Surface deformation high-precision detection method and system suitable for rolling mill sheet

Abstract

The application relates to the technical field of rolling mill sheet deformation detection, in particular to a high-precision surface deformation amount detection method and system suitable for rolling mill sheets, which comprises the following steps: acquiring the laser distance of each scanning point under each scanning path on the surface of the rolling mill sheet; calculating the high-frequency fluctuation degree of each scanning path; calculating the spatial consistency degree of each scanning path, combining the high-frequency fluctuation degree to obtain the vibration evaluation value of each scanning path; determining the interference coefficient of each scanning point; screening all the scanning points to obtain each significant point; obtaining the deformation evaluation value of the surface of the rolling mill sheet to evaluate the surface deformation amount of the rolling mill sheet. The application can effectively reduce the deformation amount detection error caused by the mechanical vibration or surface pollution residue of the sheet, reduce the misjudgment or missed judgment of the surface of the rolling mill sheet, and improve the detection precision of the surface deformation amount of the rolling mill sheet.

Description

TECHNICAL FIELD

[0001] The application relates to a rolling mill sheet deformation detection technology field, in particular to a surface deformation high-precision detection method and system suitable for a rolling mill sheet. BACKGROUND

[0002] The rolling mill sheet is a kind of plate material rolled by a rolling mill from a metal blank to a thinner specification. Due to the interference of various factors such as rolling pressure, roll crown, cooling uniformity and the like in the production process of the rolling mill sheet, the sheet is prone to uneven deformation, edge wave and other quality problems. The surface deformation detection of the rolling mill sheet can timely screen unqualified products, and meet the use demand of high-quality sheets.

[0003] When the surface deformation of the rolling mill sheet is detected by the laser displacement sensor, the vibration of the sheet caused by the rolling mill sheet processing process can cause the reflected signal received by the laser displacement sensor to be unstable, so that the deformation measurement of the rolling mill sheet fluctuates. At the same time, the pollutants remaining on the surface of the sheet can affect the reflection efficiency of the laser, resulting in a larger measurement error in the local area, causing great interference to the surface deformation detection of the rolling mill sheet, and affecting the precision of the deformation detection of the rolling mill sheet. SUMMARY

[0004] In order to solve the above technical problems, the surface deformation high-precision detection method and system suitable for the rolling mill sheet are provided to solve the existing problems.

[0005] The technical problem of the application is solved by providing a surface deformation high-precision detection method and system suitable for a rolling mill sheet, comprising the following steps:

[0006] In the first aspect, the application provides a surface deformation high-precision detection method suitable for a rolling mill sheet, which comprises the following steps:

[0007] Obtain the laser distance of each scanning point under each scanning path on the surface of the rolling mill sheet;

[0008] Analyze the high-frequency energy distribution of the laser distance of different scanning points under each scanning path in the frequency domain, and the discrete condition of the periodic change of the laser distance, and calculate the high-frequency fluctuation degree of each scanning path;

[0009] Calculate the spatial consistency degree of each scanning path through the interval distance between each scanning path and different scanning paths in its neighborhood range, and the correlation condition of the laser distance of all scanning points, and obtain the vibration evaluation value of each scanning path in combination with the high-frequency fluctuation degree;

[0010] The interference coefficient of each scanning point is determined by analyzing the deviation of laser distance between each scanning point and the rest of the scanning points in its neighborhood, and the difference in gradient distribution of laser distance in the neighborhood of each scanning point;

[0011] Based on the interference coefficient, all scanning points are screened to obtain each significant point. According to the dispersion of laser distance of all scanning points on the surface of the rolling mill sheet, the position distribution and the interference coefficient of all significant points, and in combination with the vibration evaluation value, a deformation evaluation value of the surface of the rolling mill sheet is obtained to evaluate the surface deformation of the rolling mill sheet.

[0012] Preferably, the high-frequency fluctuation degree of each scanning path is calculated, including:

[0013] The first dispersion degree is calculated by analyzing the dispersion of the difference in extreme value of laser distance of all scanning points in each scanning path;

[0014] The frequency domain analysis is performed on the laser distance of all scanning points in each scanning path to obtain a frequency spectrum. The frequency components greater than a preset frequency in the frequency spectrum are recorded as high-frequency components. The dispersion degree of the energy of all high-frequency components in the frequency spectrum is recorded as the second dispersion degree.

[0015] The high-frequency fluctuation degree is the product of the first dispersion degree and the second dispersion degree.

[0016] Preferably, the first dispersion degree is calculated, including:

[0017] The maximum and minimum values of the laser distance of all scanning points in each scanning path are obtained. The difference between each maximum value and the adjacent minimum value after it is recorded as the difference amount.

[0018] The dispersion degree of all difference amounts in each scanning path is calculated and recorded as the first dispersion degree.

[0019] Preferably, the spatial consistency degree of each scanning path is calculated, including:

[0020] The plurality of scanning paths adjacent to each scanning path are recorded as adjacent paths. The correlation degree of the laser distance of all scanning points between each scanning path and its adjacent paths is calculated.

[0021] The vertical distance between each scanning path and its adjacent paths is calculated. The absolute values of the correlation degree of each scanning path and all adjacent paths are weighted and summed with the inverse of the vertical distance as the weight to obtain the spatial consistency degree of each scanning path.

[0022] Preferably, the vibration evaluation value is the normalized result of the ratio of the spatial consistency degree to the high-frequency fluctuation degree.

[0023] Preferably, the determining of the interference coefficient of each scanning point comprises:

[0024] Taking each scanning point as a pixel point and the laser distance of each scanning point as a pixel value, a gray-scale image is generated based on the laser distances of all scanning points on the surface of the rolling sheet;

[0025] A local window of a preset size is constructed with any scanning point in the gray-scale image as the center, and the average value of the laser distances of all scanning points in the local window is calculated; the difference between the laser distance of the any scanning point and the average value is recorded as a relative deviation;

[0026] The gradient of the any scanning point in the gray-scale image in each direction is calculated, and the cumulative sum of the difference of the gradients corresponding to any two directions of the any scanning point is calculated;

[0027] The product of the relative deviation and the cumulative sum is taken as the interference coefficient of each scanning point.

[0028] Preferably, the obtaining process of the each significant point comprises: obtaining a segmentation threshold of the interference coefficient of all scanning points on the surface of the rolling sheet; selecting all scanning points with the interference coefficient greater than or equal to the segmentation threshold as each significant point.

[0029] Preferably, the obtaining process of the each significant point comprises: obtaining a segmentation threshold of the interference coefficient of all scanning points on the surface of the rolling sheet; selecting all scanning points with the interference coefficient greater than or equal to the segmentation threshold as each significant point.

[0030] The dispersion degree of the laser distances of all scanning points on the surface of the rolling sheet is calculated and recorded as a third dispersion degree;

[0031] The two-dimensional information entropy of the position coordinates of the positions of all significant points on the surface of the rolling sheet is calculated, the sum of the interference coefficients of all significant points on the surface of the rolling sheet is calculated and recorded as a first sum value, the sum of the vibration evaluation values of all scanning paths on the surface of the rolling sheet is calculated and recorded as a second sum value, and the product value of the three between the two-dimensional information entropy, the first sum value and the second sum value is calculated;

[0032] The deformation evaluation value is the normalized result of the ratio of the third dispersion degree and the product value.

[0033] Preferably, the evaluating of the surface deformation of the rolling sheet comprises: if the deformation evaluation value is greater than or equal to a preset threshold, the surface of the rolling sheet has deformation, otherwise, the surface of the rolling sheet has no deformation.

[0034] In a second aspect, the embodiments of the present application further provide a high-precision surface deformation amount detection system for a rolling mill sheet, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor implements the steps of the high-precision surface deformation amount detection method for the rolling mill sheet according to any one of the above aspects when executing the computer program.

[0035] The present application has at least the following beneficial effects:

[0036] The present application calculates the high-frequency fluctuation degree of each scanning path by analyzing the high-frequency energy concentration distribution of the laser distance of different scanning points under each scanning path in the frequency domain and the periodic fluctuation of the laser distance, which has the beneficial effect of considering the fluctuation regularity of the laser distance under each scanning path and preliminarily evaluating the interference influence of the sheet vibration on the rolling mill sheet surface; then the spatial consistency degree of each scanning path is calculated through the correlation of the laser distance between different scanning paths, which has the beneficial effect of considering that the fluctuation of the laser distance in the neighborhood range presents strong spatial consistency, reflecting the influence of the fluctuation of the laser distance between different scanning paths on the sheet vibration interference, obtaining the vibration evaluation value of each scanning path, which has the beneficial effect of comprehensively evaluating the fluctuation of the laser distance under the interference influence of the mechanical vibration or conveying vibration of the rolling mill; determining the interference coefficient of each scanning point, which has the beneficial effect of the significant laser distance mutation and the randomness of the gradient direction at the scanning point, reflecting the interference degree of the pollution residues on the laser distance measurement at the scanning point and evaluating the influence of the pollution residues on the laser distance measurement; screening all scanning points to obtain each significant point, which has the beneficial effect of the position of the laser distance mutation, so as to subsequently evaluate the dispersion of the position distribution of the significant point; further, obtaining the deformation evaluation value of the rolling mill sheet surface and evaluating the surface deformation amount of the rolling mill sheet, which has the beneficial effect of detecting the surface deformation amount of the rolling mill sheet by considering the interference influence of the sheet vibration and the interference influence of the pollution residues, which can effectively reduce the deformation amount detection error caused by the mechanical vibration of the sheet or the surface pollution residues, reduce the misjudgment or omission of the surface of the rolling mill sheet, and improve the detection accuracy of the surface deformation amount of the rolling mill sheet. BRIEF DESCRIPTION OF DRAWINGS

[0037] The high-precision surface deformation amount detection method for a rolling mill sheet according to the present application will be further described in detail below with reference to the accompanying drawings.

[0038] Figure 1 The step flow chart of the high-precision surface deformation amount detection method for a rolling mill sheet provided by the embodiments of the present application is shown in the following table:

[0039] Figure 2A step flowchart of a method for obtaining a deformation evaluation value is provided in the embodiments of the present application. DETAILED DESCRIPTION

[0040] For the purpose, technical solutions and advantages of the present application to be more clearly and intelligibly understood, the method and system for high-precision detection of surface deformation of a rolled sheet are further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0041] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0042] Referring to Figure 1 , a step flowchart of a method for high-precision detection of surface deformation of a rolled sheet is shown, which comprises the following steps:

[0043] Step 1: Obtain the laser distance of each scanning point under each scanning path on the surface of the rolled sheet.

[0044] A guide rail is arranged above the rolled sheet, a laser displacement sensor is deployed on the guide rail, and the laser displacement sensor is translated at a constant speed along the rolling direction of the sheet to perform scanning, i.e., the X-axis is first continuously scanned, and the spacing between adjacent scanning points is set to 2 cm. As an alternative, the implementer can set it according to the actual situation; after a row of scanning is completed, the laser displacement sensor moves 2 cm downward along the direction perpendicular to the rolling direction, i.e., the Y-axis, and then continuously scans along the X-axis. Thus, all scanning points of the laser displacement sensor scanned along a row are recorded as a scanning path, and the vertical distance between two adjacent scanning paths is 2 cm. Therefore, the laser displacement sensor measures the distance between each scanning point on the surface of the rolled sheet and the sensor by emitting a laser beam and receiving reflected light using the principle of optics, and then obtains the laser distance of each scanning point under each scanning path.

[0045] In this embodiment, in order to ensure the spatial consistency of the data, the laser distance of each scanning point on the rolled sheet obtained by the laser displacement sensor at different times is aligned in the same coordinate system, and a bilinear interpolation method is used for missing value filling processing. The bilinear interpolation method is a known technology and will not be described here. As an alternative, the implementer can use other methods of existing technology, such as mean filling method, etc., and the present embodiment does not make special restrictions on this.

[0046] At this point, the laser distance of each scanning point under each scanning path on the surface of the rolled sheet is obtained.

[0047] Step 2, analyze the high-frequency energy distribution of the laser distance of different scanning points under each scanning path in the frequency domain, and the discrete situation of the periodic change of the laser distance, calculate the high-frequency fluctuation degree of each scanning path; calculate the spatial consistency degree of each scanning path through the interval distance between each scanning path and different scanning paths in its neighborhood range, and the correlation of the laser distances of all scanning points, and obtain the vibration evaluation value of each scanning path by combining the high-frequency fluctuation degree.

[0048] When detecting the deformation of the surface of the rolled sheet, there is a relatively obvious difference between the fluctuation of the laser distance caused by the vibration of the rolled sheet and the fluctuation of the laser distance caused by the real deformation or defects on the rolled sheet, wherein the vibration of the rolled sheet is mainly affected by the mechanical vibration of the rolling mill and the vibration generated in the conveying process of the rolled sheet, and such vibration has strong periodicity. Therefore, when the deformation of the surface of the rolled sheet is more likely to be caused by the vibration of the rolled sheet, the laser distance will present stronger periodic high-frequency fluctuation, the regularity of the peak and valley of the laser distance will be stronger, and the energy distribution of the laser distance in the frequency domain will be concentrated in the mechanical vibration frequency band due to the influence of the mechanical high-frequency vibration. Meanwhile, in the two-dimensional plane of the surface of the rolled sheet, the fluctuation of the laser distance in the neighborhood range presents strong spatial consistency, the correlation of the laser distance changes at adjacent positions is relatively large, and the vibration of the rolled sheet forms obvious spatial domain correlation.

[0049] Based on the above analysis, the peak and valley change of the laser distance of different scanning points under each scanning path and the high-frequency distribution in the frequency domain are analyzed, and the high-frequency fluctuation degree is calculated to evaluate the periodic high-frequency fluctuation change of each scanning path, specifically as follows:

[0050] The maximum and minimum values of the laser distance of all scanning points under each scanning path are obtained.

[0051] In this embodiment, the AMPD algorithm (Automatic multiscale-based peak detection) is used to obtain the maximum and minimum values, wherein the AMPD algorithm is a known technology, and will not be described here. As other embodiments, the implementer can use other methods of the prior art, for example, a wave peak and wave trough second-order difference identification algorithm, and the present embodiment does not specially limit this.

[0052] The difference between each maximum value and the adjacent minimum value after it is recorded as a difference amount, and the discrete degree of all the difference amounts under each scanning path is calculated as a first discrete degree.

[0053] In the embodiment, the dispersion degree is measured by calculating the approximate entropy of all the difference amounts under each scanning path, wherein the calculation of the approximate entropy is a known technology and will not be described here again. As other embodiments, the implementer can use other methods in the prior art, such as variance, standard deviation, etc., and the embodiment does not specially limit this.

[0054] The laser distances of all the scanning points under each scanning path are subjected to frequency domain analysis to obtain a frequency spectrum diagram.

[0055] In the embodiment, the fast Fourier transform is used to obtain the frequency spectrum diagram, wherein the fast Fourier transform is a known technology and will not be described here again. As other embodiments, the implementer can use other methods in the prior art, such as discrete Fourier transform, Hilbert Huang transform, etc., and the embodiment does not specially limit this.

[0056] The frequency component greater than a preset frequency in the frequency spectrum diagram is recorded as a high-frequency component.

[0057] In the embodiment, the preset frequency is 50 Hz, and as other embodiments, the implementer can set it according to the actual situation.

[0058] The dispersion degree of the energy of all the high-frequency components in the frequency spectrum diagram is calculated and recorded as a second dispersion degree.

[0059] In the embodiment, the dispersion degree is measured by calculating the variance of the energy of all the high-frequency components in the frequency spectrum diagram, and as other embodiments, the implementer can use other methods in the prior art, such as standard deviation, etc., and the embodiment does not specially limit this.

[0060] The product of the first dispersion degree and the second dispersion degree is taken as the high-frequency fluctuation degree of each scanning path.

[0061] It should be noted that the smaller the first dispersion degree is, the higher the periodicity of the laser distance of the rolling mill sheet surface under the scanning path is, and the more significant the regular change of the fluctuation of the laser distance is. The smaller the second dispersion degree is, the more concentrated the high-frequency energy distribution of the laser distance in the frequency domain is, and the smaller the high-frequency fluctuation degree is, the more significant the periodic high-frequency fluctuation change of each scanning path is, and the greater the influence of the interference of the sheet vibration on the rolling mill sheet surface is.

[0062] Secondly, the similar change of the laser distance of all the scanning points between adjacent scanning paths is analyzed, and the spatial consistency degree is calculated, which is specifically:

[0063] The plurality of scanning paths adjacent to each scanning path is recorded as adjacent paths.

[0064] In the embodiment, 10 scanning paths adjacent to each scanning path are recorded as adjacent paths, that is, all scanning paths with a vertical distance of 10 cm from each scanning path are selected. Alternatively, the implementer can set it according to the actual situation.

[0065] The correlation degree of the laser distance of all scanning points between each scanning path and its adjacent paths is calculated.

[0066] In the embodiment, the correlation degree is calculated by calculating the cosine similarity of the laser distance of all scanning points between each scanning path and its adjacent paths. The calculation of the cosine similarity is a known technology and will not be described here. Alternatively, the implementer can use other methods of prior art, such as Pearson correlation coefficient method, and the embodiment does not make special limitations.

[0067] The vertical distance between each scanning path and its adjacent paths is calculated. The absolute values of the correlation degrees of each scanning path and all adjacent paths are weighted and summed with the inverse of the vertical distance as the weight to obtain the spatial consistency degree of each scanning path.

[0068] It should be noted that the greater the absolute value of the correlation degree, the higher the correlation change of the laser distance between the adjacent two scanning paths. Based on the vertical distance, the adjacent paths closer to each scanning path are given higher weights to evaluate the fluctuation correlation of the laser distance. The greater the spatial consistency degree, the stronger the spatial consistency of the fluctuation of the laser distance in the neighborhood range in the two-dimensional plane of the surface of the rolling mill sheet, and the more likely the fluctuation of the laser distance is caused by sheet vibration interference.

[0069] Further, based on the high-frequency fluctuation degree and the spatial consistency degree, a vibration evaluation value is determined to reflect the abnormal degree of the laser distance caused by sheet processing vibration in each line scanning path on the surface of the rolling mill sheet. Specifically:

[0070] The normalized result of the ratio of the spatial consistency degree to the high-frequency fluctuation degree is taken as the vibration evaluation value of each scanning path.

[0071] In the embodiment, sigmoid function is used for normalization processing. The sigmoid function is a known technology and will not be described here. Alternatively, the implementer can use other methods of prior art, such as softmax function, tanh function, and the embodiment does not make special limitations.

[0072] It should be noted that the greater the vibration evaluation value, the greater the influence of the fluctuation of the laser distance on the scanning path caused by the interference of the rolling mill mechanical vibration or the conveying vibration.

[0073] At this point, the vibration evaluation value of each scanning path is obtained.

[0074] Step 3, analyze the deviation of the laser distance between each scanning point and the rest of the scanning points in its neighborhood, and the difference in the gradient distribution of the laser distance in the neighborhood at each scanning point, to determine the interference coefficient of each scanning point.

[0075] Further, in the process of detecting the deformation of the rolling mill sheet surface, the oil stains, dust and various pollutants remaining on the sheet surface have a significant impact on the laser reflection efficiency. In the actual detection of the deformation of the rolling mill sheet surface, the pollutants will cause the laser reflection signal in the local area of the sheet surface to be abnormal, so that the laser distance is more likely to deviate from the actual value, and then it may be misjudged as a real deformation or defect area, increasing the error of the sheet deformation detection, affecting the accuracy and reliability of the detection result. Secondly, when there is residual interference of pollutants on the surface of the rolling mill sheet, the more serious the local area laser distance high amplitude mutation caused by the residual pollutants on the surface of the rolling mill sheet, and in the two-dimensional plane of the rolling mill sheet surface, the higher the randomness of the gradient direction of the laser distance of different scanning points in the local area. Therefore, by analyzing the deviation of the laser distance of different scanning points in the local area, and the gradient direction change, the interference coefficient is calculated, which is specifically:

[0076] Taking each scanning point as a pixel point and the laser distance of each scanning point as a pixel value, a gray image is generated based on the laser distance of all scanning points on the surface of the rolling mill sheet;

[0077] Taking any scanning point in the gray image as the center, a local window of a predetermined size is constructed, and the average value of the laser distance of all scanning points in the local window is calculated;

[0078] In this embodiment, the size of the local window is 5x5, and as other embodiments, the implementer can set it according to the actual situation.

[0079] The difference between the pixel value of the any scanning point in the gray image and the average value is calculated, which is denoted as the relative deviation;

[0080] In this embodiment, the absolute value of the difference between the laser distance of the any scanning point in the gray image and the average value is calculated, which is denoted as the relative deviation;

[0081] The gradient of the any scanning point in the gray image in each direction is calculated.

[0082] In the embodiment, the Laplacian operator is used to calculate the gradient in each direction, and the gradient in 8 directions is included. The Laplacian operator is a known technology, and will not be described here. As other embodiments, the implementer can use other methods in the prior art, such as the sobel operator, and the embodiment does not make special restrictions.

[0083] The accumulation sum of the difference of the gradients corresponding to any two directions of any scanning point in the gray-scale image is calculated.

[0084] In the embodiment, the accumulation sum of the absolute value of the difference of the gradients corresponding to any two directions of any scanning point in the gray-scale image is calculated.

[0085] The product of the relative deviation and the accumulation sum is taken as the interference coefficient of each scanning point.

[0086] It should be noted that the greater the relative deviation, the greater the difference between the laser distance of the scanning point and the scanning points in the surrounding area, indicating that there is a more significant laser distance mutation at this point, which is likely to be caused by the local change of the laser reflection characteristics due to the pollutant residue. The greater the accumulation sum, the greater the difference between the gradients in different directions in the local window, and the higher the randomness of the gradient direction, reflecting that the change direction of the laser distance at the scanning point in the surrounding area is more chaotic and lacks consistency. The greater the interference coefficient, the more significant the laser distance mutation and the higher the gradient direction randomness at the scanning point, reflecting that the degree of interference of the pollutant residue on the laser distance measurement at the scanning point is greater, and the influence of the pollutant residue on the laser distance at this point is more serious, which is more likely to cause the reliability and accuracy of the laser distance to decrease.

[0087] At this point, the interference coefficient of each scanning point is obtained.

[0088] Step 4, based on the interference coefficient, all scanning points are screened to obtain each significant point; according to the dispersion of the laser distance of all scanning points on the surface of the rolling sheet, the position distribution and the interference coefficient of all significant points, and in combination with the vibration evaluation value, a deformation evaluation value of the surface of the rolling sheet is obtained, and the surface deformation of the rolling sheet is evaluated.

[0089] Further, based on the interference coefficient, all scanning points are screened, specifically:

[0090] A segmentation threshold of the interference coefficient of all scanning points on the surface of the rolling sheet is obtained.

[0091] In the embodiment, the threshold segmentation algorithm of Otsu is used to obtain the segmentation threshold, which is a known technology and will not be described here. As other embodiments, the implementer can use other methods of prior art, such as cross-validation method, and the embodiment does not make special restrictions on this.

[0092] All scanning points with the interference coefficient greater than or equal to the segmentation threshold are selected and recorded as each salient point.

[0093] Secondly, the randomness of the residual position of the surface contaminants of the rolling sheet is high, so when the spatial distribution of different salient points in the two-dimensional plane of the rolling sheet is more random, it is more likely that the fluctuation of the laser distance at the salient point is caused by the residual contaminants on the surface of the rolling sheet, and the abnormal condition of the laser distance caused by the residual contaminants is more serious, which is more likely to be misjudged as the deformation or defect of the rolling sheet, thereby affecting the detection accuracy of the deformation amount of the rolling sheet.

[0094] Based on the above analysis, the deformation evaluation value is calculated through the vibration evaluation value, the interference coefficient and the distribution dispersion of the salient point, and specifically:

[0095] The dispersion degree of the laser distance of all scanning points on the surface of the rolling sheet is calculated and recorded as the third dispersion degree.

[0096] In the embodiment, the variance of the laser distance of all scanning points on the surface of the rolling sheet is used to measure the dispersion degree, and as other embodiments, the implementer can use other methods of prior art, such as standard deviation, and the embodiment does not make special restrictions on this.

[0097] The two-dimensional information entropy of the position coordinates of all salient points on the surface of the rolling sheet is calculated.

[0098] It should be noted that the calculation of the two-dimensional information entropy is a known technology and will not be described here.

[0099] The sum of the interference coefficients of all salient points on the surface of the rolling sheet is calculated and recorded as the first sum value.

[0100] The sum of the vibration evaluation values of all scanning paths on the surface of the rolling sheet is calculated and recorded as the second sum value.

[0101] The product value of the two-dimensional information entropy, the first sum value and the second sum value is calculated, and the normalized result of the ratio of the third dispersion degree to the product value is used as the deformation evaluation value of the surface of the rolling sheet.

[0102] In the embodiment, the calculation formula of the deformation evaluation value of the surface of the rolling sheet is:

[0103]

[0104] Wherein, B is a deformation evaluation value of the rolling mill sheet surface, σ is the third dispersion, H is the two-dimensional information entropy, G q is a disturbance coefficient of the qth significant point in the rolling mill sheet surface, Z k is a vibration evaluation value of the qth scanning path in the rolling mill sheet surface, Q is the number of all significant points in the rolling mill sheet surface, K is the number of all scanning paths in the rolling mill sheet surface, norm is a normalization function, in this embodiment, a sigmoid function is used for normalization processing, wherein the sigmoid function is a known technology and will not be described here, as other embodiments, the implementer can use other methods of prior art, for example, softmax function, tanh function, etc., this embodiment does not make special restrictions, wherein, is a first sum, is a second sum.

[0105] It should be noted that the greater the third dispersion, the greater the difference in laser distance at different scanning points, the smaller the two-dimensional information entropy, the lower the randomness of the spatial distribution of different significant points, the more likely the distribution is more dense, the smaller the second sum, the smaller the interference of the rolling mill sheet surface by the rolling mill mechanical vibration or conveying vibration, the greater the possibility of deformation, the smaller the first sum, the smaller the interference degree of the laser distance anomaly at the significant point of the rolling mill sheet surface by the contamination residue, and the higher the possibility of the existence of real deformation of the rolling mill sheet surface, and the greater the obtained deformation evaluation value; wherein the step flow chart of the deformation evaluation value acquisition method provided by the embodiment of the present application is as shown in Figure 2 .

[0106] Further, based on the deformation evaluation value, the surface deformation amount of the rolling mill sheet is evaluated, specifically:

[0107] If the deformation evaluation value is greater than or equal to a preset threshold value, the surface of the rolling mill sheet exists deformation, otherwise, the surface of the rolling mill sheet does not exist deformation;

[0108] In this embodiment, the preset threshold value is 0.6, as other embodiments, the implementer can set it according to the actual situation.

[0109] Based on the same inventive concept as the above method, the embodiment of the present application also provides a surface deformation amount high-precision detection system suitable for rolling mill sheet, which comprises a memory, a processor and a computer program stored in the memory and running on the processor, and the processor executes the computer program to realize the steps of any one of the above-mentioned surface deformation amount high-precision detection methods suitable for rolling mill sheet.

[0110] It should be understood that, although Figure 1The steps in the flowcharts are shown in sequence according to the arrows, but the steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, the steps are not strictly limited in sequence, and the steps can be executed in other sequences. Moreover, Figure 1 At least one part of the steps in the flowcharts can include a plurality of sub-steps or a plurality of stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence of the sub-steps or stages is not necessarily sequential, but can be executed in rotation or alternation with at least one part of other steps or sub-steps or stages of other steps.

[0111] The technical features of the above embodiments can be combined in any manner. For brevity, not all possible combinations of the technical features in the above embodiments are described, but as long as the combinations of the technical features do not contradict each other, they should be considered within the scope of the present disclosure.

[0112] The above embodiments only express several implementation manners of the present application, and the description is relatively specific and detailed, but should not be construed as a limitation on the present application. It should be noted that, for those skilled in the art, several modifications and improvements can be made without departing from the concept of the present application, therefore, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application, without departing from the technical solution of the present application, should be considered within the protection scope of the present application.

Claims

1. A method for high-precision detection of surface deformation of a sheet produced by a rolling mill, characterized in that, The method comprises the following steps: Obtaining the laser distance of each scanning point under each scanning path on the surface of the rolling mill sheet; Analyzing the high-frequency energy distribution of the laser distance of different scanning points under each scanning path in the frequency domain, and the discrete condition of the periodic change of the laser distance, and calculating the high-frequency fluctuation degree of each scanning path; Calculating the spatial consistency degree of each scanning path through the interval distance between each scanning path and different scanning paths in its neighborhood range, and the correlation condition of the laser distance of all scanning points, and obtaining the vibration evaluation value of each scanning path in combination with the high-frequency fluctuation degree; Analyzing the deviation condition of the laser distance between each scanning point and the remaining scanning points in its neighborhood, and the gradient distribution difference condition of the laser distance at each scanning point in the neighborhood, and determining the interference coefficient of each scanning point; Based on the interference coefficient, screening all scanning points to obtain each significant point; obtaining the deformation evaluation value of the surface of the rolling mill sheet according to the discrete condition of the laser distance of all scanning points on the surface of the rolling mill sheet, and the position distribution condition and the interference coefficient of all significant points, and in combination with the vibration evaluation value, and evaluating the surface deformation of the rolling mill sheet.

2. The method for detecting surface deformation of a sheet for a rolling mill according to claim 1, wherein The calculation of the high-frequency fluctuation degree of each scanning path comprises: Calculating a first dispersion degree by analyzing the discrete condition of the difference change of the extreme value of the laser distance of all scanning points under each scanning path; Performing frequency domain analysis on the laser distance of all scanning points under each scanning path to obtain a frequency spectrum diagram; regarding the frequency components greater than a preset frequency in the frequency spectrum diagram as high-frequency components; calculating the discrete degree of the energy of all high-frequency components in the frequency spectrum diagram as a second dispersion degree; The high-frequency fluctuation degree is the product of the first dispersion degree and the second dispersion degree.

3. The method for detecting surface deformation of a sheet for a rolling mill according to claim 2, wherein The calculation of the first dispersion degree comprises: Obtaining the maximum value and the minimum value of the laser distance of all scanning points under each scanning path; regarding the difference value between each maximum value and the adjacent minimum value after the maximum value as a difference amount; Calculating the discrete degree of all difference amounts under each scanning path as the first dispersion degree.

4. The method for detecting surface deformation of a sheet for a rolling mill according to claim 1, wherein The calculation of the spatial consistency degree of each scanning path comprises: Regarding a plurality of scanning paths adjacent to each scanning path as adjacent paths; calculating the correlation degree of the laser distance of all scanning points between each scanning path and each adjacent path thereof; Calculating the vertical distance between each scanning path and each adjacent path thereof; performing weighted summation on the absolute values of the correlation degrees of each scanning path and all adjacent paths thereof with the inverse of the vertical distance as the weight to obtain the spatial consistency degree of each scanning path.

5. The method for detecting surface deformation of a sheet for a rolling mill according to claim 1, wherein The vibration evaluation value is the normalized result of the ratio of the spatial consistency degree to the high-frequency fluctuation degree.

6. The method for detecting surface deformation of a sheet for a rolling mill according to claim 1, wherein The determination of the interference coefficient of each scanning point comprises: Generating a gray-scale image based on the laser distance of all scanning points on the surface of the rolling mill sheet, with each scanning point as a pixel point and the laser distance of each scanning point as a pixel value; Regarding any scanning point in the gray-scale image as the center, constructing a local window of a preset size, calculating the average value of the laser distance of all scanning points in the local window; regarding the difference between the laser distance of the any scanning point and the average value as a relative deviation; calculating the gradient of any scanning point in the gray image in each direction; calculating the accumulated sum of the difference of the gradients corresponding to any two directions of the any scanning point; multiplying the relative deviation and the accumulated sum as the interference coefficient of each scanning point.

7. The method for detecting surface deformation of a sheet for a rolling mill according to claim 1, wherein The acquisition process of the significant points is: acquiring the segmentation threshold of the interference coefficient of all scanning points on the surface of the rolling sheet; selecting all scanning points with the interference coefficient greater than or equal to the segmentation threshold as the significant points.

8. The method for detecting surface deformation of a sheet for a rolling mill according to claim 1, wherein The deformation evaluation value of the surface of the rolling sheet is obtained by: calculating the dispersion degree of the laser distance of all scanning points on the surface of the rolling sheet as the third dispersion degree; calculating the two-dimensional information entropy of the position coordinates of the positions of all significant points on the surface of the rolling sheet; calculating the sum of the interference coefficients of all significant points on the surface of the rolling sheet as the first sum value; calculating the sum of the vibration evaluation values of all scanning paths on the surface of the rolling sheet as the second sum value; calculating the product value of the three between the two-dimensional information entropy, the first sum value and the second sum value; The deformation evaluation value is the normalized result of the ratio of the third dispersion degree and the product value.

9. The method for detecting surface deformation of a sheet for a rolling mill according to Claim 1, wherein The evaluation of the surface deformation of the rolling sheet includes: if the deformation evaluation value is greater than or equal to the preset threshold, the surface of the rolling sheet has deformation, otherwise, the surface of the rolling sheet has no deformation.

10. A high-precision surface deformation amount detection system for a rolled sheet, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, The processor executes the computer program to realize the steps of the high-precision detection method for the surface deformation of the rolling sheet according to any one of claims 1-9. The processor executes the computer program to realize the steps of the high-precision detection method for the surface deformation of the rolling sheet according to any one of claims 1-9.

Images