An intelligent positioning system for welding points of iron towers

Abstract

The present invention relates to the field of welding technology, and in particular to an intelligent positioning system for welding points of an iron tower, the system comprising: a first acquisition module for acquiring a grayscale image of a welding workpiece image; a feature extraction module for extracting feature points; an edge detection module for analyzing the relative position of feature points according to feature point gradients to obtain edge points; a positioning module for positioning welding points of a welding workpiece; a second acquisition module for acquiring marking point information and welding workpiece information; an abnormality analysis module for analyzing the positioning abnormality of welding points; an analysis optimization module for optimizing the analysis process of the positioning abnormality of welding points according to the acquired surface reflectivity of the welding workpiece and the surface roughness of the welding workpiece; and a machine learning module for updating the positioning analysis process of welding points in the next monitoring cycle. The present invention improves the accuracy of welding point positioning and simultaneously improves the welding quality of the iron tower.

Description

Technical Field

[0001] The invention relates to the technical field of welding, and in particular to an intelligent positioning system for welding points of an iron tower. Background Art

[0002] As the manufacturing industry develops towards intelligence and automation, the welding process, as one of the key technologies in the manufacturing field, has a direct impact on product quality and production efficiency through its automation level. At present, the traditional method of locating welding points usually relies on manual visual inspection or simple sensor assistance, which is not only inefficient but also difficult to meet the growing high-precision requirements. Especially in the welding process of complex structural parts, how to accurately and quickly locate welding points has become an urgent problem to be solved.

[0003] Chinese patent publication number CN115430956A discloses a welding point positioning method, system, device and storage medium, the welding point positioning method includes: obtaining a plane image of a target battery in response to battery arrival information, locating the coordinates of a calibration point in the plane image, and calculating the welding point position of the target battery according to the calibration point coordinates; obtaining the depth data of the target battery to determine the welding point depth, and generating welding parameters in combination with the welding point position and sending them to a welding device, so that the welding device performs corresponding welding operations on the target battery according to the welding parameters. The present invention calculates the welding point position and welding point depth in the target battery by combining two-dimensional detection and three-dimensional detection. Even if the welding point is blocked, the welding point can be accurately welded to avoid the occurrence of cold welding during the welding process, thereby improving the welding quality and welding accuracy of the iron tower; it can be seen that when locating the welding point, the scheme only analyzes the calibration point, and there is a problem of low welding point positioning accuracy. Summary of the invention

[0004] To this end, the present invention provides an intelligent positioning system for welding points of an iron tower, so as to overcome the problem of low positioning accuracy of welding points in the prior art.

[0005] To achieve the above object, the present invention provides an intelligent positioning system for welding points of an iron tower, the system comprising:

[0006] A first acquisition module is used to acquire a grayscale image of a welding workpiece image;

[0007] A feature extraction module is used to extract feature points according to the grayscale image of the acquired welding workpiece image;

[0008] An edge detection module is used to analyze the relative positions of feature points according to the gradient of feature points to obtain edge points;

[0009] A positioning module is used to extract the straight line features of the edge points, perform local peak detection on the straight line features of the edge points, and locate the welding points of the welding workpiece according to the results of the peak detection of the straight line features of the edge points;

[0010] The second acquisition module is used to acquire the marking point information and the welding workpiece information;

[0011] An abnormality analysis module is used to analyze the positioning abnormality of the welding point according to the positioning result of the welding point and the acquired marking point information;

[0012] An analysis and optimization module, used to optimize the analysis process of the abnormality of the positioning of the welding point based on the obtained reflectivity of the welding workpiece surface, and also used to optimize the analysis process of the abnormality of the positioning of the welding point based on the roughness of the welding workpiece surface;

[0013] The machine learning module is used to update the positioning analysis process of the welding points in the next monitoring period according to the analysis results of the abnormality of the positioning of the welding points in the monitoring period.

[0014] Furthermore, the feature extraction module compares the grayscale value G(x, y) of each pixel in the grayscale image of the welding workpiece image with each preset grayscale threshold, and extracts the feature point according to the comparison result, wherein:

[0015] When g1≤G(x,y)≤g2, the feature extraction module determines that the pixel point is a feature point;

[0016] When G(x,y)<g1 or G(x,y)>g2, the feature extraction module determines that the pixel point is a feature point;

[0017] Among them, g1 is the first preset grayscale value, g2 is the second preset grayscale value, x is the horizontal coordinate of the pixel point, and y is the vertical coordinate of the pixel point.

[0018] Furthermore, the edge detection module calculates the gradient amplitude F(x,y) of the feature point and sets F(x,y)=(Fx 2 +Fy 2 ) 0.5 ;

[0019] The edge detection module compares the gradient amplitude F(x, y) of the feature point with the preset amplitude f0, and analyzes the relative position of the feature point according to the comparison result, wherein:

[0020] When F(x,y)≤f0, the edge detection module determines that the feature point is a non-edge point;

[0021] When F(x,y)>f0, the edge detection module determines that the feature point is an edge point;

[0022] Among them, Fx is the gradient of the feature point (x, y) in the horizontal direction, and Fy is the gradient of the feature point (x, y) in the vertical direction.

[0023] Furthermore, the positioning module is provided with a feature analysis unit and a positioning unit, and the feature analysis unit is used to perform Hough transform on the edge points and extract straight line features from the edge points, and the calculation formula is:

[0024] ρ=Bx×cos(θ)+By×sin(θ)

[0025] Among them, ρi is the distance between the extracted i-th straight line and the origin, θi is the angle between the extracted i-th straight line and the x-axis, Bx is the horizontal coordinate of the edge point plane rectangular coordinate system, and By is the vertical coordinate of the edge point plane rectangular coordinate system.

[0026] Furthermore, the positioning unit is used to locate the welding point of the welding workpiece according to the result of the edge point straight line feature peak detection, and the positioning unit sets the coordinates of the welding point to (xw, yw), setting:

[0027] xw=[ρz2×sin(θz1)-ρz1 ×sin(θz2)] / [cos(θz2)×sin(θz1)-cos(θz1)×sin(θz2)];

[0028] yw=[ρz1×cos(θz2)-ρz2 ×cos(θz1)] / [cos(θz2)×sin(θz1)-cos(θz1)×sin(θz2)].

[0029] Further, the abnormality analysis module is provided with a calculation unit and an abnormality analysis unit, the calculation unit is used to calculate the abnormality index of the welding point positioning according to the positioning result of the welding point coordinates and the horizontal coordinate Jx and the vertical coordinate Jy of the marking point in the welding workpiece image, and the calculation unit sets the abnormality index of the welding point positioning to Yc, and sets: Among them, k0 is the preset error threshold.

[0030] Further, the abnormality analysis unit compares the welding point positioning abnormality index with the preset abnormality index y0, and analyzes the abnormality of the welding point positioning according to the comparison result, wherein:

[0031] When Yc≤y0, the abnormality analysis unit determines that the welding point positioning is normal;

[0032] When Yc>y0, the abnormality analysis unit determines that the welding point positioning is abnormal.

[0033] Furthermore, the analysis and optimization module is provided with a first optimization unit and a second optimization unit, wherein the first optimization unit is used to compare the obtained surface reflectivity s0 of the welding workpiece with the preset reflectivity s1, and optimize the analysis process of the abnormality of the welding point positioning according to the comparison result, wherein:

[0034] When s0≤s1, the first optimization unit determines that the surface reflectivity of the welding workpiece is normal and does not perform optimization;

[0035] When s0>s1, the first optimization unit determines that the surface reflectivity of the welding workpiece is abnormal, and optimizes the analysis process of the abnormality of the welding point positioning, and sets the optimized preset abnormality index to y0', setting: y0'=y0×{1+exp[3×(s0-s1) / (s0+s1)-3]}.

[0036] Furthermore, the second optimization unit is used to compare the obtained surface roughness c0 of the welding workpiece with the preset roughness c1, and perform secondary optimization on the analysis process of the abnormality of the welding point positioning according to the comparison result, wherein:

[0037] When c0≤c1, the second optimization unit determines that the surface roughness of the welding workpiece is normal and does not perform optimization;

[0038] When c0>c1, the second optimization unit determines that the surface roughness of the welding workpiece is abnormal, and performs secondary optimization on the analysis process of the abnormality of the welding point positioning, and sets the optimized preset reflectivity to s1', setting s1'=s1×{1+ln[(c0-c1) / (c0+c1)+1]}, when s1'≥s2, the value of s1' is s2, and s2 is the preset reflectivity threshold.

[0039] Furthermore, the machine learning module updates the positioning analysis process of the welding points in the next monitoring period with the analysis results of the abnormality of the positioning of the welding points in the monitoring period, wherein:

[0040] When r0 / r1≤E, the machine learning module determines that the positioning of the welding point in the current monitoring period is normal and does not update;

[0041] When r0 / r1>E, the machine learning module determines that the positioning of the welding point in the current monitoring period is abnormal, and updates the positioning analysis process of the welding point in the next monitoring period, and sets the updated preset amplitude to f0', setting f0'=f0×{1-sin[(r0 / r1-E)×(π / 2)]};

[0042] Among them, r0 is the number of abnormal welding point positioning within the monitoring period, r1 is the total number of welding point positioning within the monitoring period, and E is the preset abnormal proportion.

[0043] Compared with the prior art, the beneficial effects of the present invention are that the feature extraction module sets a preset gray value to improve the efficiency of pixel point extraction of the cross projection in the gray image of the welding workpiece, the edge detection module converts the straight line in the image space into a point in the parameter space through the Hough transform, simplifies the complexity of straight line detection, and improves the detection efficiency, the positioning module calculates the intersection coordinates of two straight lines detected by the Hough transform to improve the accuracy of welding point positioning, and at the same time, the calculation amount of coordinate extraction is relatively small, and the welding point coordinates can be obtained in real time, which improves the welding efficiency, and the abnormality analysis module sets a preset error threshold to improve the calculation accuracy of the welding point positioning abnormality index, thereby improving the welding abnormality The accuracy of the abnormality analysis is improved by setting a preset abnormality index to improve the accuracy of the welding positioning abnormality analysis, thereby improving the accuracy of the welding point positioning. The analysis optimization module sets a preset reflectivity to reduce the influence of the surface reflectivity of the welding workpiece on the welding point positioning abnormality analysis, thereby improving the accuracy of the welding positioning abnormality analysis. The preset roughness and preset reflectivity thresholds are set to reduce the influence of the surface roughness of the welding workpiece on the welding point positioning abnormality analysis, thereby improving the accuracy of the welding positioning abnormality analysis. The machine learning module sets a preset abnormality ratio to improve the accuracy of the welding positioning abnormality analysis, thereby improving the accuracy of the welding point positioning, and ultimately improving the welding efficiency and the tower welding quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 Schematic diagram of the structure of the intelligent positioning system for welding points of the iron tower according to this embodiment.

[0045] Figure 2 It is a structural diagram of the positioning module of this embodiment.

[0046] Figure 3 Schematic diagram of the structure of the abnormality analysis module in this embodiment.

[0047] Figure 4 This is a schematic diagram of the structure of the analysis and optimization module of this embodiment. DETAILED DESCRIPTION

[0048] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0049] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0050] It should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected" and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0051] See also Figure 1 As shown, it is a structural schematic diagram of the welding point intelligent positioning system for the iron tower in this embodiment, and the system includes:

[0052] A first acquisition module is used to acquire a grayscale image of the welding workpiece image. The first acquisition module establishes a plane rectangular coordinate system with the lower left corner pixel of the grayscale image of the welding workpiece image as the origin, the positive direction of the y-axis from bottom to top, and the positive direction of the x-axis from left to right, and the position of each pixel is represented by the coordinates of the pixel points; in this embodiment, no specific limitation is made on the acquisition method of the welding workpiece image, and those skilled in the art can freely set it, as long as the acquisition requirements of the welding workpiece image are met, wherein the welding workpiece image can be acquired through interaction; in this embodiment, no specific limitation is made on the acquisition method of the grayscale image, and those skilled in the art can freely set it, as long as the acquisition requirements of the grayscale image are met, wherein the image can be read and converted through the OpenCV library in Python, and the grayscale value of the pixel point can be acquired;

[0053] A feature extraction module, used for extracting feature points according to the acquired grayscale image of the welding workpiece image, the feature extraction module is connected to the first acquisition module;

[0054] An edge detection module is used to analyze the relative position of the feature point according to the feature point gradient to obtain the edge point. The edge detection module is connected to the feature extraction module. In this embodiment, the method for obtaining the feature point gradient is not specifically limited, and those skilled in the art can freely set it as long as the requirements for obtaining the feature point gradient are met. The feature point gradient can be obtained by applying a 3x3 convolution kernel through a Sobel edge detection algorithm;

[0055] A positioning module, used for extracting straight line features of edge points, performing local peak detection on the straight line features of edge points, and locating welding points of welding workpieces according to the results of the peak detection of the straight line features of edge points, the positioning module being connected to the edge detection module;

[0056] A second acquisition module is used to acquire the marking point information and the welding workpiece information, wherein the marking point information includes the horizontal coordinate and the vertical coordinate of the marking point in the welding workpiece image, and the welding workpiece information includes the surface reflectivity of the welding workpiece and the surface roughness of the welding workpiece. The second acquisition module is connected to the positioning module. In this embodiment, no specific limitation is made on the acquisition method of the marking point information and the welding workpiece information, and those skilled in the art can freely set it as long as the acquisition requirements of the marking point information and the welding workpiece information are met. The marking point information can be acquired through interaction, and the welding workpiece information can be acquired through equipment such as a spectrometer and a laser measuring instrument;

[0057] An abnormality analysis module, used for analyzing the positioning abnormality of the welding point according to the positioning result of the welding point and the acquired marking point information, the abnormality analysis module is connected to the second acquisition module;

[0058] An analysis and optimization module, used for optimizing the analysis process of the abnormality of the welding point positioning according to the obtained reflectivity of the welding workpiece surface, and for optimizing the analysis process of the abnormality of the welding point positioning according to the roughness of the welding workpiece surface, the analysis and optimization module is connected to the abnormality analysis module;

[0059] The machine learning module is used to update the positioning analysis process of the welding points in the next monitoring period according to the analysis results of the abnormality of the positioning of the welding points in the monitoring period. The machine learning module is connected to the analysis optimization module. In this embodiment, there is no specific limitation on the setting of the monitoring period, and technicians in this field can set it freely as long as the setting requirements of the monitoring period are met. Among them, the monitoring period can be set to 30 days, 40 days, etc.

[0060] Specifically, the welding point intelligent positioning system for the iron tower described in this embodiment is applied to the welding point positioning of the iron tower based on laser vision guidance. It performs edge detection on the feature area composed of feature points, extracts the edge of the laser cross line, performs Hough transform on the extracted edge, identifies two straight lines, and locates the welding point based on the parameters of the two straight lines identified by the Hough transform, thereby improving the accuracy of welding point positioning.

[0061] Specifically, the feature extraction module compares the grayscale value G(x, y) of each pixel in the grayscale image of the welding workpiece image with each preset grayscale threshold, and extracts the feature points according to the comparison result, wherein:

[0062] When g1≤G(x,y)≤g2, the feature extraction module determines that the pixel point is a feature point;

[0063] When G(x,y)<g1 or G(x,y)>g2, the feature extraction module determines that the pixel point is a feature point;

[0064] Among them, g1 is a first preset grayscale value, g2 is a second preset grayscale value, x is the horizontal coordinate of the pixel point, y is the vertical coordinate of the pixel point, and g1<g2.

[0065] Specifically, the feature extraction module sets a preset grayscale value to improve the efficiency of pixel extraction of the cross projection in the grayscale image of the welding workpiece, thereby improving the accuracy of edge point analysis, thereby improving the accuracy of welding point positioning, and ultimately improving welding efficiency and tower welding quality; in this embodiment, the setting of the preset grayscale value is not specifically limited, and technical personnel in this field can set it freely as long as the setting requirements of the preset grayscale value are met, among which the optimal value of g1 is 200, and the optimal value of g2 is 245.

[0066] Specifically, the edge detection module calculates the gradient amplitude F(x,y) of the feature point and sets F(x,y)=(Fx 2 +Fy 2 ) 0.5 ;

[0067] The edge detection module compares the gradient amplitude F(x, y) of the feature point with the preset amplitude f0, and analyzes the relative position of the feature point according to the comparison result, wherein:

[0068] When F(x,y)≤f0, the edge detection module determines that the feature point is a non-edge point;

[0069] When F(x,y)>f0, the edge detection module determines that the feature point is an edge point;

[0070] Among them, Fx is the gradient of the feature point (x, y) in the horizontal direction, and Fy is the gradient of the feature point (x, y) in the vertical direction.

[0071] Specifically, the edge detection module converts a straight line in the image space into a point in the parameter space through the Hough transform, simplifies the complexity of straight line detection, improves the detection efficiency, and improves the accuracy of edge point analysis by setting a preset amplitude, thereby improving the accuracy of edge point straight line feature extraction, thereby improving the accuracy of welding point positioning, and ultimately improving welding efficiency and tower welding quality; in this embodiment, the setting of the preset amplitude is not specifically limited, and technical personnel in this field can set it freely, as long as the setting requirements of the preset amplitude are met, where the optimal value of f0 is 100.

[0072] Specifically, the machine learning module updates the positioning analysis process of the welding points in the next monitoring period with the analysis results of the abnormality of the positioning of the welding points in the monitoring period, wherein:

[0073] When r0 / r1≤E, the machine learning module determines that the positioning of the welding point in the current monitoring period is normal and does not update;

[0074] When r0 / r1>E, the machine learning module determines that the positioning of the welding point in the current monitoring period is abnormal, and updates the positioning analysis process of the welding point in the next monitoring period, and sets the updated preset amplitude to f0', setting f0'=f0×{1-sin[(r0 / r1-E)×(π / 2)]};

[0075] Among them, r0 is the number of abnormal welding point positioning within the monitoring period, r1 is the total number of welding point positioning within the monitoring period, and E is the preset abnormal proportion.

[0076] Specifically, the machine learning module improves the accuracy of welding positioning abnormality analysis by setting a preset abnormality ratio, thereby improving the accuracy of welding point positioning, and ultimately improving welding efficiency and tower welding quality; in this embodiment, the setting of the preset abnormality ratio is not specifically limited, and technical personnel in this field can set it freely, as long as the setting requirements of the preset abnormality ratio are met, where the optimal value of E is 0.03.

[0077] See also Figure 2 As shown, it is a schematic diagram of the structure of the positioning module of this embodiment, and the positioning module includes:

[0078] A feature analysis unit, used to extract straight line features of edge points and perform local peak detection on the straight line features of edge points;

[0079] The positioning unit is used to locate the welding point of the welding workpiece according to the result of the edge point straight line feature peak detection, and the positioning unit is connected to the feature analysis unit.

[0080] Specifically, the feature analysis unit performs Hough transform on the edge points and extracts straight line features from the edge points. The calculation formula is:

[0081] ρ=Bx×cos(θ)+By×sin(θ)

[0082] Among them, ρi is the distance between the extracted i-th straight line and the origin, θi is the angle between the extracted i-th straight line and the x-axis, Bx is the abscissa of the rectangular coordinate system of the edge point plane, and By is the ordinate of the rectangular coordinate system of the edge point plane;

[0083] The feature analysis unit performs local peak detection on the edge point straight line features to obtain two local peaks (ρz1, θz1) and (ρz2, θz2).

[0084] Specifically, the method for detecting local peak values ​​is not specifically limited in this embodiment, and those skilled in the art can freely set it as long as the detection requirements of local peak values ​​are met. The parameter space can be quantized and divided into a two-dimensional accumulator array. Whenever a curve passes through a cell, the value of the cell is increased by one. The cell with the largest value is found in the accumulator array, and its coordinates correspond to the parameters (ρ, θ) of the detected straight line. The local peak value described in this embodiment is to create a parameter space when performing Hough transform on the edge point, in which one dimension is ρ and the other dimension is θ. Then, for each edge point, a curve is drawn in the parameter space: ρ=Bx×cos(θ)+By×sin(θ). This curve represents the mapping of all straight lines passing through the point (Bx, By) in the parameter space. In the parameter space, points on the same straight line will converge at a point (ρ, θ) to form a local peak value.

[0085] Specifically, the positioning unit locates the welding point of the welding workpiece according to the result of the edge point straight line feature peak detection, and the positioning unit sets the coordinates of the welding point to (xw, yw), setting:

[0086] xw=[ρz2×sin(θz1)-ρz1 ×sin(θz2)] / [cos(θz2)×sin(θz1)-cos(θz1)×sin(θz2)];

[0087] yw=[ρz1×cos(θz2)-ρz2 ×cos(θz1)] / [cos(θz2)×sin(θz1)-cos(θz1)×sin(θz2)].

[0088] Specifically, the positioning unit calculates the coordinates of the intersection of two straight lines detected by the Hough transform to improve the accuracy of welding point positioning. At the same time, the calculation amount of coordinate extraction is relatively small, and the welding point coordinates can be obtained in real time, which improves the welding efficiency and the quality of tower welding. Technical personnel in this field can transmit the positioning results of the welding points to the control system of the welding robot to complete precise welding.

[0089] See also Figure 3 As shown, it is a schematic diagram of the structure of the abnormality analysis module of this embodiment, and the abnormality analysis module includes:

[0090] A calculation unit, used to calculate the welding point positioning abnormality index according to the positioning result of the welding point coordinates and the horizontal coordinate and vertical coordinate of the marking point in the welding workpiece image;

[0091] The abnormality analysis unit is used to analyze the abnormality of the welding point positioning according to the calculation result of the welding point positioning abnormality index, and the abnormality analysis unit is connected to the calculation unit.

[0092] Specifically, the calculation unit calculates the welding point positioning abnormality index according to the positioning result of the welding point coordinates and the horizontal coordinate Jx and the vertical coordinate Jy of the marking point in the welding workpiece image. The calculation unit sets the welding point positioning abnormality index to Yc, and sets: Among them, k0 is the preset error threshold.

[0093] Specifically, the calculation unit sets a preset error threshold to improve the calculation accuracy of the welding point positioning anomaly index, thereby improving the accuracy of welding abnormality analysis, thereby improving the accuracy of welding point positioning, and ultimately improving welding efficiency and tower welding quality; in this embodiment, there is no specific limitation on the setting of the preset error threshold, and technical personnel in this field can set it freely, as long as the setting requirements of the preset error threshold are met, where the optimal value of k0 is 6 pixel units.

[0094] Specifically, the abnormality analysis unit compares the welding point positioning abnormality index with the preset abnormality index y0, and analyzes the abnormality of the welding point positioning according to the comparison result, wherein:

[0095] When Yc≤y0, the abnormality analysis unit determines that the welding point positioning is normal;

[0096] When Yc>y0, the abnormality analysis unit determines that the welding point positioning is abnormal.

[0097] Specifically, the abnormality analysis unit improves the accuracy of welding positioning abnormality analysis by setting a preset abnormality index, thereby improving the accuracy of welding point positioning, and ultimately improving welding efficiency and tower welding quality; in this embodiment, there is no specific limitation on the setting of the preset abnormality index, and technical personnel in this field can set it freely, as long as the setting requirements of the preset abnormality index are met, where the optimal value of y0 is 0.52.

[0098] See also Figure 4 As shown, it is a schematic diagram of the structure of the analysis and optimization module of this embodiment, and the analysis and optimization module includes:

[0099] A first optimization unit is used to optimize the analysis process of welding point positioning abnormality according to the obtained reflectivity of the welding workpiece surface;

[0100] The second optimization unit is used to perform secondary optimization on the analysis process of welding point positioning abnormality according to the acquired surface roughness of the welding workpiece, and the second optimization unit is connected to the first optimization unit.

[0101] Specifically, the first optimization unit compares the obtained reflectivity s0 of the welding workpiece surface with the preset reflectivity s1, and optimizes the analysis process of the abnormality of the welding point positioning according to the comparison result, wherein:

[0102] When s0≤s1, the first optimization unit determines that the surface reflectivity of the welding workpiece is normal and does not perform optimization;

[0103] When s0>s1, the first optimization unit determines that the surface reflectivity of the welding workpiece is abnormal, and optimizes the analysis process of the abnormality of the welding point positioning, and sets the optimized preset abnormality index to y0', setting: y0'=y0×{1+exp[3×(s0-s1) / (s0+s1)-3]}.

[0104] Specifically, the first optimization unit sets a preset reflectivity to reduce the influence of the surface reflectivity of the welding workpiece on the abnormality analysis of the welding point positioning, thereby improving the accuracy of the welding point positioning, and ultimately improving the welding efficiency and the quality of the tower welding. In this embodiment, there is no specific limitation on the setting of the preset reflectivity, and technical personnel in this field can set it freely, as long as the setting requirements of the preset reflectivity are met, where the optimal value of s1 is 0.7.

[0105] Specifically, the second optimization unit compares the obtained surface roughness c0 of the welding workpiece with the preset roughness c1, and performs secondary optimization on the analysis process of the abnormality of the welding point positioning according to the comparison result, wherein:

[0106] When c0≤c1, the second optimization unit determines that the surface roughness of the welding workpiece is normal and does not perform optimization;

[0107] When c0>c1, the second optimization unit determines that the surface roughness of the welding workpiece is abnormal, and performs secondary optimization on the analysis process of the abnormality of the welding point positioning, and sets the optimized preset reflectivity to s1', setting s1'=s1×{1+ln[(c0-c1) / (c0+c1)+1]}, when s1'≥s2, the value of s1' is s2, and s2 is the preset reflectivity threshold.

[0108] Specifically, the second optimization unit reduces the influence of the surface roughness of the welding workpiece on the abnormality analysis of welding point positioning by setting preset roughness and preset reflectivity thresholds, thereby improving the accuracy of welding positioning abnormality analysis, thereby improving the accuracy of welding point positioning, and ultimately improving welding efficiency and tower welding quality; in this embodiment, there is no specific limitation on the setting of the preset roughness and preset reflectivity thresholds, and technical personnel in this field can set them freely, as long as the setting requirements of the preset roughness and preset reflectivity thresholds are met, where the optimal value of c1 is 2.6um, and the optimal value of s2 is 0.9.

[0109] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. An intelligent positioning system for welding points of a tower, characterized in that: include, A first acquisition module is used to acquire a grayscale image of a welding workpiece image; A feature extraction module is used to extract feature points according to the grayscale image of the acquired welding workpiece image; An edge detection module is used to analyze the relative positions of feature points according to the gradient of feature points to obtain edge points; A positioning module is used to extract the straight line features of the edge points, perform local peak detection on the straight line features of the edge points, and locate the welding points of the welding workpiece according to the results of the peak detection of the straight line features of the edge points; The second acquisition module is used to acquire the marking point information and the welding workpiece information; An abnormality analysis module is used to analyze the positioning abnormality of the welding point according to the positioning result of the welding point and the acquired marking point information; An analysis and optimization module, used to optimize the analysis process of the abnormality of the positioning of the welding point based on the obtained reflectivity of the welding workpiece surface, and also used to optimize the analysis process of the abnormality of the positioning of the welding point based on the roughness of the welding workpiece surface; A machine learning module is used to update the location analysis process of the welding points in the next monitoring period according to the analysis results of the abnormality of the welding point location in the monitoring period; The analysis and optimization module is provided with a first optimization unit and a second optimization unit. The first optimization unit is used to compare the obtained surface reflectivity s0 of the welding workpiece with the preset reflectivity s1, and optimize the analysis process of the abnormality of the welding point positioning according to the comparison result, wherein: When s0≤s1, the first optimization unit determines that the surface reflectivity of the welding workpiece is normal and does not perform optimization; When s0>s1, the first optimization unit determines that the reflectivity of the welding workpiece surface is abnormal, and optimizes the analysis process of the abnormality of the welding point positioning, and sets the optimized preset abnormality index to y0', setting: y0'=y0×{1+exp[3×(s0-s1) / (s0+s1)-3]}, y0 is the preset abnormality index; The second optimization unit is used to compare the obtained surface roughness c0 of the welding workpiece with the preset roughness c1, and perform secondary optimization on the analysis process of the abnormality of the welding point positioning according to the comparison result, wherein: When c0≤c1, the second optimization unit determines that the surface roughness of the welding workpiece is normal and does not perform optimization; When c0>c1, the second optimization unit determines that the surface roughness of the welding workpiece is abnormal, and performs secondary optimization on the analysis process of the abnormality of the welding point positioning, and sets the optimized preset reflectivity to s1', and sets s1'=s1×{1+ln[(c0-c1) / (c0+c1)+1]}. When s1'≥s2, the value of s1' is s2, and s2 is the preset reflectivity threshold.

2. The intelligent positioning system for welding points of a tower according to claim 1, characterized in that: The edge detection module calculates the gradient amplitude F(x,y) of the feature point and sets F(x,y)=(Fx 2 +Fy 2 ) 0.5 ; The edge detection module compares the gradient amplitude F(x, y) of the feature point with the preset amplitude f0, and analyzes the relative position of the feature point according to the comparison result, wherein: When F(x,y)≤f0, the edge detection module determines that the feature point is a non-edge point; When F(x,y)>f0, the edge detection module determines that the feature point is an edge point; Among them, Fx is the gradient of the feature point (x, y) in the horizontal direction, and Fy is the gradient of the feature point (x, y) in the vertical direction.

3. The intelligent positioning system for welding points of a tower according to claim 2, characterized in that: The positioning module is provided with a feature analysis unit and a positioning unit. The feature analysis unit is used to perform Hough transform on the edge points and extract straight line features from the edge points. The calculation formula is: ρ=Bx×cos(θ)+By×sin(θ) Among them, ρi is the distance between the extracted i-th straight line and the origin, θi is the angle between the extracted i-th straight line and the x-axis, Bx is the horizontal coordinate of the edge point plane rectangular coordinate system, and By is the vertical coordinate of the edge point plane rectangular coordinate system.

4. The intelligent positioning system for welding points of an iron tower according to claim 3 is characterized in that: The positioning unit is used to locate the welding point of the welding workpiece according to the result of the edge point straight line feature peak detection. The positioning unit sets the coordinates of the welding point to (xw, yw), and sets: xw=[ρz2×sin(θz1)-ρz1×sin(θz2)] / [cos(θz2)×sin(θz1)-cos(θz1)×sin(θz2)]; yw=[ρz1×cos(θz2)-ρz2×cos(θz1)] / [cos(θz2)×sin(θz1)-cos(θz1)×sin(θz2)].

5. The intelligent positioning system for welding points of a tower according to claim 4, characterized in that: The abnormality analysis module is provided with a calculation unit and an abnormality analysis unit. The calculation unit is used to calculate the abnormality index of the welding point positioning according to the positioning result of the welding point coordinates and the horizontal coordinate Jx and the vertical coordinate Jy of the marking point in the welding workpiece image. The calculation unit sets the abnormality index of the welding point positioning to Yc, and sets: Among them, k0 is the preset error threshold.

6. The intelligent positioning system for welding points of a tower according to claim 5, characterized in that: The abnormality analysis unit compares the welding point positioning abnormality index with the preset abnormality index y0, and analyzes the abnormality of the welding point positioning according to the comparison result, wherein: When Yc≤y0, the abnormality analysis unit determines that the welding point positioning is normal; When Yc>y0, the abnormality analysis unit determines that the welding point positioning is abnormal.

7. The intelligent positioning system for welding points of a tower according to claim 6, characterized in that: The machine learning module updates the positioning analysis process of the welding points in the next monitoring period with the analysis results of the abnormality of the positioning of the welding points in the monitoring period, wherein: When r0 / r1≤E, the machine learning module determines that the positioning of the welding point in the current monitoring period is normal and does not update; When r0 / r1>E, the machine learning module determines that the positioning of the welding point in the current monitoring period is abnormal, and updates the positioning analysis process of the welding point in the next monitoring period, and sets the updated preset amplitude to f0', setting f0'=f0×{1-sin[(r0 / r1-E)×(π / 2)]}; Among them, r0 is the number of abnormal welding point positioning within the monitoring period, r1 is the total number of welding point positioning within the monitoring period, and E is the preset abnormal proportion.

Images