A PVC buckle plate warping degree identification method based on intelligent sensor

Abstract

The application relates to the technical field of intelligent detection, and discloses a PVC buckle plate warping degree recognition method based on an intelligent sensor, which comprises the following steps: a PVC buckle plate detection coordinate system is established, axial and transverse sampling positions are arranged, ranging values and contact pressure values in the conditions of no load and plate placement are collected, and actual measured height and contact strength are obtained; based on a low contact area, an overall posture plane is fitted, the influence of placement inclination is deducted, and an actual measured warping residual error is obtained; further, in combination with neighborhood support constraint degree and free warping trend, a region affected by support flattening is compensated, and a low constraint equivalent warping residual error is obtained; finally, overall warping, edge warping, distortion and support flattening underestimation are calculated, and warping grades and support flattening risk prompts are output according to the calculation results.

Description

Technical Field

[0001] This invention relates to the field of intelligent detection technology, specifically to a method for identifying the degree of warping of PVC ceiling panels based on intelligent sensors. Background Technology

[0002] PVC ceiling panels, commonly used in interior decoration and architectural ceilings, have their flatness and warping directly affecting installation fit, joint consistency, and subsequent stability. During production, warehousing, and factory inspection, it is typically necessary to identify deformations in PVC panels, such as overall bending, edge warping, or diagonal twisting. Current testing methods often rely on visual inspection, ruler measurement, feeler gauge measurement, or single-point height testing. However, these methods are easily influenced by the inspector's experience, the panel's placement, the selection of measurement points, and the flatness of the supporting surface, making it difficult to obtain stable and repeatable quantitative results.

[0003] With the development of sensor detection technology, detection methods have emerged that use height sensors to sample the surface of panels and calculate height deviations. However, these methods typically only focus on the height distribution on the panel surface, and the height value under the default sampling state can directly reflect the free warping state of the panel. In reality, when PVC panels are placed on a supporting surface for testing, some protruding or warped areas may be flattened due to their own weight, local contact, or support pressure. This causes the actual height residual collected to be less than the true warping amplitude of the panel under low constraint, especially for edge warping and local twisting areas, which easily leads to an underestimation of the degree of warping. Furthermore, judging the degree of warping solely based on the overall maximum height difference makes it difficult to distinguish between different deformation forms such as overall bending, edge warping, and diagonal twisting, resulting in detection results that do not fully reflect the actual installation risks of the panel. Therefore, there is an urgent need for a detection method that can combine height sampling and contact pressure information to identify and compensate for the impact of support flattening and comprehensively evaluate the degree of warping of PVC panels. Summary of the Invention

[0004] This invention provides a method for identifying the degree of warping of PVC panels based on intelligent sensors, which helps to solve the problems mentioned in the background art.

[0005] This invention provides the following technical solution: a method for identifying the degree of warping of PVC ceiling panels based on intelligent sensors, comprising: The number of axial and lateral sampling positions is determined based on the detection length and the maximum allowable spacing, and the coordinates of each sampling position are evenly divided. Collect height and pressure values ​​at each sampling location under no-load and working conditions, calculate the measured height value, and determine the contact strength coefficient at each sampling location based on the no-load pressure reference threshold and the contact reference pressure value. The attitude fitting weights are calculated from the contact strength coefficient, and the overall attitude plane is fitted by weighted least squares. The difference between the measured height value and the overall attitude plane height value is used as the measured warping residual under the supported state. The support influence base radius is determined based on the actual sampling interval, and a neighborhood set is constructed. The support constraint coverage coefficient at each sampling location is obtained using the contact strength coefficient within the neighborhood. The free reliable weight is calculated based on the contact strength coefficient. The free trend reference set is determined by the free weight sum of the neighborhood set or the candidate reference position, and then the free warping trend value is calculated. Compensation is calculated based on the consistency of the direction of the free warping trend with the measured residual, the contact strength coefficient, and the support constraint coverage coefficient to obtain the low-constraint equivalent warping residual. The overall warping amplitude, edge warping amplitude, torsion amplitude, and support flattening underestimation ratio are calculated using the low-constraint equivalent warping residual. The warping degree identification value of the PVC panel is obtained by weighted summation with preset weights. The over-limit ratio is determined by comparing the various warp amplitude and warp degree identification values ​​with the allowable threshold, and the warp level is determined. Combined with the support flattening underestimation ratio, the support flattening risk warning is output as the final detection result.

[0006] Optionally, the step of determining the number of axial and lateral sampling positions based on the detection length and the maximum allowable spacing, and uniformly dividing the coordinates of each sampling position, includes: By defining the length and width directions of the PVC panel to be tested as the axial and transverse directions respectively, and taking the intersection of the axial starting point and the transverse starting edge as the origin of the detection coordinates, the axial coordinate increasing direction and the transverse coordinate increasing direction are established. Based on the axial detection length, the transverse detection width, and the maximum allowable sampling interval, the number of axial sampling positions and the number of transverse sampling positions are determined respectively. By uniformly dividing the detection length and detection width, the axial coordinates and transverse coordinates corresponding to each sampling position are determined, thereby forming a two-dimensional sampling coordinate grid for warpage detection.

[0007] Optionally, the step of collecting height and pressure values ​​at each sampling location under no-load and working conditions, calculating the measured height value, and determining the contact strength coefficient at each sampling location based on the no-load pressure reference threshold and the contact reference pressure value includes: When no PVC panels are installed, collect no-load distance and no-load pressure values, and determine the no-load pressure benchmark threshold based on the no-load pressure value; After the PVC panels are placed, the working distance and contact pressure values ​​are collected, and the measured height value at each sampling position is determined by the difference between the no-load distance value and the working distance value. The effective contact pressure set is determined based on the relationship between the contact pressure value and the no-load pressure reference threshold. When the effective contact pressure set is not empty, the contact reference pressure value is calculated. Then, the contact pressure intensity at each sampling location is normalized to a contact intensity coefficient with a value range of zero to one.

[0008] Optionally, the step of calculating the attitude fitting weights from the contact strength coefficient, using weighted least squares to fit the overall attitude plane, and using the difference between the measured height value and the overall attitude plane height value as the measured warping residual under supported conditions includes: The attitude fitting weight is determined based on the contact strength coefficient, so that the influence of the strong contact flattening area on the overall attitude plane fitting is reduced. Based on the axial coordinates, lateral coordinates, measured height values, and attitude fitting weights of each sampling location, the overall attitude plane coefficient is determined by minimizing the weighted sum of squared height deviations. The measured height value is then subtracted from the corresponding overall attitude plane height value to obtain the measured warping residual under the supported state, thereby eliminating the interference of overall tilt placement on local warping identification.

[0009] Optionally, the step of determining the support-affected foundation radius according to the actual sampling interval and constructing a neighborhood set, and using the contact strength coefficient within the neighborhood to calculate the support constraint coverage coefficient at each sampling location, includes: The radius of the support-affected foundation is determined based on the actual axial sampling interval and the actual lateral sampling interval. Sampling positions whose planar distance from the current sampling position is not greater than the radius of the support-affected foundation are identified as neighborhood positions. Subsequently, based on the contact strength coefficient of each sampling position in the neighborhood set and its planar distance from the current sampling position, a distance-weighted calculation is performed to obtain the support constraint coverage coefficient of the current sampling position, which is used to characterize the degree of coverage of the support contact flattening effect near the current sampling position.

[0010] Optionally, the step of calculating the free reliable weight based on the contact strength coefficient, determining the free trend reference set through the free weight sum of the neighborhood set or candidate reference positions, and then calculating the free warping trend value includes: The free and reliable weights of each sampling location are determined based on the contact strength coefficient, and the sum of the free and reliable weights in the neighborhood set is used to determine whether the neighborhood set can be directly used as the free trend reference set. When there are insufficient free references in the neighborhood, sampling positions with free confidence weights greater than the preset free confidence threshold are further selected from the entire detection area as candidate free reference positions. If there are still no candidate free reference positions, the fallback free reference positions are selected according to the free confidence weights. The free warping trend value that the current sampling position should exhibit under low constraint is obtained by weighting the measured warping residuals, free confidence weights, and planar distances of each reference sampling position within the final free trend reference set.

[0011] Optionally, the step of calculating compensation based on the consistency of the direction of the free warping trend and the measured residual, the contact strength coefficient, and the support constraint coverage coefficient to obtain the low-constraint equivalent warping residual includes: By comparing the directional relationship between the free warp trend value and the measured warp residual under the support state, the directional consistency judgment value is determined to avoid mistaking local sudden changes in opposite directions as support flattening. The flattening difference is determined by the difference between the amplitude of the free warping trend value and the measured warping residual. The counter-flattening compensation amount of the support is calculated by combining the contact strength coefficient, support constraint coverage coefficient, direction consistency judgment value and the sign function value of the free warping trend value. The measured warpage residual under the supported condition is added to the support reverse flattening compensation amount to obtain the low-constraint equivalent warpage residual.

[0012] Optionally, the step of calculating the overall warpage amplitude, edge warpage amplitude, torsion amplitude, and support flattening underestimation ratio using low-constraint equivalent warpage residuals, and then weighting and summing them with preset weights to obtain the PVC panel warpage identification value, includes: The overall warping amplitude is calculated based on the maximum and minimum values ​​of the low-constraint equivalent warping residuals at all sampling locations. The edge warping amplitude is calculated based on the absolute value of the low-constraint equivalent warping residuals at the boundary sampling locations. The twist amplitude is calculated based on the low-constraint equivalent warping residuals at the four corner locations. The underestimation ratio of support flattening is determined by comparing the maximum absolute value of the equivalent warp residual under low constraint with the measured warp residual under support conditions. The warping degree of PVC panels is identified by a weighted combination of the overall warping amplitude, edge warping amplitude, twisting amplitude, and the underestimation ratio of support flattening.

[0013] Optionally, the step of comparing the various warp amplitude and warp degree identification values ​​with the allowable threshold to obtain the excess ratio and determine the warp level, and combining the support flattening underestimation ratio to output a support flattening risk warning as the final detection result, includes: Obtain the overall warping allowable threshold, edge warping allowable threshold, twist allowable threshold, and comprehensive warping degree identification threshold for the corresponding batch of PVC panels to be tested. Calculate the ratio of the overall warping amplitude, edge warping amplitude, twist amplitude, and the PVC panel warping degree identification value to the corresponding threshold, and use the maximum value as the excess ratio. Based on the proportion of warping exceeding the limit, the output indicates whether the warping is qualified, slightly warped, moderately warped, or severely warped. Based on the proportion of underestimation of support flattening, the output indicates whether the support flattening risk is low, medium, or high. The final test result includes the warping level of the PVC panel and the support flattening risk warning.

[0014] The present invention has the following beneficial effects: 1. In traditional flat-lay testing, the support-flattening effect of PVC panels due to their own weight or clamp fixation forcibly compresses real local warping to near the reference plane, causing severe distortion of the height data collected by the sensor and rendering subsequent evaluation unfounded. This invention creatively constructs a compensation mechanism from identifying flattening intensity to restoring free form. First, the degree of constraint of any test point on the surrounding contact and pressing area is accurately quantified by the support constraint coverage coefficient, and a "free trend reference set" with extremely low support influence is selected based on this. Then, using the warping residuals of these high-confidence reference points, a distance-weighted spatial interpolation method is used to deduce the free warping trend value that each point affected by flattening should exhibit under low constraint. Crucially, by introducing a direction consistency judgment value, the misjudgment of normal local depressions in the opposite direction as flattening and subsequent incorrect compensation is avoided, ensuring a high degree of consistency between the compensation direction and the actual deformation direction. Finally, the flattening difference is combined with the trend direction sign function to generate the support reverse flattening compensation amount, which corrects the original measured residual and obtains the low-constraint equivalent warping residual. This series of processes is equivalent to physically "stripping" away the external force constraint brought by the support contact, making the obtained warping data infinitely close to the true contour of the board when it is placed naturally without constraints.

[0015] 2. A reasonable sampling grid is fundamental to ensuring no omissions in warp detection and that the data is usable. Fixed step sizes or arbitrarily set sampling methods can easily lead to skipped features in key deformation areas or excessive sampling, resulting in data redundancy. This invention proposes an adaptive sampling position planning method that can automatically generate the optimal sampling grid layout based on the actual dimensions of the PVC panel to be inspected and the maximum allowable sampling interval set in the engineering. This method divides the axial detection length by the maximum allowable axial sampling interval, rounds up, and then adds one to obtain the total number of sampling positions along the length direction. This ensures that even if the detection length is not divisible by the interval, the sampling interval will never exceed the allowable upper limit. At the same time, both ends are forcibly included in the sampling positions, so that the warping trend of the beginning and end edges of the panel is completely captured. Similarly, the number of transverse sampling positions is determined using the same logic in the width direction. Then, the detection length and width are each evenly divided into intervals one less than the number of sampling positions, thereby ensuring that each sampling position is evenly distributed in rows and columns across the entire panel surface. This technique achieves an adaptive sampling point layout that "determines the grid based on the size of the plate and ensures accuracy by setting the upper limit of the spacing." This avoids missing detections of severe local warping due to sparse sampling points, while also eliminating the waste of detection time and data processing resources due to excessively dense sampling points. It balances detection completeness and economy, and provides a high-quality sampling point array with a regular layout and strong representativeness for subsequent complex calculations such as plane fitting and flattening compensation.

[0016] 3. Under supported testing conditions, the sampling points in the flattened area severely interfere with the accurate judgment of the overall placement posture of the board. If these forced contact points are treated the same as free suspension points, the fitted posture plane will tilt towards the flattened area, leading to an incorrect estimate of the warping degree of the entire board surface. This invention addresses this by designing a two-tiered anti-interference strategy. In the first tier, a statistically robust no-load pressure benchmark threshold is constructed using the median and absolute median difference of the no-load pressure values ​​when the board is not placed, effectively resisting sensor noise and data disturbances. After placing the board, the truly effective contact points are selected by comparing the contact pressure values ​​with the benchmark threshold, and the contact strength coefficient of each point is calculated based on the median, forming a continuous strength mapping from zero to one. In the second tier, "one minus the contact strength coefficient" is creatively used as the posture fitting weight, so that points with greater flattening strength receive lower fitting participation weights, even approaching zero, while suspension points maintain a complete weight close to one. When solving the overall tilt plane using weighted least squares, this mechanism automatically suppresses the dragging effect of strong contact areas on the plane equation, ensuring that the fitted attitude plane truly reflects the natural resting tendency of the entire sheet rather than the artifacts of forced deformation. The measured warpage residual obtained after subtracting this attitude plane eliminates attitude tilt artifacts to the greatest extent, laying a clean initial data foundation for subsequent free warpage reconstruction.

[0017] 4. Single-dimensional warp indices often only partially reflect a certain type of deformation characteristic of the board, failing to support comprehensive and accurate quality assessment. They are particularly prone to overlooking diagonal torsional deformation and detection risks caused by support flattening. This invention constructs a multi-dimensional comprehensive evaluation system, extracting four core indicators from the compensated low-constraint equivalent warp residuals. The overall warp amplitude reflects the overall arching degree of the board through the difference between the maximum and minimum residuals across the entire board surface; the edge warp amplitude is obtained by taking the maximum value from the absolute values ​​of the residuals at the four edge sampling points, specifically capturing the extreme state of edge warp; the torsion amplitude accurately quantifies the degree of torsion in the diagonal direction through cross-subtraction and halving of the residuals at the four corner points. Crucially, a support flattening underestimation ratio is introduced. By comparing the change in the maximum absolute value of the residuals before and after compensation, the underestimation caused by directly using the measured support values ​​is objectively measured. Subsequently, these physically meaningful indicators are weighted and integrated according to the following proportions: overall warp weight 45%, edge warp weight 30%, torsion weight 15%, and underestimation correction weight 10%, forming a unique warp degree identification value. Finally, by calculating the excess percentage of each indicator relative to the batch's allowable threshold and taking the maximum value, the system automatically determines four levels of warping status: qualified, mild, moderate, and severe. Based on the underestimated percentage, it outputs three levels of support flattening risk warnings: low, medium, and high. This process compresses complex and multidimensional deformation information into intuitive levels and risk language, not only improving the objectivity and comprehensiveness of quality judgment but also specifically revealing to operators the possibility of misjudgment due to flattening, providing truly reliable and interpretable decision-making basis for the factory quality control of PVC panels. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the process of the present invention; Figure 2 This is a schematic diagram of the sampling coordinate grid and the deployment of smart sensors; Figure 3 This is a schematic diagram showing the calculation and output of warpage index results; Figure 4 This is a schematic diagram of the process for obtaining low-constraint equivalent warp residuals. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Example 1, see Figures 1 to 4 A method for identifying the degree of warping of PVC ceiling panels based on smart sensors, comprising: The number of axial and lateral sampling positions is determined based on the detection length and the maximum allowable spacing, and the coordinates of each sampling position are evenly divided. Collect height and pressure values ​​at each sampling location under no-load and working conditions, calculate the measured height value, and determine the contact strength coefficient at each sampling location based on the no-load pressure reference threshold and the contact reference pressure value. The attitude fitting weights are calculated from the contact strength coefficient, and the overall attitude plane is fitted by weighted least squares. The difference between the measured height value and the overall attitude plane height value is used as the measured warping residual under the supported state. The support influence base radius is determined based on the actual sampling interval, and a neighborhood set is constructed. The support constraint coverage coefficient at each sampling location is obtained using the contact strength coefficient within the neighborhood. The free reliable weight is calculated based on the contact strength coefficient. The free trend reference set is determined by the free weight sum of the neighborhood set or the candidate reference position, and then the free warping trend value is calculated. Compensation is calculated based on the consistency of the direction of the free warping trend with the measured residual, the contact strength coefficient, and the support constraint coverage coefficient to obtain the low-constraint equivalent warping residual. The overall warping amplitude, edge warping amplitude, torsion amplitude, and support flattening underestimation ratio are calculated using the low-constraint equivalent warping residual. The warping degree identification value of the PVC panel is obtained by weighted summation with preset weights. The over-limit ratio is determined by comparing the various warp amplitude and warp degree identification values ​​with the allowable threshold, and the warp level is determined. Combined with the support flattening underestimation ratio, the support flattening risk warning is output as the final detection result.

[0021] The process of determining the number of axial and lateral sampling positions based on the detection length and the maximum allowable spacing, and uniformly dividing the coordinates of each sampling position, includes: The length direction of the PVC panel to be tested is defined as the axial direction, and the width direction of the PVC panel to be tested is defined as the transverse direction. One end of the PVC panel to be tested along its length is defined as the axial starting end. In the observation direction from the axial starting end to the other end, the width edge located on the left side of the PVC panel to be tested is defined as the transverse starting edge. The intersection of the axial starting end and the transverse starting edge is defined as the origin of the detection coordinates. The direction away from the origin of the detection coordinate along the length of the PVC panel is defined as the direction of axial coordinate increase, and the direction away from the origin of the detection coordinate along the width of the PVC panel is defined as the direction of lateral coordinate increase. Obtain the axial detection length, lateral detection width, maximum allowable axial sampling interval, and maximum allowable lateral sampling interval of the PVC panel to be inspected; The number of axial sampling positions is determined based on the axial detection length and the maximum allowable axial sampling interval. Specifically, the quotient obtained by dividing the axial detection length by the maximum allowable axial sampling interval is rounded up, and then one is added to the rounded result to obtain the number of axial sampling positions. Wherein, the number of axial sampling positions represents the total number of sampling positions set along the length direction of the PVC panel to be tested, the axial detection length represents the detection length of the PVC panel to be tested along the length direction, the maximum allowable axial sampling interval represents the maximum allowable interval between two adjacent axial sampling positions, and rounding up means taking the calculation result as the smallest integer not less than the calculation result; The number of horizontal sampling positions is determined based on the horizontal detection width and the maximum allowable horizontal sampling interval. Specifically, the quotient obtained by dividing the horizontal detection width by the maximum allowable horizontal sampling interval is rounded up, and then one is added to the rounded result to obtain the number of horizontal sampling positions. Among them, the number of lateral sampling positions represents the total number of sampling positions set along the width direction of the PVC panel to be tested, the lateral detection width represents the detection width of the PVC panel to be tested along the width direction, and the maximum allowable lateral sampling interval represents the maximum allowable interval between two adjacent lateral sampling positions. The axial detection length is evenly divided into axial intervals that are one less than the number of axial sampling positions. The axial coordinate of any axial sampling position is determined as follows: the axial sampling position number of the axial sampling position is reduced by one, the result is multiplied by the axial detection length, and then divided by the difference obtained by subtracting one from the number of axial sampling positions to obtain the axial coordinate of the axial sampling position. The value range of the axial sampling position number is all positive integers from one to the number of axial sampling positions; Wherein, the axial coordinate represents the distance of the corresponding axial sampling position relative to the origin of the detection coordinate in the direction of increasing axial coordinate, and the axial sampling position number represents the sequential number of the corresponding axial sampling positions arranged in ascending order of axial coordinate. The horizontal detection width is evenly divided into horizontal intervals that are one less than the number of horizontal sampling positions. The horizontal coordinate of any horizontal sampling position is determined as follows: subtract one from the horizontal sampling position number of the horizontal sampling position, multiply the result by the horizontal detection width, and then divide by the difference obtained by subtracting one from the number of horizontal sampling positions to obtain the horizontal coordinate of the horizontal sampling position. The range of values ​​for the horizontal sampling position number is all positive integers from one to the number of horizontal sampling positions; Wherein, the horizontal coordinate represents the distance of the corresponding horizontal sampling position relative to the origin of the detection coordinate in the direction of increasing horizontal coordinate, and the horizontal sampling position number represents the sequential number of the corresponding horizontal sampling positions arranged in ascending order of horizontal coordinate.

[0022] A reasonable sampling grid is fundamental to ensuring no omissions in warp detection and that the data is usable. Fixed step sizes or arbitrarily set sampling methods can easily lead to skipped features in key deformation areas or excessive sampling density, resulting in data redundancy. This invention proposes an adaptive sampling position planning method that can automatically generate an optimal sampling grid layout based on the actual dimensions of the PVC panel to be inspected and the maximum allowable sampling interval set in the engineering. This method divides the axial detection length by the maximum allowable axial sampling interval, rounds up, and then adds one to obtain the total number of sampling positions along the length direction. This ensures that even if the detection length is not divisible by the interval, the sampling interval will never exceed the allowable upper limit. At the same time, both ends are forcibly included in the sampling positions, so that the warping trend of the beginning and end edges of the panel is completely captured. Similarly, the number of transverse sampling positions is determined using the same logic in the width direction. Then, the detection length and width are each evenly divided into intervals one less than the number of sampling positions, thereby ensuring that each sampling position is evenly distributed in rows and columns across the entire panel surface. This technique achieves an adaptive sampling point layout that "determines the grid based on the size of the plate and ensures accuracy by setting the upper limit of the spacing." This avoids missing detections of severe local warping due to sparse sampling points, while also eliminating the waste of detection time and data processing resources due to excessively dense sampling points. It balances detection completeness and economy, and provides a high-quality sampling point array with a regular layout and strong representativeness for subsequent complex calculations such as plane fitting and flattening compensation.

[0023] The process involves collecting height and pressure values ​​at each sampling location under both unloaded and working conditions, calculating the measured height value, and determining the contact strength coefficient at each sampling location based on the unloaded pressure reference threshold and the contact reference pressure value. This includes: When no PVC panels are installed, the distance to the support reference surface corresponding to each sampling position is measured by a height sensor to obtain the unloaded distance value of each sampling position. The no-load pressure value at each sampling point is collected by a pressure sensor installed on the support surface; The no-load pressure benchmark threshold is determined based on the no-load pressure values ​​at each sampling location. Specifically, the median of all no-load pressure values ​​is calculated first, then the absolute value of the difference between each no-load pressure value and the median is calculated, and the median of all absolute values ​​is calculated. The median of the absolute value is multiplied by three and then added to the median of all no-load pressure values ​​to obtain the no-load pressure benchmark threshold. Among them, the no-load pressure reference threshold is the pressure judgment benchmark used to distinguish between no-load pressure disturbance and actual contact pressure; the no-load pressure value is the pressure sensor value collected at the corresponding sampling position when no PVC panel is placed; the median is the value in the middle position after sorting the values ​​involved in the calculation according to their size; and the absolute value is the value after removing the positive and negative directions. After the PVC panel is placed at the support detection position, the working distance value from the sensor to the upper surface of the PVC panel at each sampling position is collected by the height sensor, and the contact pressure value after the PVC panel is placed at each sampling position is collected by the pressure sensor. The measured height of the upper surface of the PVC buckle relative to the supporting reference surface is calculated based on the unloaded distance measurement value and the working distance measurement value. Specifically, the working distance measurement value is subtracted from the unloaded distance measurement value at the same sampling position to obtain the measured height of the upper surface of the PVC buckle at that sampling position. Among them, the measured height value represents the height of the upper surface of the PVC buckle at the corresponding sampling position relative to the supporting reference surface, the no-load distance measurement value represents the distance from the height sensor to the supporting reference surface when no PVC buckle is placed, and the working distance measurement value represents the distance from the height sensor to the upper surface of the PVC buckle after the PVC buckle is placed. Based on the contact pressure value at each sampling location and the no-load pressure reference threshold, an effective contact pressure set is determined. The effective contact pressure set consists of contact pressure values ​​that meet the following conditions: the contact pressure value is greater than the no-load pressure reference threshold, and the axial sampling location number corresponding to the contact pressure value is within the range of one to the number of axial sampling locations, and the lateral sampling location number is within the range of one to the number of lateral sampling locations. Among them, the effective contact pressure set represents the set of pressure values ​​that are determined to be in effective contact between the PVC panel and the supporting surface, and the contact pressure value represents the pressure sensor value collected at the corresponding sampling position after the PVC panel is placed. When the effective contact pressure set is not empty, the contact reference pressure value is determined according to the effective contact pressure set. Specifically, the median of all contact pressure values ​​in the effective contact pressure set is added to the minimum positive number of the pressure calculation to obtain the contact reference pressure value. Among them, the contact reference pressure value represents the reference pressure used to normalize the contact pressure intensity, and the minimum positive number in the pressure calculation represents a positive number used to avoid the denominator being zero in the pressure normalization calculation. When the effective contact pressure set is empty, the contact strength coefficient at each sampling location is set to zero. The contact strength coefficient represents the contact strength between the PVC panel and the supporting surface at the corresponding sampling location; When the effective contact pressure set is not empty, the contact strength coefficient of each sampling location is determined based on the no-load pressure reference threshold and the contact reference pressure value, specifically as follows: When the contact pressure value at the corresponding sampling location is less than or equal to the no-load pressure reference threshold, the contact strength coefficient at that sampling location is determined to be zero. When the contact pressure value at the corresponding sampling location is greater than the no-load pressure reference threshold and less than the contact reference pressure value, the difference between the contact pressure value and the no-load pressure reference threshold is divided by the difference between the contact reference pressure value and the no-load pressure reference threshold to obtain the contact strength coefficient at that sampling location. When the contact pressure value at the corresponding sampling location is greater than or equal to the contact reference pressure value, the contact strength coefficient at that sampling location is determined to be one. The contact strength coefficient is greater than or equal to zero and less than or equal to one.

[0024] The process of calculating the attitude fitting weights from the contact strength coefficient, using weighted least squares to fit the overall attitude plane, and using the difference between the measured height value and the overall attitude plane height value as the measured warping residual under supported conditions includes: To reduce the impact of the strong contact flattening area on the attitude plane fitting, the attitude fitting weight is determined according to the contact strength coefficient. Specifically, the contact strength coefficient at the corresponding sampling position is subtracted from one, and the minimum positive number of the height is added to obtain the attitude fitting weight at that sampling position. Among them, the attitude fitting weight represents the weight of the corresponding sampling position when participating in the overall attitude plane fitting, and the height calculation is a very small positive number to ensure that the strong contact area still retains a very low weight without causing the calculation to be interrupted. Based on the coordinates of each sampling location, the measured height value, and the attitude fitting weights, the overall attitude plane coefficient is determined to minimize the weighted sum of squared height deviations. Specifically: For each sampling position, the tilt coefficient of the overall attitude plane in the axial direction is multiplied by the axial coordinate of the sampling position, the tilt coefficient of the overall attitude plane in the lateral direction is multiplied by the lateral coordinate of the sampling position, and then added to the height intercept of the overall attitude plane to obtain the height value of the overall attitude plane at the sampling position. Subtract the overall attitude plane height value from the measured height value at the sampling location, square the result, and then multiply by the attitude fitting weight at the sampling location. The calculation results corresponding to all sampling positions are accumulated, and the axial tilt coefficient, lateral tilt coefficient and height intercept corresponding to the minimum value of the accumulated result are selected as the overall attitude plane coefficient. Among them, the tilt coefficient of the overall attitude plane in the axial direction represents the degree of height change of the overall attitude plane of the PVC panel in the axial direction, the tilt coefficient of the overall attitude plane in the lateral direction represents the degree of height change of the overall attitude plane of the PVC panel in the lateral direction, and the height intercept of the overall attitude plane represents the height value of the overall attitude plane at the origin of the detection coordinate. Subtract the measured height value from the overall attitude plane height value to obtain the measured warping residual under the supported state; Among them, the measured warpage residual under the supported state represents the local height deviation of the corresponding sampling position under the supported detection state after deducting the overall placement posture.

[0025] Under supported testing conditions, the sampling points in the flattened area severely interfere with the accurate judgment of the overall placement posture of the board. If these forced contact points are treated the same as free suspension points, the fitted posture plane will tilt towards the flattened area, leading to an incorrect estimate of the warping degree of the entire board surface. This invention addresses this by designing a two-tiered anti-interference strategy. In the first tier, a statistically robust no-load pressure benchmark threshold is constructed using the median and absolute median difference of the no-load pressure values ​​before board placement, effectively resisting sensor noise and data disturbances. After board placement, the truly effective contact points are selected by comparing the contact pressure values ​​with the benchmark threshold, and the contact strength coefficient of each point is calculated based on the median, forming a continuous strength mapping from zero to one. In the second tier, "one minus the contact strength coefficient" is creatively used as the posture fitting weight, so that points with greater flattening strength receive lower fitting participation weights, even approaching zero, while suspension points maintain a complete weight close to one. When solving the overall tilt plane using weighted least squares, this mechanism automatically suppresses the dragging effect of strong contact areas on the plane equation, ensuring that the fitted attitude plane truly reflects the natural resting tendency of the entire sheet rather than the artifacts of forced deformation. The measured warpage residual obtained after subtracting this attitude plane eliminates attitude tilt artifacts to the greatest extent, laying a clean initial data foundation for subsequent free warpage reconstruction.

[0026] The process of determining the support influence base radius based on the actual sampling interval and constructing a neighborhood set, and using the contact strength coefficient within the neighborhood to calculate the support constraint coverage coefficient at each sampling location, includes: To determine whether there is a support flattening effect near a certain sampling location, the radius of the support-affected foundation is determined based on the actual sampling interval, specifically: The actual axial sampling interval is obtained by dividing the axial detection length by the difference between the number of axial sampling positions minus one. The actual horizontal sampling interval is obtained by dividing the horizontal detection width by the difference between the number of horizontal sampling positions minus one. Add the square of the actual axial sampling interval to the square of the actual lateral sampling interval, take the square root of the sum, and multiply the result by two to obtain the radius of the support-affected foundation. For any current sampling position, sampling positions whose planar distance from the current sampling position is not greater than the radius of the support-affected foundation are determined as neighborhood positions; The planar distance between any neighboring location and the current sampling location is calculated as follows: The difference between the axial coordinate of the current sampling position and the axial coordinate of the neighboring position is squared, and the difference between the lateral coordinate of the current sampling position and the lateral coordinate of the neighboring position is squared. The two squared results are added together and then the square root is taken to obtain the planar distance. Wherein, the planar distance represents the distance between the neighboring location and the current sampling location within the detection plane; For any current sampling location, a neighborhood set is constructed, which consists of sampling locations that satisfy the following conditions: The planar distance between the sampling location and the current sampling location is less than or equal to the radius of the support-affected foundation. Based on the contact strength coefficients of each sampling location within the neighborhood set, the support constraint coverage coefficient of the current sampling location is calculated as follows: For each sampling location within the neighborhood set, the contact strength coefficient at that location is divided by the sum of the minimum positive numbers obtained by adding the plane distance and height between that location and the current sampling location to obtain the contact distance weighted value; all contact distance weighted values ​​within the neighborhood set are summed to obtain the contact distance weighted sum; for each sampling location within the neighborhood set, the distance weight is calculated by dividing the contact strength coefficient at that location by the sum of the minimum positive numbers obtained by adding the plane distance and height between that location and the current sampling location to obtain the distance weight value; all distance weight values ​​within the neighborhood set are summed to obtain the distance weighted sum; the contact distance weighted sum is divided by the distance weighted sum to obtain the support constraint coverage coefficient at the current sampling location.

[0027] The calculation of free reliable weights based on contact strength coefficients, determining the free trend reference set through the sum of free weights in the neighborhood set or candidate reference positions, and then calculating the free warping trend value includes: For each sampling location, the free and reliable weight is determined based on the contact strength coefficient. Specifically, the contact strength coefficient at the corresponding sampling location is subtracted from one to obtain the free and reliable weight at that sampling location. The larger the value of the free and reliable weight, the smaller the influence of the support contact effect on the corresponding sampling position, and the more suitable the actual warpage residual is as a reference data for the free warpage trend value. The smaller the value of the free and reliable weight, the greater the influence of the support contact effect on the corresponding sampling position, and the lower the weight of the actual warping residual when participating in the calculation of the free warping trend value. To calculate the sum of free weights within the neighborhood set corresponding to any sampling location, specifically, the free and reliable weights of each neighborhood location within the neighborhood set corresponding to the sampling location are summed to obtain the sum of free weights. Set a free reference lower bound, and set the free reference lower bound value to three; When the sum of free weights is greater than or equal to the lower limit of the free reference, the neighborhood set of the corresponding sampling location is determined as the free trend reference set; When the sum of free weights is less than the lower limit of free reference, the sampling position with a free confidence weight greater than the preset free confidence threshold is selected as the candidate free reference position. The preset free and credible threshold is set to 0.5. The preset free and credible threshold is used to filter the lower limit of the free and credible weight of candidate free reference positions. The candidate free reference position set consists of sampling positions that meet the following conditions: the axial sampling position number of the sampling position is within the range of one to the number of axial sampling positions, the lateral sampling position number of the sampling position is within the range of one to the number of lateral sampling positions, and the free confidence weight at the sampling position is greater than the preset free confidence threshold. When the candidate free reference position set is not empty, the candidate free reference position set is determined as the free trend reference set; When the candidate free reference location set is empty, the free confidence weights of each sampling location in the entire detection area are sorted from largest to smallest, and a preset number of sampling locations with the highest ranking are selected as the free trend reference set. The preset quantity is determined as follows: multiply the number of axial sampling positions by the number of lateral sampling positions to obtain the total number of sampling positions; multiply the total number of sampling positions by 5% and round up; compare the rounded result with three and take the larger value as the number of fallback free reference positions. The final free trend reference set is determined according to the following rules: When the sum of free weights is greater than or equal to the lower limit of free reference, the final free trend reference set is the neighborhood set; when the sum of free weights is less than the lower limit of free reference and the candidate free reference position set is not empty, the final free trend reference set is the candidate free reference position set; when the sum of free weights is less than the lower limit of free reference and the candidate free reference position set is empty, the final free trend reference set is a set of sampling positions formed by selecting the number of fallback free reference positions in the entire detection area after sorting them from largest to smallest according to their free confidence weights. For any current sampling location, based on the measured warpage residuals of each low-constraint sampling location in the free trend reference set, calculate the free warpage trend value that the current sampling location should exhibit under low-constraint conditions, specifically: For each reference sampling position within the final free trend reference set, the free confidence weight at that reference sampling position is multiplied by the measured warpage residual under the support state at that reference sampling position, and then divided by the sum of the smallest positive numbers obtained by calculating the plane distance plus height between that reference sampling position and the current sampling position to obtain the residual weight value; all residual weight values ​​are summed to obtain the residual weighted sum value; for each reference sampling position within the final free trend reference set, the free confidence weight at that reference sampling position is divided by the sum of the smallest positive numbers obtained by calculating the plane distance plus height between that reference sampling position and the current sampling position to obtain the trend weight value; all trend weight values ​​are summed to obtain the trend weight sum value; the residual weighted sum value is divided by the trend weight sum value to obtain the free warpage trend value at the current sampling position.

[0028] The calculation of compensation based on the consistency of the direction of the free warping trend and the measured residual, the contact strength coefficient, and the support constraint coverage coefficient to obtain the low-constraint equivalent warping residual includes: To avoid misinterpreting local abrupt changes in opposite directions as flattening, a consistent direction criterion is established, specifically as follows: When the product of the free warping trend value at the same sampling location and the measured warping residual under the support state is greater than or equal to zero, the direction consistency judgment value at that sampling location is determined to be one. When the product of the free warping trend value at the same sampling location and the measured warping residual under the supported state is less than zero, and the absolute value of the measured warping residual under the supported state is less than or equal to the minimum positive number of the height calculation, the direction consistency judgment value at that sampling location is determined to be one. When the product of the free warping trend value at the same sampling location and the measured warping residual under the supported state is less than zero, and the absolute value of the measured warping residual under the supported state is greater than the minimum positive number of the height calculation, the direction consistency judgment value at that sampling location is determined to be zero. Based on the contact strength coefficient, support constraint coverage coefficient, direction consistency judgment value, and flattening difference value, the support reverse flattening compensation amount is calculated. Specifically, the sign function values ​​corresponding to the contact strength coefficient, support constraint coverage coefficient, direction consistency judgment value, flattening difference value, and free warping trend value at the same sampling location are multiplied in sequence to obtain the support reverse flattening compensation amount at that sampling location. The flattening difference is calculated as follows: the absolute value of the free warping trend value is subtracted from the absolute value of the measured warping residual under the support state to obtain the amplitude difference; the amplitude difference is compared with zero, and the larger of the two is taken as the flattening difference. The sign function corresponding to the sign function value is specifically as follows: when the free warping trend value is greater than zero, the sign function value is one; when the free warping trend value is equal to zero, the sign function value is zero; when the free warping trend value is less than zero, the sign function value is negative one. The measured warpage residual under the supported condition is added to the support reverse flattening compensation amount to obtain the low-constraint equivalent warpage residual. Among them, the low-constraint equivalent warp residual represents the equivalent warp residual at the corresponding sampling position under low-constraint conditions after support reverse flattening compensation.

[0029] In traditional flat-lay testing, the support-flattening effect of PVC panels due to their own weight or clamp fixation forcibly compresses real local warping to near the reference surface, causing severe distortion of the height data collected by the sensor and rendering subsequent evaluation unfounded. This invention creatively constructs a compensation mechanism from identifying flattening intensity to restoring free form. First, the degree of constraint of any test point on the surrounding contact and pressing area is accurately quantified through a support constraint coverage coefficient, and based on this, a "free trend reference set" with extremely low support influence is selected. Next, using the warping residuals of these high-confidence reference points, a distance-weighted spatial interpolation method is used to deduce the free warping trend value that each point affected by flattening should exhibit under low constraint. Crucially, by introducing a direction consistency judgment value, the misjudgment of normal local indentations in the opposite direction as flattening and subsequent incorrect compensation is avoided, ensuring a high degree of consistency between the compensation direction and the actual deformation direction. Finally, the flattening difference is combined with the trend direction sign function to generate the support reverse flattening compensation amount, which corrects the original measured residual and obtains the low-constraint equivalent warping residual. This series of processes is equivalent to physically "stripping" away the external force constraint brought by the support contact, making the obtained warping data infinitely close to the true contour of the board when it is placed naturally without constraints.

[0030] The process of calculating the overall warpage amplitude, edge warpage amplitude, torsion amplitude, and support flattening underestimation ratio using low-constraint equivalent warpage residuals, and then weighting and summing these values ​​with preset weights to obtain the PVC panel warpage identification value, includes: The overall warpage amplitude is calculated based on the maximum and minimum values ​​of the low-constraint equivalent warpage residuals. Specifically, the maximum value is selected from the low-constraint equivalent warpage residuals corresponding to all sampling positions, and the minimum value is selected from the low-constraint equivalent warpage residuals corresponding to all sampling positions. The maximum value is then subtracted from the minimum value to obtain the overall warpage amplitude of the PVC panel. The edge warping amplitude is calculated based on the low-constraint equivalent warping residual, specifically as follows: The maximum value is selected from the absolute values ​​of all low-constraint equivalent warp residuals corresponding to the first lateral sampling position, the maximum value is selected from the absolute values ​​of all low-constraint equivalent warp residuals corresponding to the last lateral sampling position, the maximum value is selected from the absolute values ​​of all low-constraint equivalent warp residuals corresponding to the first axial sampling position, and the maximum value is selected from the absolute values ​​of all low-constraint equivalent warp residuals corresponding to the last axial sampling position. The maximum value is then selected again from the above four maximum values ​​as the edge warp amplitude of the PVC panel. To identify whether PVC panels exhibit diagonal torsion, the torsion amplitude is calculated based on the equivalent warpage residuals of the low-constraint corners. Specifically: Subtract the low-constraint equivalent warp residuals at the first axial sampling position and the last lateral sampling position from the low-constraint equivalent warp residuals at the first axial sampling position and the first lateral sampling position, then subtract the low-constraint equivalent warp residuals at the last axial sampling position and the first lateral sampling position, and add the low-constraint equivalent warp residuals at the last axial sampling position and the last lateral sampling position. Take the absolute value of the result and divide the absolute value by two to obtain the torsion amplitude of the PVC panel. The torsion amplitude represents the relative degree of torsion between two diagonal directions of the PVC panel. The underestimation ratio of support level is calculated as follows: The maximum value is selected from the absolute values ​​of the low-constraint equivalent warp residuals corresponding to all sampling locations, and the maximum value is also selected from the absolute values ​​of the measured warp residuals under the support state corresponding to all sampling locations. The maximum value of the low-constraint equivalent warp residual is subtracted from the maximum value of the measured warp residual under the support state to obtain the underestimation difference. The maximum value of the low-constraint equivalent warp residual is added to the minimum positive number of the height calculation to obtain the underestimation normalized denominator. The underestimation difference is divided by the underestimation normalized denominator to obtain the support flattening underestimation ratio. Among them, the underestimation ratio of support flattening indicates the degree of underestimation of the true degree of warping when the measured warping residual under the support state is directly used for judgment; the larger the underestimation ratio of support flattening, the more obvious the underestimation of the true degree of warping when the measured warping residual under the support state is directly used for judgment. Set the overall warp amplitude weight, edge warp amplitude weight, twist amplitude weight, and support flattening underestimation correction weight, and make the four weights satisfy the following conditions: the sum of the overall warp amplitude weight, edge warp amplitude weight, twist amplitude weight, and support flattening underestimation correction weight is equal to one, and the overall warp amplitude weight, edge warp amplitude weight, twist amplitude weight, and support flattening underestimation correction weight are all greater than zero; Among them, the overall warp amplitude weight represents the weight of the overall warp amplitude in the warp degree identification value, the edge warp amplitude weight represents the weight of the edge warp amplitude in the warp degree identification value, the twist amplitude weight represents the weight of the twist amplitude in the warp degree identification value, and the support flattening underestimation correction weight represents the correction weight of the support flattening underestimation ratio in the warp degree identification value. Based on the overall warpage amplitude, edge warpage amplitude, torsion amplitude, and the underestimation ratio of support flattening, the warpage identification value of the PVC panel is calculated as follows: Multiply the overall warp amplitude weight by the overall warp amplitude to obtain the overall warp amplitude weighted term; multiply the edge warp amplitude weight by the edge warp amplitude to obtain the edge warp amplitude weighted term; multiply the torsion amplitude weight by the torsion amplitude to obtain the torsion amplitude weighted term; multiply the support flattening underestimation correction weight by the support flattening underestimation ratio, and then multiply by the overall warp amplitude to obtain the support flattening underestimation correction term; add the overall warp amplitude weighted term, the edge warp amplitude weighted term, the torsion amplitude weighted term, and the support flattening underestimation correction term to obtain the PVC panel warp degree identification value; Among them, the PVC panel warping degree identification value represents the quantitative result of the warping degree, which comprehensively reflects the overall warping, edge warping, twisting, and the underestimated impact of support flattening; As a preferred approach, the overall warping amplitude is weighted at 45%, the edge warping amplitude at 30%, the twist amplitude at 15%, and the support flattening underestimation correction at 10%.

[0031] The process of comparing various warpage amplitude and warpage degree identification values ​​with allowable thresholds to obtain the excess ratio and determine the warpage level, and combining the underestimation ratio of support flattening to output a support flattening risk warning as the final detection result, includes: Obtain the overall warpage allowable threshold, edge warpage allowable threshold, twist allowable threshold, and comprehensive warpage identification threshold for the corresponding batch of PVC panels to be inspected; Wherein, the overall warping allowable threshold is used to limit the upper limit of the overall warping amplitude, the edge warping allowable threshold is used to limit the upper limit of the edge warping amplitude, the twist allowable threshold is used to limit the upper limit of the twist amplitude, and the comprehensive warping degree identification threshold is used to limit the upper limit of the PVC panel warping degree identification value; The excess ratio is calculated based on the degree to which the overall warping amplitude, edge warping amplitude, twisting amplitude, and PVC panel warping identification value exceed the corresponding threshold. Specifically: Divide the overall warpage amplitude by the overall warpage allowable threshold to obtain the overall warpage exceedance ratio; divide the edge warpage amplitude by the edge warpage allowable threshold to obtain the edge warpage exceedance ratio; divide the torsion amplitude by the torsion allowable threshold to obtain the torsion exceedance ratio; divide the PVC panel warpage degree identification value by the comprehensive warpage degree identification threshold to obtain the comprehensive warpage exceedance ratio; select the maximum value among the overall warpage exceedance ratio, edge warpage exceedance ratio, torsion exceedance ratio, and comprehensive warpage exceedance ratio as the exceedance ratio; Among them, the excess ratio indicates the maximum extent to which the PVC panel under test exceeds the various allowable thresholds; The warping level of the PVC panel is output based on the excess ratio. The method for determining the warping level of the PVC panel is as follows: When the over-limit ratio is less than or equal to one, the output shows a qualified warping state. When the over-limit ratio is greater than 1 and less than or equal to 1.5, output a slightly warped state; When the over-limit ratio is greater than 1.5 and less than or equal to 2, output a moderate warping state. When the over-limit ratio is greater than two, the output shows a severe warping state. The support flattening risk warning is output based on the underestimation ratio of support flattening. The method for determining the support flattening risk warning is as follows: When the undervaluation ratio of support flattening is less than 0.10, output the risk of low support flattening. When the underestimation ratio of support flattening is greater than or equal to 0.10 and less than 0.25, the risk of support flattening is output. When the undervaluation ratio of support flattening is greater than or equal to 0.25, output a high support flattening risk. Among them, the risk warning of support flattening indicates the risk that the degree of warpage may be underestimated due to support contact flattening. The final test results are output, including the PVC panel warping level and support flattening risk warning.

[0032] Single-dimensional warp indices often only partially reflect a certain type of deformation characteristic of the board material, failing to support comprehensive and accurate quality assessment. They are particularly prone to overlooking diagonal torsional deformation and detection risks caused by support flattening. This invention constructs a multi-dimensional comprehensive evaluation system, extracting four core indicators from the compensated low-constraint equivalent warp residuals. The overall warp amplitude reflects the overall arching degree of the board material through the difference between the maximum and minimum residuals across the entire board surface; the edge warp amplitude is obtained by taking the maximum value from the absolute values ​​of the residuals at the four edge sampling points, specifically capturing the extreme state of edge warp; the torsion amplitude accurately quantifies the degree of torsion in the diagonal direction through cross-subtraction and halving of the residuals at the four corner points. Crucially, a support flattening underestimation ratio is introduced. By comparing the change in the maximum absolute value of the residuals before and after compensation, the underestimation caused by directly using the measured support values ​​is objectively measured. Subsequently, these physically meaningful indicators are weighted and integrated according to the following proportions: overall warp weight 45%, edge warp weight 30%, torsion weight 15%, and underestimation correction weight 10%, forming a unique warp degree identification value. Finally, by calculating the excess percentage of each indicator relative to the batch's allowable threshold and taking the maximum value, the system automatically determines four levels of warping status: qualified, mild, moderate, and severe. Based on the underestimated percentage, it outputs three levels of support flattening risk warnings: low, medium, and high. This process compresses complex and multidimensional deformation information into intuitive levels and risk language, not only improving the objectivity and comprehensiveness of quality judgment but also specifically revealing to operators the possibility of misjudgment due to flattening, providing truly reliable and interpretable decision-making basis for the factory quality control of PVC panels.

[0033] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0034] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for identifying the degree of warping of PVC ceiling panels based on intelligent sensors, characterized in that, include: The number of axial and lateral sampling positions is determined based on the detection length and the maximum allowable spacing, and the coordinates of each sampling position are evenly divided. Collect height and pressure values ​​at each sampling location under no-load and working conditions, calculate the measured height value, and determine the contact strength coefficient at each sampling location based on the no-load pressure reference threshold and the contact reference pressure value. The attitude fitting weights are calculated from the contact strength coefficient, and the overall attitude plane is fitted by weighted least squares. The difference between the measured height value and the overall attitude plane height value is used as the measured warping residual under the supported state. The support influence base radius is determined based on the actual sampling interval, and a neighborhood set is constructed. The support constraint coverage coefficient at each sampling location is obtained using the contact strength coefficient within the neighborhood. The free reliable weight is calculated based on the contact strength coefficient. The free trend reference set is determined by the free weight sum of the neighborhood set or the candidate reference position, and then the free warping trend value is calculated. Compensation is calculated based on the consistency of the direction of the free warping trend with the measured residual, the contact strength coefficient, and the support constraint coverage coefficient to obtain the low-constraint equivalent warping residual. The overall warping amplitude, edge warping amplitude, torsion amplitude, and support flattening underestimation ratio are calculated using the low-constraint equivalent warping residual. The warping degree identification value of the PVC panel is obtained by weighted summation with preset weights. The over-limit ratio is determined by comparing the various warp amplitude and warp degree identification values ​​with the allowable threshold, and the warp level is determined. Combined with the support flattening underestimation ratio, the support flattening risk warning is output as the final detection result.

2. The method for identifying the degree of warping of PVC panels based on intelligent sensors according to claim 1, characterized in that: The process of determining the number of axial and lateral sampling positions based on the detection length and the maximum allowable spacing, and uniformly dividing the coordinates of each sampling position, includes: By defining the length and width directions of the PVC panel to be tested as the axial and transverse directions respectively, and taking the intersection of the axial starting point and the transverse starting edge as the origin of the detection coordinates, the axial coordinate increasing direction and the transverse coordinate increasing direction are established. Based on the axial detection length, the transverse detection width, and the maximum allowable sampling interval, the number of axial sampling positions and the number of transverse sampling positions are determined respectively. By uniformly dividing the detection length and detection width, the axial coordinates and transverse coordinates corresponding to each sampling position are determined, thereby forming a two-dimensional sampling coordinate grid for warpage detection.

3. The method for identifying the degree of warping of PVC panels based on intelligent sensors according to claim 1, characterized in that: The process involves collecting height and pressure values ​​at each sampling location under both unloaded and working conditions, calculating the measured height value, and determining the contact strength coefficient at each sampling location based on the unloaded pressure reference threshold and the contact reference pressure value. This includes: When no PVC panels are installed, collect no-load distance and no-load pressure values, and determine the no-load pressure benchmark threshold based on the no-load pressure value; After the PVC panels are placed, the working distance and contact pressure values ​​are collected, and the measured height value at each sampling position is determined by the difference between the no-load distance value and the working distance value. The effective contact pressure set is determined based on the relationship between the contact pressure value and the no-load pressure reference threshold. When the effective contact pressure set is not empty, the contact reference pressure value is calculated. Then, the contact pressure intensity at each sampling location is normalized to a contact intensity coefficient with a value range of zero to one.

4. The method for identifying the degree of warping of PVC panels based on intelligent sensors according to claim 1, characterized in that: The process of calculating the attitude fitting weights from the contact strength coefficient, using weighted least squares to fit the overall attitude plane, and using the difference between the measured height value and the overall attitude plane height value as the measured warping residual under supported conditions includes: The attitude fitting weight is determined based on the contact strength coefficient, so that the influence of the strong contact flattening area on the overall attitude plane fitting is reduced. Based on the axial coordinates, lateral coordinates, measured height values, and attitude fitting weights of each sampling location, the overall attitude plane coefficient is determined by minimizing the weighted sum of squared height deviations. The measured height value is then subtracted from the corresponding overall attitude plane height value to obtain the measured warping residual under the supported state, thereby eliminating the interference of overall tilt placement on local warping identification.

5. The method for identifying the degree of warping of PVC panels based on intelligent sensors according to claim 1, characterized in that: The process of determining the support influence base radius based on the actual sampling interval and constructing a neighborhood set, and using the contact strength coefficient within the neighborhood to calculate the support constraint coverage coefficient at each sampling location, includes: The radius of the support-affected foundation is determined based on the actual axial sampling interval and the actual lateral sampling interval. Sampling positions whose planar distance from the current sampling position is not greater than the radius of the support-affected foundation are identified as neighborhood positions. Subsequently, based on the contact strength coefficient of each sampling position in the neighborhood set and its planar distance from the current sampling position, a distance-weighted calculation is performed to obtain the support constraint coverage coefficient of the current sampling position, which is used to characterize the degree of coverage of the support contact flattening effect near the current sampling position.

6. The method for identifying the degree of warping of PVC panels based on intelligent sensors according to claim 1, characterized in that: The calculation of free reliable weights based on contact strength coefficients, determining the free trend reference set through the sum of free weights in the neighborhood set or candidate reference positions, and then calculating the free warping trend value includes: The free and reliable weights of each sampling location are determined based on the contact strength coefficient, and the sum of the free and reliable weights in the neighborhood set is used to determine whether the neighborhood set can be directly used as the free trend reference set. When there are insufficient free references in the neighborhood, sampling positions with free confidence weights greater than the preset free confidence threshold are further selected from the entire detection area as candidate free reference positions. If there are still no candidate free reference positions, the fallback free reference positions are selected according to the free confidence weights. The free warping trend value that the current sampling position should exhibit under low constraint is obtained by weighting the measured warping residuals, free confidence weights, and planar distances of each reference sampling position within the final free trend reference set.

7. The method for identifying the degree of warping of PVC panels based on intelligent sensors according to claim 1, characterized in that: The calculation of compensation based on the consistency of the direction of the free warping trend and the measured residual, the contact strength coefficient, and the support constraint coverage coefficient to obtain the low-constraint equivalent warping residual includes: By comparing the directional relationship between the free warp trend value and the measured warp residual under the support state, the directional consistency judgment value is determined to avoid mistaking local sudden changes in opposite directions as support flattening. The flattening difference is determined by the difference between the amplitude of the free warping trend value and the measured warping residual. The counter-flattening compensation amount of the support is calculated by combining the contact strength coefficient, support constraint coverage coefficient, direction consistency judgment value and the sign function value of the free warping trend value. The measured warpage residual under the supported condition is added to the support reverse flattening compensation amount to obtain the low-constraint equivalent warpage residual.

8. The method for identifying the degree of warping of PVC panels based on intelligent sensors according to claim 1, characterized in that: The process of calculating the overall warpage amplitude, edge warpage amplitude, torsion amplitude, and support flattening underestimation ratio using low-constraint equivalent warpage residuals, and then weighting and summing these values ​​with preset weights to obtain the PVC panel warpage identification value, includes: The overall warping amplitude is calculated based on the maximum and minimum values ​​of the low-constraint equivalent warping residuals at all sampling locations. The edge warping amplitude is calculated based on the absolute value of the low-constraint equivalent warping residuals at the boundary sampling locations. The twist amplitude is calculated based on the low-constraint equivalent warping residuals at the four corner locations. The underestimation ratio of support flattening is determined by comparing the maximum absolute value of the equivalent warp residual under low constraint with the measured warp residual under support conditions. The warping degree of PVC panels is identified by a weighted combination of the overall warping amplitude, edge warping amplitude, twisting amplitude, and the underestimation ratio of support flattening.

9. The method for identifying the degree of warping of PVC panels based on intelligent sensors according to claim 1, characterized in that: The process of comparing various warpage amplitude and warpage degree identification values ​​with allowable thresholds to obtain the excess ratio and determine the warpage level, and combining the underestimation ratio of support flattening to output a support flattening risk warning as the final detection result, includes: Obtain the overall warping allowable threshold, edge warping allowable threshold, twist allowable threshold, and comprehensive warping degree identification threshold for the corresponding batch of PVC panels to be tested. Calculate the ratio of the overall warping amplitude, edge warping amplitude, twist amplitude, and the PVC panel warping degree identification value to the corresponding threshold, and use the maximum value as the excess ratio. Based on the proportion of warping exceeding the limit, the output indicates whether the warping is qualified, slightly warped, moderately warped, or severely warped. Based on the proportion of underestimation of support flattening, the output indicates whether the support flattening risk is low, medium, or high. The final test result includes the warping level of the PVC panel and the support flattening risk warning.

Images