Profiling welding path planning method based on image processing and geometric forming simulation, electronic equipment and storage medium

Abstract

The invention provides a profiling welding path planning method based on image processing and geometric forming simulation, electronic equipment and a storage medium, and relates to the technical field of welding. The profiling welding path planning method can more accurately simulate complex welding bead shapes, boundary calculation and posture changes in the actual welding process, and comprises the steps that a welding bead forming prediction model is constructed; establishing a groove section image; the groove section image is read based on the image processing technology, and weld bead layering is conducted on the groove; calculating the number of pixels occupied by a weld bead section formed by the molten drops in the groove; calculating the pixels of the current layer of welding bead based on an image processing technology, and completing welding bead partitioning of the current layer of welding bead according to the calculated number of the pixels occupied by the welding bead section formed by the molten drops in the groove; calculating the gravity center of the current welding bead, and taking the gravity center as a molten drop target drop point of the current welding bead; and optimizing the starting point of the droplet transition, and taking the direction of the gravity center point of the current welding bead pointing to the starting point of the droplet transition as a planning posture.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of welding, and in particular to a profiling welding path planning method based on image processing and geometric profiling simulation, an electronic device and a storage medium. BACKGROUND

[0002] With the continuous development of welding automation, welding robots play an increasingly important role in the field of welding. However, unlike traditional manual multi-layer multi-pass welding, robot welding needs to strictly specify the position and pose of the welding path. This requires reasonable path planning technology, especially in the bevel welding, multi-layer multi-pass path planning has become one of the key technologies.

[0003] At present, in the field of bevel welding, the existing multi-layer multi-pass welding path planning technology mainly includes equal height planning and equal area planning algorithm. This kind of algorithm simplifies the geometric shape of the welding path, and regards the weld as a simple geometric figure (such as triangle, parallelogram, rectangle, etc.), so as to calculate the path. Specifically, the principle of equal height planning is to assume that the height of each layer of weld is equal, and then plan the position of the weld in the layer according to equal interval or other principles, while the equal area planning is to make the area of each weld equal to plan.

[0004] Although these algorithms improve the efficiency of path planning to some extent, there are still the following significant shortcomings:

[0005] (1) Simplification of weld shape: existing algorithms usually simplify the weld into a simple geometric shape such as triangle, parallelogram or rectangle, ignoring the true appearance of the weld. However, the actual weld surface usually presents a circular arc shape, and there may be a complex lap relationship between the welds. Simplifying the weld into a regular geometric shape cannot truly reflect the lap between the welds and the complex shape in the welding process, which may cause deviation in the welding quality and affect the final weld formation.

[0006] (2) Dependence on bevel curve boundary equation: these path planning methods usually require the boundary equation of the bevel to be known, which is complex for designers. Designers usually use CAD software to design the bevel, and to obtain the accurate boundary equation of the bevel, secondary development or complex calculation of the CAD software is often required, increasing the difficulty of design work. In order to simplify this process, designers usually choose to use straight lines and other simplified methods to represent the bevel boundary, but this simplified treatment often cannot accurately reflect the real welding conditions.

[0007] (3) Ignoring the influence of key factors such as droplet transfer direction on forming: The posture in the welding process has an important influence on the welding effect and weld forming. Different postures will affect the transfer direction of the droplet, and then affect the forming quality of the weld. However, in the current multi-layer multi-pass path planning method, the posture planning generally adopts fixed or preset posture, ignoring the change of the welding posture under specific working conditions, limiting the flexibility and accuracy of path planning.

[0008] In summary, the existing multi-layer multi-pass path planning algorithm has certain limitations and cannot accurately simulate the complex weld shape, boundary calculation and posture change in the actual welding process. In order to improve the welding quality and efficiency, it is urgent to develop a more accurate and flexible multi-layer multi-pass path planning technology that can consider the weld topography, groove boundary and posture change. SUMMARY

[0009] The technical problem to be solved by the present application is to provide a contour welding path planning method based on image processing and geometric shaping simulation, which can more accurately simulate the complex weld shape, boundary calculation and posture change in the actual welding process.

[0010] The technical scheme adopted by the present application to solve the technical problem is: a contour welding path planning method based on image processing and geometric shaping simulation, comprising the steps of:

[0011] S1, constructing a weld forming prediction model, the weld forming prediction model is based on the following forming mechanism: the total volume of the droplet remains unchanged during the falling and spreading process; the groove profile and the surface of the previous weld remain unchanged during the forming process; the droplet begins to solidify after contacting the entity, and the solidification time and the droplet movement speed are fixed; the weld cross section is an equivalent circular arc; the weld forming prediction model is used to simulate the movement process of the droplet in the groove, including droplet radius correction and movement direction correction, and setting the droplet solidification termination condition;

[0012] Droplet radius correction: when the droplet is invaded by the entity area, the area loss is compensated by increasing the radius until the weld cross-sectional area reaches the droplet area;

[0013] Movement direction correction: the movement direction of the droplet will deviate in the groove width direction during the falling process, and the movement direction needs to be corrected constantly;

[0014] Wherein, the droplet movement direction correction formula is:

[0015] ;

[0016] V is the droplet transfer direction, T is the transverse correction vector, and V is the corrected droplet movement direction.

[0017] ,

[0018] In the formula, : the center of gravity point of the solid support area in the circle where the droplet is located; : the center of gravity point of the weld bead, : correction coefficient; : unit vector, the direction from to , indicating the direction of the support eccentricity; : component of the vector x direction; : unit column vector, indicating the x direction;

[0019] Droplet solidification termination condition: the time of the center of the circle where the droplet is located from the time when the droplet starts to contact the solid to the time when the droplet is completely solidified is equal to the solidification compensation termination condition threshold value, or the area ratio of the weld bead to the area of the circle where the weld bead outer surface arc is located is equal to the area ratio termination condition coefficient;

[0020] S2, a groove cross-sectional image is established, and the pixel of the solid part of the groove cross-sectional image is distinguished from the pixel of the groove area;

[0021] S3, based on the image processing technology, the above-mentioned groove cross-sectional image is read, and based on the weld bead cross-sectional height formed by the droplet, the groove is layered to determine the number of layers of the weld bead;

[0022] S4, the number of pixels occupied by the weld bead cross section formed by the droplet in the groove is calculated: according to the weld bead cross-sectional area formed by the droplet, an area equal to the area is demarcated in the groove, and then the number of pixels contained in the area is calculated based on the image processing technology. The number of pixels is the number of pixels occupied by the weld bead cross section formed by the droplet in the groove;

[0023] S5, the determined weld bead layers are traversed in the welding order, the pixels of the current layer weld bead are calculated based on the image processing technology, and the weld bead partitioning of the current layer weld bead is completed according to the number of pixels occupied by the weld bead cross section formed by the droplet in the groove;

[0024] S6, the centers of gravity of the weld beads of the current layer weld bead are calculated in the welding order, and the center of gravity is taken as the target landing point of the droplet of the current weld bead;

[0025] S7, the starting point of the droplet transition is optimized, and the direction from the center of gravity point of the current weld bead to the droplet transition starting point is taken as the planned pose:

[0026] S71, the droplet transition starting point is calculated according to the center of gravity of the current weld bead and the droplet transition direction:

[0027] ;

[0028] in, Center point of weld; The initial transition direction of the molten droplet is vertically downward. Transition distance; : Droplet transition initiation point;

[0029] S72, the distance the molten droplet moves along the droplet transition direction by one pixel;

[0030] S73. Using image processing methods, count whether the pixels in the above droplet contain solid points; if they contain solid points, proceed to step S74; if they do not contain solid points, return to steps S72 and S73.

[0031] S74. The first solid point of the pixel within the droplet is taken as the tangent point between the droplet and the solid. Calculate the tangency point. With the current weld center of gravity The distance d; if d is less than the droplet radius, the current droplet transition direction is substituted into the established weld bead forming prediction model to correct the motion direction and droplet radius, the simulation modifies the actual forming of the current weld bead, and the image is updated; if d is greater than or equal to the droplet radius, step S75 is executed.

[0032] S75, with the current weld center of gravity Using the initial radius *r* of the molten droplet as the center, several sampling points are uniformly generated on its upper semicircle. Then, using these sampling points as centers and *r* as the radius, a test circle is constructed. The number of pixels within the bevel of each test circle is counted, and the test circle with the largest number of pixels within the bevel is identified, with its center set as... Calculate the new droplet transition direction V / and with a new droplet transition direction V / Return to steps S61 to S63;

[0033] Among them, V / ;

[0034] S8. Return to execute steps S6 and S7 in sequence until all welds in the current layer have been traversed.

[0035] S9. Return to the previous step and execute steps S5, S6 and S7 in sequence until all weld layers have been traversed.

[0036] Furthermore, the cross-sectional area and height of the weld bead are obtained by the following method:

[0037] A weld bead is welded onto the upper surface of the plate, and the contour of the weld bead is measured using a structured light inspection instrument to obtain the cross-sectional area and height of the weld bead.

[0038] Furthermore, the correction coefficient k p The following methods were used to obtain the results, including:

[0039] The first weld bead is welded onto the upper surface of the plate.

[0040] Move the welding path 0.5 times the weld width to one side of the first weld bead and weld the second weld bead.

[0041] The weld contours of the two welds were measured using a structured light inspection instrument, and then a forming simulation experiment was conducted using the weld forming prediction model. k was adjusted. p The parameters are adjusted until the error between the simulated contour and the real contour is minimized, thus yielding k. p The value of .

[0042] Furthermore, the solidification compensation termination condition threshold is obtained using the following method:

[0043] A weld bead is horizontally welded on the side of a vertical plate. During the welding process, a high-speed camera continuously photographs the weld bead and measures the solidification time. This solidification time is the solidification compensation termination condition threshold.

[0044] Furthermore, the area ratio termination condition coefficient is obtained using the following method:

[0045] Based on the initial radius r of the molten droplet, a U-shaped bevel with a bottom radius equal to r is set;

[0046] A weld bead is welded within the U-shaped groove. The cross-sectional area S of the weld bead and the area S of the circle containing the outer surface arc of the weld bead are measured using a structured light inspection instrument. all Then calculate the area ratio termination condition coefficient bs, bs = S / S all .

[0047] An electronic device includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When executed by the processor, the computer program implements the above-described contour welding path planning method based on image processing and geometric simulation.

[0048] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described conformal welding path planning method based on image processing and geometrical simulation.

[0049] The beneficial effects of this invention are: This invention breaks through the limitations of traditional equal area and equal height planning methods, which simplify the weld bead shape to a polygon and do not consider the direction of droplet transition. By combining geometric simulation and image processing methods, it fully considers key factors such as the real morphology of the groove, the direction of droplet transition, and the accumulation of previous weld beads, and realizes automatic planning of contour paths for multi-layer and multi-pass weld beads. It can more accurately simulate the complex weld bead shape, boundary calculation, and posture changes in the actual welding process, which makes it easier for welding programmers to quickly verify and optimize welding schemes. Attached Figure Description

[0050] Figure 1 This is a diagram showing the deformation process of a falling molten droplet;

[0051] Figure 2 This is a schematic diagram of the solid support area within the circle where the molten droplet is located and the weld bead.

[0052] Figure 3 This is a schematic diagram of the weld bead layers in a V-groove.

[0053] Figure 4 This is a schematic diagram of the weld bead partitioning of a V-shaped groove;

[0054] Figure 5 This is a schematic diagram of a molten droplet falling in a vertically downward direction from its initial orientation;

[0055] Figure 6 It calculates the tangency point between the molten droplet and the solid. To the current weld center of gravity A diagram illustrating the distance;

[0056] Figure 7 A schematic diagram of constructing the test circle;

[0057] Figure 8 This is a schematic diagram of the test circle with the largest number of pixels within the bevel.

[0058] Figure 9 This is a schematic diagram of the optimized droplet transition direction;

[0059] Figure 10 This is a schematic diagram of scanning and partitioning the weld bead layer;

[0060] Figure 11 This is a schematic diagram of the area ratio termination condition coefficient test;

[0061] Figure 12 It is the effect of V-shaped groove contour welding path planning;

[0062] The figure shows: droplet 1, solid 2, test circle 4, bevel 21, circle 22, weld bead 31, and solid support area 32 within the circle where the droplet is located. Detailed Implementation

[0063] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0064] The conformal welding path planning method based on image processing and geometrical forming simulation of the present invention includes the following steps:

[0065] S1. Construct a weld bead forming prediction model, which is based on the following forming mechanism: It does not consider changes in metal density caused by temperature variations, assuming (simplified) that the total volume of the molten droplet remains constant during its fall and spread; it does not consider the remelting and geometric modification of the solid by the molten droplet, assuming (simplified) that the groove profile and the surface of the previous weld remain constant during the forming process; it does not explicitly consider changes caused by heat dissipation conditions and droplet size, assuming (simplified) that the droplet begins to solidify after contacting the solid, and the solidification time and droplet velocity are fixed; based on the fact that well-formed welds generally exhibit a continuous and smooth outer contour in engineering practice, the weld bead cross-section is simplified to an equivalent circular arc; the weld bead forming prediction model is used to simulate the movement process of the molten droplet within the groove. Based on the above mechanism, the molten droplet falling deformation process of this model is shown in the attached figure. Figure 1 As shown:

[0066] (1) The molten droplet 1 transitions from O1 to O2: along the direction of motion The movement does not encounter entity 2, and its direction and radius remain unchanged.

[0067] (2) The molten droplet transitions from O2 to O3: along the direction of motion The motion does not change direction or radius unless it encounters a solid object.

[0068] (3) The molten droplet transitions from O3 to O4: along the direction of movement When the molten droplet encounters an entity, it needs to be corrected for radius and direction of motion, and the solidification termination condition of the droplet needs to be set.

[0069] Droplet radius correction: When the droplet is encroached upon by the solid area, the area loss is compensated by increasing the radius until the weld cross-sectional area reaches the droplet area;

[0070] Motion direction correction: During the falling process of the molten droplet, it is supported by the asymmetrical entities on both sides, and its motion direction will deviate in the bevel width direction. The motion direction needs to be constantly corrected.

[0071] The formula for correcting the direction of droplet movement is as follows:

[0072] ;

[0073] The corrected direction of droplet movement is V, where V is the droplet transition direction and T is the transverse correction vector.

[0074] ,

[0075] See Figure 2 In the formula, : The center of gravity of the solid support region 32 within the circle where the molten droplet is located; The center of gravity of weld bead 31 Correction factor; Unit vector, direction from point to , indicating the direction of support for eccentricity; : The component of the vector in the x-direction (horizontal component); : Unit column vector, representing the x-direction;

[0076] The solidification termination condition of the droplet is: the time it takes for the droplet to move to the center of the solidification compensation termination condition after it starts to contact the solidification entity is equal to the solidification compensation termination condition threshold, or the area ratio of the weld bead within the circle where the droplet is located is equal to the area ratio termination condition coefficient. The solidification compensation termination condition threshold refers to the time from when the droplet starts to contact the solidification entity to when it is completely solidified. The area ratio termination condition coefficient refers to the ratio of the cross-sectional area of ​​the weld bead to the area of ​​the circle containing the arc on the outer surface of the weld bead.

[0077] S2. Establish a bevel cross-section image, wherein the pixels of the solid part of the bevel cross-section image are set differently from the pixels of the bevel region, that is, the pixels of the solid part of the bevel cross-section image are set differently from the pixels of the bevel region.

[0078] S3. Based on image processing technology, read the above-mentioned bevel cross-section image, and based on the weld bead cross-section height h formed by the molten droplets, perform weld bead layering on bevel 21 to determine the number of weld bead layers (see...). Figure 3 The weld bead cross-sectional height h formed by the molten droplets can be obtained experimentally. It can be understood that the number of weld bead layers = (groove depth / weld bead cross-sectional height h) rounded up.

[0079] S4. Calculate the number of pixels occupied by the weld bead cross section formed by the droplet within the groove: Based on the area of ​​the weld bead cross section formed by the droplet, delineate an area with the same area within the groove, and then calculate the number of pixels contained in this area based on image processing technology. This number of pixels is the number of pixels occupied by the weld bead cross section formed by the droplet within the groove.

[0080] S5. According to the welding sequence, traverse each determined weld layer in turn, calculate the pixels of the current layer weld based on image processing technology, and complete the weld partitioning of the current layer weld according to the number of pixels occupied by the weld cross section formed by the molten droplet in the groove. In this way, the previous layer weld is divided into one or more welds with the same number of pixels (area) as the number of pixels (area) occupied by the weld cross section formed by the molten droplet in the groove. Figure 4 The current layer of weld beads is divided into three weld beads;

[0081] S6. Traverse each weld bead in the current layer according to the welding sequence, calculate the centroid of the current weld bead, and use this centroid as the target droplet landing point for the current weld bead (see...). Figure 5 );

[0082] S7. Optimize the starting point of droplet transition, and use the direction from the current weld bead's center of gravity to the starting point of droplet transition as the planning orientation:

[0083] S71, such as Figure 5 As shown, the droplet transfer initiation point is calculated based on the current weld bead's centroid and the droplet transfer direction:

[0084] ;

[0085] in, Center point of weld; : Droplet transition direction, the initial transition direction (V0) is vertically downward; Transition distance, set as arc length, is generally 5mm based on experience; : Droplet transition initiation point;

[0086] S72, the distance the molten droplet moves along the droplet transition direction by one pixel;

[0087] S73. Using image processing methods, count whether the pixels in the above droplet contain solid points; if they contain solid points, proceed to step S74; if they do not contain solid points, return to steps S72 and S73.

[0088] S74, such as Figure 6 As shown, the first solid point of the pixel within the droplet is taken as the tangent point between the droplet and the solid. Calculate the tangency point. With the current weld center of gravity The distance d; if d is less than the droplet radius, the current droplet transition direction is substituted into the established weld bead forming prediction model, the actual forming of the current weld bead is simulated and modified, and the image is updated; if d is greater than or equal to the droplet radius, step S75 is executed.

[0089] S75, such as Figures 7 to 9 As shown, with the current weld bead centroid point Using the initial radius *r* of the molten droplet as the center, several sampling points are uniformly generated on its upper semicircle. Then, using these sampling points as centers and *r* as the radius, a test circle 4 is constructed. The number of pixels located within the bevel of each test circle 4 is counted, and the test circle 4 with the largest number of pixels located within the bevel is identified, with its center set as... Calculate the new droplet transition direction V / and with a new droplet transition direction V / Return to steps S71 to S73;

[0090] Among them, V / ;

[0091] S8. Return to execute steps S6 and S7 in sequence until all welds in the current layer have been traversed.

[0092] S9. Return to the previous step and execute steps S5, S6 and S7 in sequence until all weld layers have been traversed.

[0093] The effect of using the method of this invention for contour welding path planning of V-shaped grooves is as follows: Figure 12 As shown.

[0094] The bevel cross-sectional image of the present invention can be drawn by CAD or acquired by a structured light weld seam tracker. In some embodiments, the bevel cross-sectional image is drawn by CAD, with the solid portion area filled in black and the bevel in white.

[0095] like Figure 10 As shown, specifically, the scanning boundary is taken from both ends of the current weld bead, and pixels are traversed horizontally column by column. For each column, pixels are scanned downwards from the straight line position of that column. If a pixel is a bevel area pixel, it is included in the area of ​​that layer and the scanning continues downwards. When a solid part pixel is detected for the first time, the scanning of that column is terminated and the next column is processed. Through this column-by-column vertical scanning method, accurate weld fill area can be obtained when the bottom of the layer is irregular.

[0096] This invention overcomes the limitations of traditional equal-area and equal-height planning methods, which simplify weld bead shapes to polygons and fail to consider droplet transition directions. By combining geometric simulation and image processing methods, it fully considers key factors such as the true morphology of the groove, droplet transition direction, and previous weld bead accumulation, and realizes automatic contour path planning for multi-layer and multi-pass welds. It can more accurately simulate the complex weld bead shapes, boundary calculations, and posture changes in the actual welding process, making it easier for welding programmers to quickly verify and optimize welding schemes.

[0097] The present invention obtains the cross-sectional area and height of the weld bead using the following method:

[0098] A weld bead is welded onto the upper surface of the plate, and the contour of the weld bead is measured using a structured light inspection instrument to obtain the cross-sectional area and height of the weld bead.

[0099] Correction coefficient k pThe following method was used to obtain the result: A first weld bead was welded onto the upper surface of the flat plate. Then, the welding path was moved 0.5 times the weld bead width to one side of the first weld bead to weld a second weld bead. The weld bead profiles of the two welds were then measured using a structured light inspection instrument. Finally, a forming simulation experiment was conducted using the aforementioned weld bead forming prediction model, and k was adjusted. p The parameters are adjusted until the error between the simulated contour and the real contour is minimized, at which point k... p The parameter is the k we are looking for. p The value of .

[0100] The solidification compensation termination condition threshold can be measured on a horizontal, inclined, or vertical plate. Due to the tension between the weld beads, the tension significantly affects the weld bead morphology when measured on a flat or inclined plate. Therefore, to more accurately measure the solidification compensation termination condition threshold, this invention preferably employs the following method: a weld bead is horizontally welded to the side of a vertical plate. During the welding process, a high-speed camera continuously photographs the weld bead, thereby measuring the solidification time. This solidification time is the solidification compensation termination condition threshold. Specifically, the starting point for timing the weld bead solidification time is when the weld bead contacts the substrate, and the ending point is when the shape of the weld bead no longer changes.

[0101] like Figure 11 As shown, in this embodiment of the invention, the area ratio termination condition coefficient is obtained using the following method:

[0102] a. Based on the initial radius r of droplet 1, set a U-shaped bevel with a bottom radius equal to r;

[0103] b. Weld a weld bead 31 within the U-shaped groove, and use a structured light inspection instrument to measure the cross-sectional area S of the weld bead 31 and the area S of the circle 22 containing the outer surface arc of the weld bead. all Then calculate the area ratio termination condition coefficient bs, bs = S / S all .

[0104] It should be noted that the specific value of the initial radius r of the molten droplet mentioned above is set accordingly based on the actual radius of the molten droplet to be planned for simulated welding according to the present invention.

Claims

1. A contour welding path planning method based on image processing and geometric forming simulation, characterized in that, Including the following steps: S1. Construct a weld bead forming prediction model, which is based on the following forming mechanism: the total volume of the molten droplet remains unchanged during its fall and spread; the groove profile and the surface of the previous weld remain unchanged during the forming process; the molten droplet begins to solidify after contacting the solid, and the solidification time and the molten droplet movement speed are fixed; the weld bead cross section is an equivalent circular arc; the weld bead forming prediction model is used to simulate the movement process of the molten droplet in the groove, including molten droplet radius correction and movement direction correction, and sets the molten droplet solidification termination condition; Droplet radius correction: When the droplet is encroached upon by the solid area, the area loss is compensated by increasing the radius until the weld cross-sectional area reaches the droplet area; Motion direction correction: During the falling process of the molten droplet, it is supported by the asymmetrical entities on both sides, and its motion direction will deviate in the bevel width direction. The motion direction needs to be constantly corrected. The formula for correcting the direction of droplet movement is as follows: ； : Corrected droplet motion direction; V : Droplet transition direction; T : Lateral correction vector; ， In the formula, : The center of gravity of the solid support region within the circle where the droplet is located; The center of gravity of the weld bead. Correction factor; Unit vector, direction from point to , indicating the direction of support for eccentricity; : The component of the vector in the x-direction; : Unit column vector, representing the x-direction; The solidification termination condition of the droplet is: the time it takes for the droplet to move to the center of the solidification compensation termination condition after it starts to contact the solidification entity is equal to the solidification compensation termination condition threshold, or the area ratio of the weld bead within the circle where the droplet is located is equal to the area ratio termination condition coefficient. The solidification compensation termination condition threshold refers to the time from when the droplet starts to contact the solidification entity to when it is completely solidified. The area ratio termination condition coefficient refers to the ratio of the cross-sectional area of ​​the weld bead to the area of ​​the circle containing the arc on the outer surface of the weld bead. S2. Create a bevel cross-section image, wherein the pixels of the solid part of the bevel cross-section image are set differently from the pixels of the bevel region; S3. Based on image processing technology, read the above-mentioned bevel cross-section image, and based on the weld bead cross-section height formed by the molten droplets, perform weld bead layering on the bevel to determine the number of weld bead layers; S4. Calculate the number of pixels occupied by the weld bead cross section formed by the molten droplet within the groove: Based on the area of ​​the weld bead cross section formed by the molten droplet, delineate an area with the same area within the groove, and then calculate the number of pixels contained in this area based on image processing technology. This number of pixels is the number of pixels occupied by the weld bead cross section formed by the molten droplet within the groove. S5. According to the welding sequence, traverse each determined weld layer in turn, calculate the pixels of the current layer weld based on image processing technology, and complete the weld partitioning of the current layer weld based on the number of pixels occupied by the weld cross section formed by the molten droplet in the groove. S6. In the welding sequence, traverse each weld bead of the current layer, calculate the centroid of the current weld bead, and use the centroid as the target landing point of the molten droplet of the current weld bead. S7. Optimize the starting point of droplet transition, and use the direction from the current weld bead's center of gravity to the starting point of droplet transition as the planning orientation: S71. Calculate the droplet transfer initiation point based on the current weld bead's center of gravity and the droplet transfer direction: ； in, Center point of weld; : Droplet transition direction, the initial transition direction is vertically downward; Transition distance; : Droplet transition initiation point; S72, the distance the molten droplet moves along the droplet transition direction by one pixel; S73. Using image processing methods, count whether the pixels in the above droplet contain solid points; if they contain solid points, proceed to step S74; if they do not contain solid points, return to steps S72 and S73. S74. The first solid point of the pixel within the droplet is taken as the tangent point between the droplet and the solid. Calculate the tangency point. With the current weld center of gravity The distance d; if d is less than the droplet radius, the current droplet transition direction is substituted into the established weld bead forming prediction model to correct the motion direction and droplet radius, the simulation modifies the actual forming of the current weld bead, and the image is updated; if d is greater than or equal to the droplet radius, step S75 is executed. S75, with the current weld center of gravity Using the initial radius *r* of the molten droplet as the center, several sampling points are uniformly generated on its upper semicircle. Then, using these sampling points as centers and *r* as the radius, a test circle is constructed. The number of pixels within the bevel of each test circle is counted, and the test circle with the largest number of pixels within the bevel is identified, with its center set as... Calculate the new droplet transition direction V / and with a new droplet transition direction V / Return to steps S61 to S63; Among them, V / ; S8. Return to execute steps S6 and S7 in sequence until all welds in the current layer have been traversed. S9. Return to the previous step and execute steps S5, S6 and S7 in sequence until all weld layers have been traversed.

2. The conformal welding path planning method based on image processing and geometrical forming simulation as described in claim 1, characterized in that, The cross-sectional area and height of the weld bead are obtained using the following method: A weld bead is welded onto the upper surface of the plate, and the contour of the weld bead is measured using a structured light inspection instrument to obtain the cross-sectional area and height of the weld bead.

3. The contour welding path planning method based on image processing and geometrical forming simulation as described in claim 1, characterized in that, The correction coefficient k p The following methods were used to obtain the results, including: The first weld bead is welded onto the upper surface of the plate. Move the welding path 0.5 times the weld width to one side of the first weld bead and weld the second weld bead. The weld contours of the two welds were measured using a structured light inspection instrument, and then a forming simulation experiment was conducted using the weld forming prediction model. k was adjusted. p The parameters are adjusted until the error between the simulated contour and the real contour is minimized, thus yielding k. p The value of .

4. The conformal welding path planning method based on image processing and geometrical forming simulation as described in claim 1, characterized in that, The solidification compensation termination condition threshold is obtained using the following method: A weld bead is horizontally welded on the side of a vertical plate. During the welding process, a high-speed camera continuously photographs the weld bead and measures the solidification time. This solidification time is the solidification compensation termination condition threshold.

5. The contour welding path planning method based on image processing and geometrical forming simulation as described in claim 1, characterized in that, The area ratio termination condition coefficient is obtained using the following method: Based on the initial radius r of the molten droplet, a U-shaped bevel with a bottom radius equal to r is set; A weld bead is welded within the U-shaped groove. The cross-sectional area S of the weld bead and the area S of the circle containing the outer surface arc of the weld bead are measured using a structured light inspection instrument. all Then calculate the area ratio termination condition coefficient bs, bs = S / S all .

6. An electronic device, characterized in that, The method includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When executed by the processor, the computer program implements the contour welding path planning method based on image processing and geometry simulation as described in any one of claims 1 to 5.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the contour welding path planning method based on image processing and geometrical simulation as described in any one of claims 1 to 5.

Images