waves propagating within the steel strip to determine the thickness value at each location, forming a continuous thickness gradient map.

[0028] Specifically, the process of constructing a bending stress residual distribution model using the finite element method includes: dividing the steel strip into several micro-units, assigning corresponding material property parameters to each unit, including elastic modulus, Poisson's ratio, and yield strength. Based on the roll diameter parameters in the winding history, the radius of curvature of the steel strip in the winding state is calculated, and the degree of stress relaxation is determined in conjunction with the storage time. By solving the stress balance equation for each unit, the residual stress distribution after the steel strip is unwound is obtained. The residual stress numerical matrix records the magnitude and direction of stress at various locations on the steel strip, providing basic data support for subsequent anomaly identification. This modeling method can accurately reflect the internal stress distribution characteristics of the steel strip caused by long-term winding and storage, avoiding the limitations of traditional single-point measurements that cannot obtain the full picture of the stress field.

[0029] It should be noted that the correlation analysis uses the Pearson correlation coefficient method to determine the mapping relationship between stress concentration and deformation degree. By extracting the stress value sequence of the high-stress region and the corresponding warp height deviation value sequence, the linear correlation between the two sequences is calculated. When the absolute value of the correlation coefficient is greater than a preset threshold, a mapping relationship between stress value and deformation is established, which describes the degree of deformation response caused by a unit stress change.

[0030] In one possible implementation, the construction process of the anomaly feature vector comprehensively considers multiple dimensions of anomaly indicators. The marking of the plastic deformation risk zone is based on the principles of material mechanics. When the tension fluctuation amplitude exceeds a preset multiple of the elastic deformation safety threshold, it indicates that the steel strip has locally approached or entered the plastic deformation stage, at which point the material deformation will not be fully recoverable. The definition of the friction coefficient anomaly segment is achieved by analyzing surface roughness data. The higher the roughness of the area, the greater the friction coefficient between the steel strip and the conveyor roller, affecting the uniformity of tension transmission. The anomaly feature vector encodes this multi-dimensional information into a unified data structure, facilitating subsequent cluster analysis. The implementation process of the K-means clustering algorithm in grouping and classifying anomaly regions is as follows: First, a cluster number K value is set, which is determined based on empirical values ​​of the steel strip width and anomaly type. K anomaly feature vectors are randomly selected as initial cluster centers. The Euclidean distance from each anomaly region to each cluster center is calculated, and the anomaly region is assigned to the group containing the nearest cluster center. The mean of all abnormal feature vectors within each group is recalculated as the new cluster center. Distance calculation and center update processes are iteratively executed until the cluster centers no longer change significantly or the maximum number of iterations is reached. This clustering method groups regions with similar abnormal features together, facilitating targeted treatment. The direction of anomaly propagation is determined by analyzing the temporal variation of tension values. During continuous steel strip transport, anomalies at a certain location can propagate downstream with the strip's movement. By comparing the intensity changes of the same abnormal feature at different locations at adjacent sampling times, it is determined whether the anomaly spreads along the transport direction or propagates in the reverse direction. Combined with the steel strip transport speed, the time span from the anomaly's occurrence to its disappearance is calculated, assessing the potential impact of the anomaly on subsequent processing. The generation of the segment anomaly distribution map involves extracting the spatial features of the anomaly segments. The segment anomaly distribution map is represented in two-dimensional matrix form, where rows correspond to the length direction of the steel strip, columns correspond to the width direction, and matrix element values ​​represent the severity level of the anomaly at that location. Through visualization, operators can intuitively understand the overall quality distribution of the steel strip.

[0031] For example, when calculating the proportion of different anomaly types, the percentages of warpage, thickness, and surface anomalies in the total area of ​​the steel strip are calculated separately. When the proportion of a certain type of anomaly exceeds a preset threshold, a corresponding quality warning mechanism is triggered. The ratio of the area of ​​the abnormal area to the area of ​​the normal area reflects the overall usability of the steel strip; the smaller the ratio, the better the quality of the steel strip, and the more readily it can be used for subsequent slitting processing.

[0032] In one embodiment, the determination of the comprehensive abnormal condition assessment result of the steel strip involves assigning different weight coefficients based on the severity of the abnormality type. Plastic deformation risk is given the highest weight because this type of abnormality can lead to permanent damage to the steel strip. Thickness abnormalities are the next most significant, affecting the dimensional accuracy of the slitting product. Surface abnormalities, while not affecting material strength, can impact the product's appearance quality. A comprehensive score is calculated using a weighted summation method, and the quality grade of the steel strip is determined based on the score range.

[0033] Step S102: Based on the comprehensive abnormal state assessment results, extract the local deformation distribution of the steel strip in the longitudinal and transverse directions from the warp height value and thickness distribution data. Combine the friction coefficient variation area identified by the surface finish difference to assess the degree of influence of residual bending stress on the centerline offset of the steel strip and determine the specific distance and location distribution of the centerline offset.

[0034] Based on the warpage height and thickness distribution data from the comprehensive anomaly assessment results, a differential calculation method is used to obtain the longitudinal and transverse deformation gradient sequences of the steel strip. A three-dimensional deformation distribution matrix is ​​constructed to record the local deformation at each spatial location. Combined with the spatial derivative of the warpage height, the longitudinal and transverse deformation distribution feature vectors are calculated to determine the local deformation distribution of the steel strip surface. For this local deformation distribution, the spatial variation law of roughness values ​​in the surface finish data is extracted. A friction coefficient distribution map is established through the physical correspondence between surface finish differences and friction coefficients, identifying the boundaries of regions with significant friction coefficient changes, and obtaining friction coefficient variation region distribution data containing both friction coefficient values ​​and spatial location information. Using the stress components in the residual bending stress numerical matrix, the contribution of bending deformation at each location is calculated through the constitutive relationship between stress and deformation in mechanics of materials. Combined with the theoretical coordinates of the steel strip's geometric centerline, the offset moments generated by residual bending stress on the centerline in the longitudinal and transverse directions are evaluated, resulting in a quantitative result of the influence of residual bending stress on the steel strip's centerline offset. Based on the quantification results of the degree of influence and the distribution data of the friction coefficient variation area, the centerline offset distance in different areas is solved by the force balance equation. The offset is mapped to the actual position coordinate system of the steel strip by the spatial coordinate transformation method, and the position distribution matrix of the centerline offset is constructed to determine the specific distance and position distribution of the centerline offset.

[0035] Specifically, the extraction process for local deformation distribution is based on the analysis of the geometric deformation characteristics of the steel strip surface. The longitudinal deformation gradient sequence is obtained by numerically differentiating the warp height value along the length of the steel strip, reflecting the rate of deformation change of the steel strip along the conveying direction. The transverse deformation gradient sequence is obtained by calculating the spatial derivative of the warp height in the width direction, characterizing the degree of non-uniform deformation of the steel strip in the transverse direction.

[0036] Specifically, the construction of the three-dimensional deformation distribution matrix uses an interpolation method to extend the discrete measurement point data into a continuous spatial distribution, and the matrix element values ​​represent the magnitude of the local deformation at the corresponding location.

[0037] It should be noted that thickness distribution data plays a crucial compensatory role in deformation analysis. When the steel strip thickness is unevenly distributed, the same external force will produce different degrees of deformation response in different thickness regions. By establishing a mapping relationship between thickness and deformation sensitivity, the systematic deviation caused by thickness differences in the warpage height measurement is corrected, thus obtaining the true local deformation distribution characteristics. The deformation distribution feature vector includes parameters such as longitudinal deformation amplitude, transverse deformation amplitude, and deformation direction angle, providing basic data support for subsequent centerline offset analysis. The identification of friction coefficient variation regions is achieved based on the physical correlation between surface finish and frictional properties.

[0038] In one possible implementation, a nonlinear relationship exists between surface roughness values ​​and friction coefficients, with regions of higher roughness typically corresponding to larger friction coefficients. By establishing empirical formulas or using lookup tables, the surface roughness values ​​are converted into corresponding friction coefficient estimates. The friction coefficient distribution map is represented in a two-dimensional grid format, where grid nodes correspond to spatial locations on the steel strip surface, and the node values ​​represent the magnitude of the friction coefficient at that location.

[0039] For example, boundary identification of regions with significant friction coefficient variations is achieved through a gradient threshold method. The spatial gradient at each location in the friction coefficient distribution map is calculated; when the gradient magnitude exceeds a preset threshold, the location is marked as a boundary point of friction coefficient variation. By connecting adjacent boundary points to form a closed region, the geometric contour of the friction coefficient variation region is obtained. The friction coefficient variation region distribution data not only includes the spatial location information of the region but also records the statistical characteristics of the friction coefficient within the region, such as the mean, standard deviation, and range of variation. Assessing the impact of residual bending stress on centerline offset involves a complex mechanical analysis process.

[0040] Preferably, the stress components in the residual bending stress numerical matrix include two components: normal stress and shear stress, corresponding to the internal stress states of the steel strip in the vertical and horizontal directions, respectively. A linear relationship between stress and strain is established using Hooke's law in mechanics of materials, and the local deformation caused by stress at each location is calculated. The bending deformation contribution represents the deformation response under unit stress, and its magnitude is related to the geometric parameters and material properties of the steel strip.

[0041] Specifically, the calculation of the centerline offset moment considers the influence of stress non-uniformity on the overall equilibrium state of the steel strip. When the stress level in a certain region of the steel strip is significantly higher than that in other regions, that region will experience greater bending deformation, causing the centerline of the steel strip to deviate from its theoretical position. The total offset moment is obtained by integrating the contributions of stress in different regions to the centerline position. The quantification of the degree of influence is expressed using a dimensionless scoring method; a higher score indicates a more significant impact of residual bending stress on the centerline offset.

[0042] In one embodiment, the determination of the centerline offset distance and position distribution comprehensively considers the coupling effects of multiple physical factors. The force balance equations are solved based on the principle that the sum of forces acting on the steel strip in all directions is zero under static equilibrium. By establishing a set of balance equations including bending stress, friction, and support force, the centerline offset at different positions is solved. The specific equation set is as follows:

[0043]

[0044] In the formula: F i M represents the force components of the steel strip in three directions. b F represents the torque generated by bending stress. f F represents frictional force. s The formula represents the supporting force, which describes the force balance principle that the sum of forces in all directions on the steel strip is zero in a static equilibrium state. The spatial coordinate transformation method described here converts the offset in the local coordinate system into the position coordinates in the global coordinate system, which is convenient for comparison and analysis with the actual geometric dimensions of the steel strip.

[0045] Understandably, the centerline offset position distribution matrix reflects the spatial distribution pattern of the offset phenomenon on the steel strip surface. The row index of the matrix corresponds to the position along the length direction of the steel strip, the column index corresponds to the position along the width direction, and the matrix element value represents the centerline offset distance at that position. By analyzing the numerical characteristics of the position distribution matrix, key areas with large offset distances and the dominant trend of offset direction can be identified.

[0046] Step S103: Based on the centerline offset distance and position distribution, and combined with the thickness distribution data to identify regional thickness differences, the deformation area in the width direction of the steel strip is divided and marked, and the stable section positions with warping height deviation values ​​less than a preset threshold and relatively uniform thickness are extracted.

[0047] Based on the variation of the centerline offset distance along the length of the steel strip, a fixed-size window is used to grid the width direction of the steel strip. The window moves according to a preset step size, and each grid records the offset value and thickness measurement value at the corresponding position. By comparing the thickness difference between adjacent grids, the thickness abrupt boundary is identified, and a region distribution matrix containing offset features and thickness gradients is obtained. For each grid cell in the region distribution matrix, the warp height value of the cell and its neighborhood is extracted, and the deviation of the warp height from a preset threshold is calculated. The coefficient of variation is obtained by calculating the ratio of the standard deviation to the average value of the thickness values ​​of adjacent grids. If the deviation value is less than the threshold and the coefficient of variation is lower than the preset standard, the grid is marked as a candidate stable region, resulting in a preliminary stable region marking map. Based on the preliminary stable region marking map, an eight-neighbor connectivity rule is used to identify adjacent candidate stable region grids. Connected grids are merged to form continuous stable blocks. The area ratio and aspect ratio of each block are calculated, and blocks with an area greater than the minimum effective area and an aspect ratio within a preset range are selected as stable segments. The starting and ending coordinates of the stable segments in the width direction of the steel strip are determined.

[0048] Specifically, in one implementation, the meshing process uses fixed-size windows to systematically scan the steel strip. The window size is determined based on the steel strip width and slitting accuracy requirements, typically set to one-tenth to one-twentieth of the steel strip width. The window moves along the steel strip width direction in steps of half the window size, achieving partial overlap between adjacent windows and avoiding the loss of boundary information. The data within each window is integrated into a grid cell, recording the average centerline offset and average thickness measurement of that area.

[0049] Specifically, during the construction of the region distribution matrix, the thickness difference between adjacent grids is obtained by calculating the absolute value of the difference between the mean thicknesses of the two grids. When the difference exceeds a preset thickness abrupt change threshold, a thickness abrupt change boundary is marked between the two grids. The region distribution matrix is ​​stored in the form of a two-dimensional array, with the row index corresponding to the position number in the length direction of the steel strip and the column index corresponding to the grid number in the width direction. The matrix elements contain two components: offset eigenvalue and thickness gradient value.

[0050] It should be noted that the coefficient of variation is calculated using statistical methods to assess stability. For each grid cell and its eight neighboring grid cells, the standard deviation of these nine values ​​is calculated, and then divided by the average value to obtain the coefficient of variation, which reflects the dispersion of the thickness distribution. The smaller the coefficient of variation, the more uniform the thickness distribution in that region, and the higher the stability.

[0051] Preferably, the eight-neighbor connectivity rule, when identifying continuous stable blocks, checks the top, bottom, left, right, and four diagonal adjacent grids of each candidate stable region grid. If adjacent grids are also marked as candidate stable regions, they are merged into the same connected block. The connected blocks are gradually expanded through recursive traversal until no more adjacent candidate stable region grids can be found. The aspect ratio is obtained by calculating the span ratio of the block in the length and width directions of the steel strip. When the ratio is between 0.5 and 2, the block is considered to have a regular shape and is suitable as a stable segment. This screening method excludes overly long or irregular regions, ensuring that the selected stable segments have good processing adaptability.

[0052] Step S104: Determine the position of the dynamic baseline that adapts to the real-time morphological changes of the steel strip based on the stable section position data.

[0053] Based on the stable section location data, the center coordinates and boundary range of each section are extracted. Combined with the stress values ​​at the corresponding positions in the bending stress residual distribution matrix, the theoretical offset distance caused by stress at each point within the section is calculated through linear interpolation. The roughness distribution of the corresponding section in the surface finish detection data is obtained, and a mapping relationship between roughness and frictional resistance is established. The contribution value of frictional resistance to the lateral offset during steel strip conveying is calculated, resulting in an initial correction vector containing stress offset components and frictional offset components. For each component value in the initial correction vector, the anomaly level identifier of the corresponding section in the comprehensive anomaly state assessment result is extracted. If the anomaly level exceeds a preset safety threshold, the correction vector of that section is attenuated. The attenuation coefficient is determined according to the ratio of the anomaly level to the maximum level. By comparing the deviation between the predicted offset value and the actual measured value before and after correction, the effectiveness of the correction vector is judged, and a verified offset correction value sequence is obtained. Based on the numerical sequence of offset corrections, a weighted average method is used to fuse the corrections for each stable segment. The weights are determined based on the segment area ratio and stability score. Discrete correction points are connected using cubic spline interpolation to form a continuous baseline correction curve. Real-time conveying speed and tension change data of the steel strip are obtained, and the morphological change rate is calculated. The update frequency of the baseline correction curve is adjusted according to the morphological change rate. When the change rate exceeds a preset threshold, the update frequency is increased. Kalman filtering is used to fuse historical baseline positions and current corrections to predict the position at the next moment. The state vector of the Kalman filter includes the baseline position and velocity, and the observation vector is the current correction. Combined with the deformation distribution characteristics in the width direction of the steel strip, a dynamic baseline position coordinate sequence adapted to the real-time morphological changes of the steel strip is determined.

[0054] Specifically, in one implementation, the initial correction vector is constructed through multi-source data fusion. The center coordinates of the stable section are determined by the arithmetic mean of the coordinates of all grids within the section, and the boundary range is obtained by identifying the coordinates of the outermost grid of the section. The residual bending stress distribution matrix records the magnitude and direction of the residual stress at each location point of the steel strip, and the stress value between any two known stress points is calculated using a linear interpolation method. During the interpolation process, the stress value of the interpolation point is determined according to the distance ratio between the point to be interpolated and the two known points, following a linear relationship. The theoretical offset distance is calculated by multiplying the ratio of the stress value to the material's elastic modulus by the steel strip thickness, reflecting the lateral displacement trend of the steel strip under stress.

[0055] Specifically, the mapping relationship between surface roughness and frictional resistance is established based on empirical formulas and fitting of measured data. Surface roughness is measured using a profilometer to obtain the Ra value, and frictional resistance is expressed as the product of the friction coefficient and the normal force. When the surface roughness of the steel strip increases, the friction coefficient between it and the conveyor roller increases non-linearly; a functional relationship between the two is established through polynomial fitting. The contribution of frictional resistance to lateral offset considers the movement speed and tension distribution of the steel strip; the faster the speed, the more significant the influence of the lateral component of friction. The initial correction vector adds the stress offset component and the friction offset component according to preset weighting coefficients, which are determined based on the statistical regularity of historical processing data.

[0056] It should be noted that the attenuation mechanism is designed to consider the impact of abnormal regions on correction accuracy. The comprehensive anomaly assessment results categorize the anomaly level into five levels, increasing from minor to severe. The attenuation coefficient is calculated using a linear relationship: when the anomaly level is the lowest, the attenuation coefficient is 1, indicating no attenuation; when the anomaly level reaches the highest, the attenuation coefficient drops to 0.2, significantly reducing the weight of the correction in that section. This attenuation strategy avoids interference from erroneous corrections in abnormal regions on the overall baseline adjustment, improving the reliability of the correction. The validity of the correction vector is verified through real-time comparison, comparing the corrected predicted offset value with the actual offset value measured by the laser displacement sensor; if the deviation is within the allowable range, the correction is considered valid.

[0057] For example, the weighted average method fully considers the representativeness of each stable segment when integrating the corrections. The segment area percentage is calculated by dividing the area of ​​that segment by the total area of ​​all stable segments; segments with larger areas play a more significant role in the overall correction. The stability score is a comprehensive evaluation based on the thickness uniformity, warpage, and surface quality within the segment; a higher score indicates more reliable measurement data for that segment. The weighting formula is the product of the area percentage and the stability score, then normalized to ensure that the sum of all weights is 1.

[0058] Preferably, cubic spline interpolation provides a smooth transition when forming a continuous baseline correction curve. Discrete correction points are the correction values ​​at the center of each stable segment, and these points are irregularly distributed along the width of the steel strip. The cubic spline function employs a cubic polynomial within each interval, ensuring the continuity of the function values, first derivative, and second derivative at the nodes. The spline coefficients are obtained by solving a tridiagonal linear system of equations, constructing a smooth curve passing through all correction points. This interpolation method avoids the broken line characteristics of linear interpolation and the oscillations of high-order polynomial interpolation.

[0059] In one possible implementation, the calculation of the morphological change rate comprehensively considers multiple dynamic indicators. The real-time conveying speed of the steel strip is obtained through encoder feedback, and tension change data is provided by a tension sensor array. The morphological change rate is defined as the amount of change in the shape parameters of the steel strip per unit time, including the centerline offset rate, thickness change rate, and warping change rate. When the morphological change rate exceeds a preset threshold, it indicates that the steel strip is in a state of rapid deformation, and the baseline update frequency needs to be increased to adapt to dynamic changes.

[0060] Understandably, Kalman filtering achieves optimized fusion of historical information and current observations when predicting the baseline position. The state vector contains the position coordinates and moving speed of the baseline at each point along the width of the steel strip, and the state transition matrix is ​​determined based on the physical model of steel strip transport. The observation vector is the correction value obtained by the sensor at the current moment, and the observation noise covariance matrix is ​​set according to the sensor accuracy. Kalman filtering estimates the state at the next moment through a prediction step, and then corrects the prediction result by combining the actual observation values ​​through an update step. The filter gain is adaptively adjusted according to the ratio of the prediction error covariance and the observation noise covariance, achieving accurate tracking and prediction of the baseline position. This method effectively suppresses the influence of measurement noise and provides a smooth and accurate baseline trajectory.

[0061] For example, the dynamic adjustment strategy for the update frequency is implemented in stages based on the rate of morphological change. When the rate of change is within the normal range, the baseline is updated every 100 milliseconds; when the rate of change exceeds a first threshold, the update interval is shortened to 50 milliseconds; and when a second threshold is reached, it is further shortened to 20 milliseconds. This adaptive adjustment mechanism avoids unnecessary computational burden while ensuring tracking accuracy.

[0062] In one embodiment, the final determination of the dynamic baseline position coordinate sequence also considers the deformation distribution characteristics along the width of the steel strip. By analyzing the symmetry, continuity, and periodicity of the deformation distribution, potential systematic deviations are identified. When a periodic deformation pattern is detected, a corresponding compensation amount is added to the baseline adjustment to correct the expected offset in advance. The coordinate sequence records the position information of the baseline across the entire width of the steel strip at fixed sampling intervals.

[0063] Step S105: Based on the position of the dynamic baseline and the thickness distribution data, generate the lateral movement distance control command and the corresponding feed depth preset value of the slitting tool, and obtain the difference value between the control command and the current actual position of the tool.

[0064] The position values ​​of each sampling point are extracted based on the dynamic baseline position coordinate sequence. The boundary of the region with a thickness deviation exceeding a preset threshold is identified from the thickness distribution data. The vertical distance from the boundary of this region to the baseline is calculated. The lateral displacement of the tool from its current position to the target cutting position is obtained through translation transformation in the Cartesian coordinate system, yielding the initial lateral movement distance. For this initial lateral movement distance, a feed depth compensation coefficient is determined based on the degree of deviation in the thickness deviation region. The compensation coefficient is calculated by dividing the deviation value by the standard thickness and then multiplying by a preset scaling factor. If the thickness deviation is positive, the feed depth is decreased; if the deviation is negative, the feed depth is increased. The preset actual feed depth is calculated by multiplying the standard feed depth by the compensation coefficient, forming a control command data package containing the lateral movement distance and feed depth. The current actual position coordinates and feed depth status of the slitting tool are obtained. The target position parameters are extracted from the control command data package, and the lateral position difference and depth difference are calculated. The total tool movement distance is determined by the square root of the sum of the squares of the differences. The movement direction angle is determined based on the ratio of the lateral difference to the depth difference, resulting in the difference between the control command and the current actual tool position.

[0065] Specifically, in one implementation, a dynamic baseline position coordinate sequence records the positions of reference points along the width direction of the steel strip at fixed spatial intervals. Each sampling point contains two components: a horizontal coordinate and a vertical coordinate, with the coordinate values ​​represented by their absolute position relative to the equipment origin. Thickness deviation regions are identified by comparing the measured thickness values ​​with the standard thickness point by point. When the deviation values ​​at multiple consecutive measurement points exceed a preset threshold, the boundary positions of these points are marked as the boundaries of the deviation region.

[0066] Specifically, translation transformation in Cartesian coordinates is achieved by establishing a mapping relationship between the tool coordinate system and the workpiece coordinate system. The vertical distance from the region boundary to the baseline is calculated using the point-to-line distance formula, and the baseline is obtained by fitting adjacent baseline points using the least squares method. The lateral displacement of the tool from its current position to the target cutting position is equal to the x-coordinate of the target position minus the x-coordinate of the current position; this displacement is the initial lateral movement distance. This direct coordinate transformation method avoids complex rotation matrix calculations and improves the real-time performance of control commands.

[0067] It should be noted that the feed depth compensation coefficient is calculated by fully considering the impact of thickness deviation on the cutting depth. The deviation value is obtained by subtracting the standard thickness from the measured thickness. The compensation coefficient is equal to the deviation value divided by the standard thickness and then multiplied by a preset scaling factor. The scaling factor is determined based on the material hardness and tool characteristics, and typically ranges from 0.5 to 1.5. When the local thickness of the steel strip is larger, the compensation coefficient is positive, and the actual feed depth increases accordingly; conversely, the feed depth decreases. The control command data packet is stored in a structured format, containing four fields: target lateral position, feed depth value, execution timestamp, and priority identifier.

[0068] Preferably, the difference numerical vector is calculated using the Euclidean distance metric. The lateral position difference and depth difference represent the displacement requirements of the tool in two dimensions, respectively, and the total distance is obtained by taking the square root of the sum of the squares of the two differences. The movement direction angle is calculated using the arctangent function, with the input being the ratio of the depth difference to the lateral difference, and the output angle value used for servo motor direction control. The difference numerical vector is represented in polar coordinates, containing two components: the distance magnitude and the direction angle, facilitating direct execution by the motion controller.

[0069] For example, when a thickness deviation of 0.2 mm is detected in a certain area of ​​the steel strip, the cutting parameters for that area are automatically adjusted.

[0070] Step S106: When the difference value does not meet the accuracy requirements, the control command is iteratively updated based on the reliability assessment results of the stable section position. Combined with the degree of influence of residual bending stress on the offset of the steel strip centerline, the feed depth control command and the optimized lateral movement control command of the steel strip slitting equipment are determined.

[0071] The lateral deviation and depth deviation components in the difference value vector are obtained and compared with preset lateral accuracy thresholds and depth accuracy thresholds. The ratio of the deviation component to the threshold is calculated. If either ratio exceeds a preset tolerance coefficient, the accuracy requirement is not met. The reliability score is extracted from the stable section position data. Each section is assigned an iterative weight according to the score. The comprehensive reliability index is calculated using a weighted summation method to obtain the reliability assessment result used for control command adjustment. Based on the reliability assessment result, sections with reliability indices higher than the preset standard are extracted as reference benchmarks. The residual bending stress values ​​corresponding to these sections are read. The theoretical offset value of each section is calculated by multiplying the stress value by the material deformation coefficient. The difference between the theoretical offset value and the measured offset value is compared to determine the stress influence correction coefficient. The lateral movement distance parameter is gradually adjusted using an iterative method. The parameter is adjusted proportionally according to the difference value in each iteration until the difference value converges to the accuracy requirement range, resulting in the updated lateral movement distance value. Based on the updated lateral movement distance value and the spatial distribution characteristics of the steel strip centerline offset, the offset data of non-sampling points is completed by linear interpolation. The actual offset of each cutting position is calculated, and the original feed depth and offset compensation are superimposed to obtain the corrected feed depth value, forming a feed depth control command sequence containing position coordinates and depth parameters. Using the feed depth control command sequence and the iteratively optimized lateral movement parameters, the motion trajectory points of the servo motor are generated through coordinate transformation. The time interval of the trajectory points is adjusted according to the equipment response characteristics, and the result is encoded into a control command format recognizable by the slitting equipment. This determines the feed depth control command and the optimized lateral movement control command for the steel strip slitting equipment.

[0072] Specifically, in one implementation, the multi-dimensional accuracy judgment of the difference values ​​is achieved by establishing layered accuracy standards. The lateral accuracy threshold is determined based on the steel strip width and the number of strips, typically set within the range of one-thousandth to five-thousandths of the steel strip width. The depth accuracy threshold is set based on the steel strip thickness and material properties; for steel strips with a thickness of 0.5 to 2 mm, the depth accuracy threshold is generally controlled between 0.01 and 0.05 mm. The ratio of the deviation component to the threshold is calculated using absolute values ​​to avoid the mutual cancellation of positive and negative deviations. A tolerance coefficient serves as the judgment boundary; when the ratio exceeds 1.1, it is determined that the accuracy requirements are not met, triggering an iterative optimization process.

[0073] Specifically, obtaining reliability assessment results involves comprehensive consideration of multiple evaluation dimensions. The reliability score of the stable section is based on three main indicators: uniformity of thickness distribution within the section, stability of warpage, and consistency of surface quality. Uniformity is measured by calculating the ratio of the thickness standard deviation to the mean, stability is characterized by the coefficient of variation of warpage height, and consistency is assessed by the range of surface roughness. The three indicators are assigned weight coefficients of 0.4, 0.3, and 0.3, respectively, and the weighted sum yields a reliability score ranging from 0 to 100. The iterative weights are positively correlated with the reliability score; sections with higher scores have greater reference value in control command adjustments. The comprehensive reliability index is obtained by weighted averaging across all sections, reflecting the overall reliability of the measurement data.

[0074] It should be noted that the influence of residual bending stress on centerline offset was quantitatively analyzed using a materials mechanics model. The theoretical offset value was obtained by multiplying the stress value by the material deformation coefficient, which is determined by the elastic modulus, Poisson's ratio, and thickness of the steel strip. For ordinary carbon steel, the elastic modulus is approximately 200 GPa, and the Poisson's ratio is between 0.25 and 0.3. The deformation coefficient was calculated using an empirical formula: the deformation coefficient equals the thickness divided by the elastic modulus and then multiplied by the stress concentration factor. The stress concentration factor was determined based on the edge condition and cutting method of the steel strip; a value of 1.0 was used for smooth edges, and 1.2 to 1.5 was used for areas with minor defects. The comparison between the theoretical and measured offset values ​​revealed the accuracy of the model predictions, and the difference between the two reflected the influence of complex factors not captured by the model.

[0075] For example, the iterative update process employs a gradual adjustment strategy to ensure convergence stability. The initial lateral movement distance parameter is set to a theoretical value based on geometric calculations. Each iteration determines the adjustment step size based on the magnitude of the difference. When the difference is greater than twice the accuracy requirement, the adjustment step size is set to 0.5 times the difference; when the difference is between 1 and 2 times the accuracy requirement, the adjustment step size is set to 0.3 times the difference; and when it approaches the accuracy requirement, the adjustment step size is reduced to 0.1 times the difference. This variable step size strategy ensures rapid convergence while avoiding oscillations near the optimal solution. The iterative process continues until the difference value for three consecutive iterations is less than the accuracy requirement, at which point convergence is considered achieved. During the iteration process, the parameter value and corresponding difference value are recorded for each adjustment, forming an optimization trajectory, which facilitates the analysis of convergence characteristics and the effectiveness of the adjustment strategy.

[0076] Preferably, linear interpolation provides a balance between computational efficiency and accuracy when completing the offset data for non-sampling points. Sampling points are distributed at fixed intervals along the width of the steel strip, with the interval determined based on the strip width and required accuracy. For a 1000 mm wide steel strip, the sampling interval is typically set to 10 to 20 mm. The offset value of non-sampling points is calculated through a linear combination of two adjacent sampling points, with the weighting coefficient determined based on the distance ratio.

[0077] In one possible implementation, the feed depth control command sequence is formed considering the continuity requirements of the cutting process. Each control command includes four elements: a timestamp, position coordinates, feed depth value, and execution status flag. The position coordinates are represented using absolute coordinates relative to the equipment origin, and the feed depth value is the corrected actual cutting depth. The command sequence is arranged in chronological order, and the time interval between adjacent commands is determined based on the equipment's response speed and the strip conveyor speed.

[0078] Understandably, the coordinate transformation process maps logical coordinates to physical motion parameters. The motion trajectory points of the servo motor are obtained by converting Cartesian coordinates to motor step counts, with the transformation coefficient determined by the mechanical transmission ratio and motor resolution. For a motor with a step angle of 1.8 degrees and a transmission mechanism with a 10:1 reduction ratio, each millimeter corresponds to 200 step pulses. The time interval between trajectory points is dynamically adjusted according to the device's response characteristics, appropriately increasing the interval during high-speed motion and decreasing it during precise positioning.

[0079] For example, the control command format uses a standardized data structure to facilitate equipment parsing and execution. The command header contains a command type identifier and length information, the body contains specific control parameters, and a checksum is appended at the end to ensure transmission reliability. Lateral movement control commands and feed depth control commands are encoded separately and distinguished by the command type identifier. This structured command format improves the reliability and maintainability of equipment control and enables precise slitting control.

[0080] In one embodiment, the entire control optimization process is dynamically adjusted through a closed-loop feedback mechanism. The deviation between the actual and commanded positions of the tool is monitored in real time, and a compensation mechanism is automatically triggered when the deviation exceeds the allowable range. The compensation amount is calculated based on historical deviation data and current working conditions to ensure that the slitting accuracy remains within the required range.

[0081] Step S107: The feed depth control command and the optimized lateral movement control command of the steel strip slitting equipment are used to schedule the actuator of the steel strip slitting equipment. The matching degree between the steel strip slitting trajectory record after execution and the preset slitting accuracy standard is evaluated. If the matching degree is insufficient, the abnormal state evaluation step is returned based on the current multi-sensor data for iterative optimization processing.

[0082] The servo motor and hydraulic actuator of the steel strip slitting equipment are driven by feed depth control commands and lateral movement control commands. The motor's stepping motion is controlled by pulse signals, and actions are executed sequentially according to the timestamps in the commands. Real-time feedback signals from the actuators, including position encoder readings and pressure sensor values, are acquired. The thickness value at the current machining position is read from the thickness distribution data, and the roughness level of the corresponding area in the surface finish detection results is extracted to obtain a set of actual tool operating state parameters. Based on this set of actual tool operating state parameters, the thickness deviation value at the current cutting position is calculated. If the deviation exceeds a preset threshold, the tool feed speed is adjusted. The contact pressure between the tool and the steel strip is corrected according to the surface roughness level. A smooth motion trajectory is generated between adjacent control points using a linear interpolation method. The tool's lateral movement speed and longitudinal feed are updated, and the adjusted tool trajectory coordinate sequence is recorded to obtain real-time optimized tool trajectory data. Based on the real-time optimized tool trajectory data, the edge coordinates of the steel strip after slitting are collected, the width and straightness parameters of the slitting gap are measured, and the positional and shape tolerance requirements in the preset slitting accuracy standard are compared. The deviation between the actual slitting trajectory and the standard trajectory is calculated, and the overall slitting quality pass rate is evaluated by calculating the mean and standard deviation, thus obtaining the slitting accuracy matching degree evaluation value. It is then determined whether the slitting accuracy matching degree evaluation value meets the quality requirement threshold. If the evaluation value is lower than the threshold, real-time data from the tension sensor, displacement sensor, and thickness measuring instrument are extracted, and multi-sensor data are integrated to form a new input parameter set. This triggers an abnormal state re-evaluation process, updates the control parameters, and regenerates the optimized control command, achieving iterative optimization until the matching degree meets the standard.

[0083] Specifically, in one implementation, the process of controlling the actuators by dispatching control commands is achieved through multi-channel parallel control. Feed depth control commands and lateral movement control commands are transmitted to their respective actuators via independent communication channels. The servo motor receives a sequence of pulse signals, each pulse corresponding to a tiny angular displacement; the pulse frequency determines the motor speed, and the number of pulses determines the total rotation angle. The hydraulic actuator controls the oil pressure through a proportional valve to achieve precise adjustment of the feed depth. The actions of the two types of actuators are synchronized and coordinated according to timestamps, with timestamp accuracy reaching the millisecond level, ensuring coordinated operation of lateral movement and longitudinal feed.

[0084] Specifically, the real-time feedback mechanism of the actuator forms the foundation for closed-loop control. The position encoder uses an incremental photoelectric encoder with a resolution of 10,000 pulses per revolution, monitoring the actual rotation angle of the servo motor in real time. Pressure sensors are installed at the inlet and outlet of the hydraulic cylinder, measuring from 0 to 25 MPa with a sampling frequency of 1 kHz, capturing dynamic pressure changes in the hydraulic system. These feedback signals are transmitted to the controller via a high-speed data bus, with a transmission delay controlled within 10 milliseconds. The controller compares the feedback values ​​with the command values ​​in real time, generating a deviation signal to adjust the control output. The set of actual tool operating state parameters includes multi-dimensional information such as position, speed, acceleration, pressure, and temperature, providing comprehensive state data for subsequent trajectory optimization.

[0085] It should be noted that the real-time adjustment mechanism based on thickness distribution and surface finish embodies the concept of adaptive control. The thickness deviation value is obtained by subtracting the standard thickness from the measured thickness at the current position. A positive deviation indicates that the steel strip is locally too thick, requiring a reduction in feed rate to prevent tool overload; a negative deviation indicates that the steel strip is too thin, allowing for an appropriate increase in feed rate to improve machining efficiency. The adjustment amount of feed rate is proportional to the deviation value, with the proportionality coefficient determined based on the hardness of the steel strip material. Higher hardness results in a smaller coefficient, preventing excessively rapid speed changes that could lead to cutting instability. Surface roughness is divided into five levels, increasing from smooth to rough, with each level corresponding to a different contact pressure correction coefficient. Higher roughness results in greater friction, requiring increased contact pressure to ensure the cutting depth; conversely, lower roughness requires reduced pressure to prevent tool wear.

[0086] For example, linear interpolation plays a crucial role in generating smooth motion trajectories. Trajectories between adjacent control points are connected by a linear function, and the coordinates of the interpolated points are uniformly distributed over time. The interpolation period is set according to the system's control frequency, typically between 1 and 10 milliseconds. Within each interpolation period, the target position corresponding to the current moment is calculated; the target position equals the starting position plus the displacement increment multiplied by the time ratio. The displacement increment is the endpoint position minus the starting position, and the time ratio is the difference between the current moment and the starting moment divided by the total time. This interpolation method is computationally simple, has good real-time performance, and can meet the requirements of high-speed slitting processing. The trajectory points generated by interpolation are managed through a buffer to ensure continuous output of control commands.

[0087] Preferably, the evaluation of the matching degree of slitting accuracy adopts a multi-index comprehensive evaluation method. The slitting gap width is measured by a laser rangefinder sensor array, with sensors arranged along the width direction of the steel strip at 50 mm intervals, achieving a measurement accuracy of 0.01 mm. The straightness parameter is obtained by fitting the slitting edge points using the least squares method, and the maximum deviation distance between the fitted straight line and the actual edge point is calculated. The positional tolerance requires that the deviation between the slitting edge and the theoretical position does not exceed ±0.5 mm, and the shape tolerance requires that the straightness deviation does not exceed 0.2 mm per meter. The deviation between the actual trajectory and the standard trajectory is obtained by calculating the Euclidean distance between corresponding points, and the root mean square value of the deviation of all points is taken as the overall deviation index.

[0088] In one possible implementation, the statistical evaluation of the pass rate reflects the quality stability of batch processing. Calculated using the mean and standard deviation, the mean reflects the average level of slitting accuracy, while the standard deviation reflects the dispersion of accuracy. The pass rate is defined as the percentage of slitting lengths meeting accuracy requirements out of the total length; a pass rate below 95% is considered substandard. The slitting accuracy matching degree evaluation value comprehensively considers deviation, pass rate, and stability, calculated using a weighted summation method with weights of 0.5, 0.3, and 0.2, respectively. When the evaluation value falls below the quality requirement threshold, it automatically enters optimization mode. Real-time extraction of multi-sensor data includes the steel strip tension distribution measured by tension sensors, the steel strip lateral offset monitored by displacement sensors, and real-time thickness data acquired by a thickness gauge. These data are filtered and fused to form a new set of input parameters. Abnormal state re-evaluation is performed based on the updated parameter set to identify the main factors causing insufficient accuracy.

[0089] For example, the control parameters are updated using a gradient optimization method. The direction and step size of parameter adjustments are determined based on the direction and magnitude of the accuracy deviation. Parameters such as lateral traverse speed, feed depth, and contact pressure are adjusted according to their respective sensitivity coefficients. The optimized control commands are then reissued to the actuator, initiating a new machining cycle. This iterative optimization continues, with the degree of accuracy improvement evaluated after each cycle. When three consecutive evaluations meet the requirements, the optimization is considered converged, and the process exits optimization mode and returns to normal machining.

[0090] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for generating and scheduling control instructions for a striping equipment, characterized in that, include: By integrating tension values, warpage height values, thickness distribution data, and surface finish data during the steel strip conveying process, a comprehensive abnormal state assessment result for the steel strip is generated. Based on the comprehensive abnormal state assessment results, the longitudinal and transverse local deformation distributions in the warping height value and the thickness distribution data are extracted. Combined with the friction coefficient variation area determined by the surface finish data, the distance and position distribution of the steel strip centerline offset are determined. Based on the distance and position distribution of the centerline offset, and combined with the regional thickness differences in the thickness distribution data, the deformation areas in the width direction of the steel strip are marked by partitioning, and the warping height deviation value and the location of stable sections where the thickness uniformity meets the preset conditions are extracted. The position of the dynamic baseline is determined based on the distance and position distribution of the centerline offset of the stable segment. Based on the position of the dynamic baseline and the thickness deviation area in the thickness distribution data, a control command for the lateral movement distance of the slitting tool and a preset value for the feed depth are generated. Based on the location of the stable section and the distance of the centerline offset, the lateral movement distance control command and the feed depth preset value are iteratively updated; Using the lateral movement distance control command and the feed depth preset value, the slitting equipment actuator is scheduled to adjust the tool running trajectory.

2. The method for generating and scheduling control instructions for slitting equipment according to claim 1, characterized in that, The integrated data on tension, warpage height, thickness distribution, and surface finish during the steel strip conveying process are used to generate a comprehensive abnormal state assessment result for the steel strip, including: Tension sensor readings at multiple locations during steel strip conveying are acquired, and the corresponding warpage height measurements are recorded. A thickness gradient map of the steel strip's transverse thickness distribution is generated using ultrasonic scanning, and a surface roughness distribution map is generated by detecting the steel strip's surface roughness. Based on the yield strength and elastic modulus of the steel strip material, a residual stress distribution matrix is ​​calculated. For regions in the residual stress distribution matrix that exceed the yield strength, the warpage height deviation and thickness variation rate at the corresponding locations are extracted to determine the mapping relationship between stress concentration and deformation degree. Based on this mapping relationship, sections with abnormalities in continuous sampling points are identified, and the standard deviation of the warpage height and the coefficient of variation of the thickness distribution within these sections are calculated. An abnormal region is grouped using a clustering algorithm to generate a section anomaly distribution map containing the anomaly type and its impact range. Based on the section anomaly distribution map, the proportion of the anomaly type in the width and length directions of the steel strip is statistically analyzed, and the area ratio of the abnormal region to the normal region is calculated to generate a comprehensive anomaly status assessment result for the steel strip.

3. The method for generating and scheduling control instructions for slitting equipment according to claim 1, characterized in that, Based on the comprehensive abnormal state assessment results, the longitudinal and transverse local deformation distributions in the warping height value and the thickness distribution data are extracted. Combined with the friction coefficient variation areas determined by the surface finish data, the distance and position distribution of the steel strip centerline offset are determined, including: Based on the comprehensive abnormal state assessment results, longitudinal and transverse deformation distribution data are extracted from the warping height values ​​to generate a deformation distribution matrix. For the deformation distribution matrix, combined with the thickness distribution data, thickness abrupt change boundaries in the steel strip width direction are identified. Based on the surface finish data, friction coefficient variation regions are extracted, and the contribution of friction coefficient variation to centerline offset is calculated. Based on the deformation distribution matrix and the friction coefficient variation regions, a linear regression method is used to calculate the centerline offset distance. Based on the centerline offset distance and the thickness abrupt change boundaries, a centerline offset position distribution map is generated, and the offset position coordinates are recorded.

4. The method for generating and scheduling control instructions for slitting equipment according to claim 1, characterized in that, The step of dividing and marking the deformation areas in the width direction of the steel strip according to the distance and position distribution of the centerline offset, combined with the regional thickness differences in the thickness distribution data, and extracting the warping height deviation value and the location of stable sections where the thickness uniformity meets preset conditions includes: Based on the distance of the centerline offset, the steel strip width direction is divided into regions using a gridding method to generate a region distribution matrix; for each grid in the region distribution matrix, the degree of deviation between the warping height value and a preset threshold is calculated; based on the thickness distribution data, the thickness variation coefficient of adjacent grids is calculated; based on the degree of deviation and the variation coefficient, grids that meet the preset conditions are marked as stable segments.

5. The method for generating and scheduling control instructions for slitting equipment according to claim 1, characterized in that, The determination of the dynamic baseline position based on the distance and position distribution of the centerline offset of the stable segment position includes: Extract the center coordinates of the stable section location, and calculate the theoretical offset value by combining the distance of the centerline offset; extract the roughness distribution based on the surface finish data, calculate the contribution value of frictional resistance to the offset, and generate an initial correction vector; adjust the initial correction vector based on the comprehensive abnormal state evaluation results to generate an offset correction amount sequence; based on the offset correction amount sequence, use an interpolation method to generate a baseline correction curve and determine the dynamic baseline position coordinate sequence.

6. The method for generating and scheduling control instructions for slitting equipment according to claim 1, characterized in that, The step of generating a lateral movement distance control command and a preset feed depth value for the slitting tool based on the position of the dynamic baseline and the thickness deviation area in the thickness distribution data includes: Extract the dynamic baseline position coordinate sequence and identify the thickness deviation region boundary in the thickness distribution data; calculate the distance from the thickness deviation region boundary to the dynamic baseline position and generate a lateral movement distance value; calculate the feed depth compensation coefficient according to the degree of deviation of the thickness deviation region and generate a preset feed depth value.

7. The method for generating and scheduling control instructions for slitting equipment according to claim 1, characterized in that, The iterative update of the lateral movement distance control command and the feed depth preset value based on the stable segment position and the centerline offset distance includes: Extract the reliability score of the stable section position to generate a comprehensive reliability index; based on the comprehensive reliability index, adjust the centerline offset distance to generate an updated lateral movement distance value; calculate the offset compensation amount according to the updated lateral movement distance value to generate a corrected feed depth control command sequence; use the feed depth control command sequence to generate a control command format recognizable by the slitting equipment.

8. The method for generating and scheduling control instructions for slitting equipment according to claim 1, characterized in that, The step of using the lateral movement distance control command and the feed depth preset value to schedule the slitting equipment actuator and adjust the tool trajectory includes: Using the lateral movement distance control command and the preset feed depth value, the slitting equipment actuator is driven, and real-time feedback signals from the actuator are obtained. Based on the thickness distribution data and the surface finish data, the tool feed speed and contact pressure are adjusted to generate real-time optimized tool trajectory data. Based on the real-time optimized tool trajectory data, the coordinates of the slitting edge position are collected, the width and straightness parameters of the slitting gap are calculated, and a slitting accuracy matching degree evaluation value is generated. Based on the slitting accuracy matching degree evaluation value, multi-sensor data is integrated to trigger an abnormal state re-evaluation process and update the control command.

Images