Adaptive welding

Abstract

A system and method for adaptive welding. In some embodiments, the system includes a welding head, a first welding monitoring camera, and a machine learning system. The machine learning system may be configured to estimate a position of an end of a filler wire relative to a groove as the welding head forms a weld layer in the groove.

Description

[0001] Cross - Reference to Related Applications

[0002] This application claims the priority and benefit of U.S. Provisional Application No. 63 / 380,686, entitled "ADAPTIVE WELDING", filed on October 24, 2022, the entire content of which is incorporated herein by reference. Technical Field

[0003] One or more aspects in accordance with embodiments of the present disclosure relate to welding, and more particularly to systems and methods for controlling a welding system. Background Art

[0004] Welding systems can be used to produce welds in an automated or semi - automated manner. During the production of such welds, various parameters can be adjusted, including, for example, the position of a heat source or a filler wire relative to a weld groove, the travel speed, or the filler wire feed speed.

[0005] Aspects related to the present disclosure are related to such a general technical environment. Summary of the Invention

[0006] According to one embodiment of the present disclosure, there is provided a system for welding, comprising: a welding head; a first welding monitoring camera; and a machine learning system, wherein the machine learning system is configured to estimate the position of the end of a filler wire relative to the groove when the welding head forms a welding layer in the groove.

[0007] In some embodiments, the estimation includes estimating the position of a first point of interest, which is the intersection of a first edge of the groove and a reference plane, based on an image obtained by the first welding monitoring camera.

[0008] In some embodiments, the estimation includes estimating the positions of four points of interest including the first point of interest, each of the four points of interest being the intersection of an edge of the groove and the reference plane.

[0009] In some embodiments, the estimation of the position of the first point of interest includes estimating the intersection in the image of a line corresponding to the first edge and a line corresponding to the reference plane.

[0010] In some embodiments, the machine learning system is further configured to estimate the depth of the groove.

[0011] In some embodiments, the machine learning system is further configured to estimate the width of the bottom of the groove.

[0012] In some embodiments, the estimation further includes estimating the position of a fifth point of interest corresponding to the end of the filler wire.

[0013] In some embodiments, the machine learning system is further configured to estimate the error of the position of the end of the filler wire relative to the target position of the end of the filler wire.

[0014] In some embodiments, the estimation further includes estimating the position of a sixth point of interest corresponding to the tip of the tungsten electrode of the welding head.

[0015] In some embodiments, the machine learning system is further configured to estimate the error of the position of the tip of the tungsten electrode relative to the target position of the tip of the tungsten electrode.

[0016] In some embodiments, the machine learning system includes a first neural network for estimating the position of the tungsten electrode in an image obtained by the first welding monitoring camera.

[0017] In some embodiments, the machine learning system further includes a second neural network for estimating the position of the end of the filler wire.

[0018] In some embodiments, the estimation of the position of the end of the filler wire includes estimating the position of the end of the filler wire based on: the estimated position of the tungsten electrode; and the image.

[0019] In some embodiments, the machine learning system further includes a third neural network for estimating the position of the upper edge of the groove.

[0020] In some embodiments, the machine learning system includes a fourth neural network for estimating the position of the lower edge of the groove.

[0021] In some embodiments, the estimation of the position of the lower edge of the groove includes estimating the position of the lower edge of the groove based on: the estimated position of the upper edge of the groove; and the image.

[0022] According to one embodiment of the present disclosure, a method is provided, which includes: calculating the traveling speed of the substrate relative to the welding head based on the following steps: the wire feeding speed of the filler wire; the size of the filler wire; and the size of the weld layer to be formed.

[0023] In some embodiments, the size of the filler wire is the diameter of the filler wire.

[0024] In some embodiments: the size of the weld layer is the thickness of the weld layer, and the calculation is further based on the width of the weld layer.

[0025] In some embodiments, the calculation is further based on an efficiency factor. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] These and other features and advantages of the present disclosure will be understood and appreciated with reference to the specification, claims, and drawings, in which:

[0027] Figure 1A is a cross-sectional view of a substrate according to an embodiment of the present disclosure;

[0028] Figure 1B is a cross-sectional view of a substrate and a weld layer according to an embodiment of the present disclosure;

[0029] Figure 1C is a side view of a gas tungsten arc welding system according to an embodiment of the present disclosure;

[0030] Figure 1D is a side view of a gas metal arc welding system according to an embodiment of the present disclosure;

[0031] Figure 2A is a schematic diagram of an image captured by a welding monitoring camera according to an embodiment of the present disclosure;

[0032] Figure 2B is a schematic diagram of an image captured by a welding monitoring camera according to an embodiment of the present disclosure;

[0033] Figure 2C is an image captured by a welding monitoring camera according to an embodiment of the present disclosure;

[0034] Figure 3A is a block diagram of a neural network according to an embodiment of the present disclosure;

[0035] Figure 3B is a block diagram of a neural network according to an embodiment of the present disclosure; and

[0036] Figure 4 is a block diagram of a welding system according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0037] The following detailed description, in conjunction with the accompanying drawings, is intended as a description of exemplary embodiments of systems and methods for controlling a welding system provided in accordance with the present disclosure and is not intended to represent the only form in which the present disclosure may be constructed or utilized. The description sets forth the features of the present disclosure in connection with the illustrated embodiments. However, it is to be understood that the same or equivalent functions and structures may be achieved by different embodiments, which are also intended to be encompassed within the scope of the present disclosure. As indicated elsewhere herein, like element numbers are intended to indicate like elements or features.

[0038] In a welding system, metal sheets (such as metal plates or pipes) can be joined by filling a groove (the cross-section of which can be triangular, rectangular, or trapezoidal) with molten metal in one or more passes along a contact line between two metal sheets. The metal sheets (along with any partially completed welds) can be collectively referred to as the substrate. The filler metal can be provided by a filler wire, and the heat can be provided by an electric arc. In gas tungsten arc welding (GTAW) (or tungsten inert gas (TIG) welding) processes, the electric arc can be formed between a tungsten electrode and the weld pool; in gas metal arc welding (GMAW) (or metal inert gas (MIG) welding) processes, the electric arc can be formed between the filler wire and the weld pool. In other processes, the heat can be provided by another heat source, such as a laser.

[0039] For example, referring to Figure 1A , when performing machine (i.e., automated) pipe welding, two pipes 100 can be placed end to end, and the ends of the two pipes can be shaped to form a groove 104 around the outer circumference of the ends of the pipes. Then, as Figure 1B shown, one or more passes can be used to partially or fully fill the groove 104, with each pass forming an additional weld layer 106 in the groove 104. The pipes 100 can be collectively referred to as the substrate 102.

[0040] Referring to Figure 1C and Figure 1D, the welding system may include a welding head that includes a heat source and a filler wire feed system. The welding control system may include a motor or other actuator for controlling the relative position of the components of the welding head and the substrate and the wire feed during the welding process. For example, the welding control system may include a travel control system for controlling the travel speed (e.g., by moving the welding head or by moving the substrate to control the movement of the welding head along the length of the groove 104). Similarly, the position of the tungsten electrode 105 (e.g., the height of the tungsten electrode 105 and the lateral position of the tungsten electrode 105), the position of the distal end of the filler wire 110 (e.g., the point where the filler wire 110 enters the weld pool), and the wire feed speed of the filler wire can be controlled. The welding control system may also include a welding power source that can control the voltage across the arc or the current flowing through the arc.

[0041] The groove 104 may not be straight, its width may vary along its length, and it may not be perfectly aligned with the drive shaft of the travel control system, such that simply driving the travel control motor without making vertical or lateral adjustments to the components of the welding head may result in misalignment between the groove 104 and the weld as welding progresses, or may result in variations in the thickness of the metal layer deposited during each pass, or may result in welding defects. Therefore, a sensing system may be employed to measure various geometric parameters of the process, and these measurements can be used to control the actuator and the welding power source.

[0042] The sensing system may employ a welding monitoring camera 115, which may be a camera configured to obtain images of the weld pool, the tungsten electrode 105, the filler wire 110, and a portion of the substrate as welding progresses. In some embodiments, the system includes a plurality of welding monitoring cameras 115 (e.g., two welding monitoring cameras, as illustrated in Figure 1C and 1D . The images may be analyzed by a machine learning system (e.g., a system including one or more neural networks, discussed in further detail below). Figure 2A and 2B show examples of such images (for GTAW and GMAW, respectively), marked with points of interest in the images that can be recognized by the machine learning system. Figure 2C is a single frame of video obtained by the welding monitoring camera 115 in a GTAW system, with points of interest 205, 210, 215 superimposed on the image.

[0043] The sensing system may (e.g., using a machine learning system) estimate (i) (in a GTAW system) the tip on the tungsten electrode 105 (at Figure 2Athe positions (e.g., height and lateral position) marked as 205) in, (ii) the end of the filler wire 110 (in Figure 2A marked as point 210 and Figure 2B the positions (e.g., height and lateral position) marked as 260) in, and (iii) the lateral position of each of the four intersection points between the four edges of the groove 104 and the reference plane (in Figure 2A marked as 215 and Figure 2B marked as 265) in. The four edges of the groove 104 can be two upper edges where the walls of the groove 104 intersect the upper surface of the substrate, and two lower edges where the walls of the groove 104 intersect the lower surface of the groove 104 (which can be the lower surface formed during machining of the groove 104, or the upper surface of the previous weld pass of the weld, e.g., the previously deposited weld layer 106). The reference plane can be a plane defined by a horizontal line in the image such that all four intersection points can fall on the horizontal line in the image.

[0044] Based on the estimated parameters, the position error of the tungsten electrode 105 (in the GTAW system) relative to its target position, (ii) the position error of the end of the filler wire 110 relative to its target position, and (iii) the depth of the groove 104 and (iv) the width of the bottom of the groove 104 can be calculated. The target position of the tip of the tungsten electrode 105 for GTAW and the target position of the end of the filler wire 110 for GMAW can be at a certain height above the center of the groove 104, or as welding progresses, e.g., if a weave bead is used, it can move from left to right. In some embodiments, the height of the end of the filler wire 110 in the GMAW system can be estimated based on the arc voltage rather than based on the image from the welding monitoring camera 115, or in addition to being estimated based on the image from the welding monitoring camera 115. For GTAW applications, the target position of the end of the filler wire 110 can be vertically offset from the tip of the tungsten electrode 105 by a fixed amount and horizontally aligned with the tip of the tungsten electrode 105. The depth of the groove 104 and the width of the bottom of the groove 104 can be calculated based on (i) the spacing between the intersection points corresponding to the two upper edges of the groove 104, (ii) the spacing between the intersection points corresponding to the two lower edges of the groove 104, and based on the knowledge of the slope of the walls of the groove 104.

[0045] Commands can be sent to the welding control system to minimize the position error of the tungsten electrode 105 and the position error of the end of the filler wire 110, or to keep these errors small enough such that the welding quality will be acceptable. The estimated depth of the groove 104 can be used to select the thickness of the metal layer to be deposited during the welding of the current weld pass, such that after completion of the welding of the current weld pass, the remaining depth of the groove 104 is equal to (or approximately equal to) the target depth. For example, if a flush weld (e.g., a weld with a bead having no crown) is to be produced, the target thickness of the metal layer to be deposited during the welding of the current weld pass can be calculated by dividing the estimated depth of the groove 104 by the number of remaining weld passes (including the current weld pass).

[0046] The estimated width of the groove 104 can be used to calculate the rate of metal deposition per unit length of the weld, which corresponds to the target thickness of the metal layer to be deposited during the welding of the current weld pass. The wire feed rate and the travel speed can then be adjusted to achieve this rate of deposition (discussed further in detail below), and the heating power can be adjusted based on the wire feed rate or the travel speed by adjusting the voltage or current (or both) of the welding power source. In the case of using a weaving (oscillating) motion pattern, the width can also be used to calculate the oscillation amplitude, speed, and dwell time. In the case of using a split bead weld formation (multiple beads per layer), the width can also be used to calculate the position of each bead weld relative to other bead welds or the sides of the weld groove 104.

[0047] Reference Figure 3A , several interconnected neural networks can be employed to estimate the position of the tungsten electrode 105, the position of the end of the filler wire 110, the position of the intersection corresponding to the upper edge of the groove 104, and the position of the intersection corresponding to the lower edge of the groove 104. The first neural network 305 receives an image from a welding monitoring camera and estimates the coordinates of the point that defines the position of the tip of the tungsten electrode 105. The estimated position of the tip of the tungsten electrode 105 is employed by the welding control system as described above, and the estimated position of the tip of the tungsten electrode 105 is also fed into a second neural network 310, which can use both the estimated position of the tip of the tungsten electrode 105 and the image to estimate the position of the end of the filler wire 110. The third neural network 315 can receive the image and estimate the position of the upper edge of the groove 104 (e.g., it can estimate the horizontal coordinate of the intersection corresponding to the upper edge of the groove 104), and the fourth neural network 320 can receive (i) the estimated horizontal coordinate of the intersection corresponding to the upper edge of the groove 104 and (ii) the image, and it can estimate the position of the lower edge of the groove 104 (e.g., it can estimate the horizontal coordinate of the intersection corresponding to the lower edge of the groove 104). Figure 3BShows the internal structures of the first neural network 305 and the second neural network 310 in some embodiments. The neural network can be implemented in a processing circuit (discussed in further detail below), or in an analog circuit, or in a hybrid analog and digital circuit.

[0048] Supervised training can be used to train each neural network. For this training, a set of labeled images can be generated by labeling the images obtained from the welding monitoring camera with the coordinates of the points of interest during the welding process. The labeling can be performed manually by an experienced operator. The size of the training dataset can be increased by adding modified versions of the labeled images to the training dataset. For example, a previously labeled image can be horizontally translated by a certain amount, and the horizontal coordinates of the labels can be adjusted accordingly to create additional labeled images. The result of the training may be a set of filters, weights, and biases (or a neural network model) that, when loaded into the neural network, cause the neural network to estimate the coordinates of the points of interest in the image of the ongoing welding. The set of weights, filters, and biases can be referred to as a "model".

[0049] Before welding begins, a similar method can be used to train the neural network to identify (and find the coordinates of) points of interest from an image of the welding setup of the welding monitoring camera. Such an image may be different from the image of the ongoing welding, its illumination may be different (provided by a suitable light source rather than mainly by the arc), and there may be no weld pool (and the point where the end of the filler wire 110 enters the weld pool). In some embodiments, Figure 3A and Figure 3B the neural network is trained twice, once with an image of the ongoing weld and once with an image of the welding setup before welding begins, to produce a first model and a second model respectively. Then, the second model (i.e., loaded into the neural network) can be used to align the tungsten electrode 105 and the filler wire 110 with the groove 104 before the weld begins. Once this alignment is complete, the first model can be loaded into the neural network, the arc can be initiated, and the weld can be performed using the first set of weights.

[0050] In a neural network-controlled welding system, or in a welding system without such control, the travel speed can be calculated based on the filler wire diameter, the cross-sectional area of the weld layer 106, and the filler wire feed speed using the formula derived below.

[0051] In terms of maintaining the volume of the filler metal, the rate at which the volume of filler metal is added to the weld pool by the filler wire 110 is equal to the rate at which the filler metal is removed from the weld pool in the form of the completed weld layer 120. The rate of adding the filler metal volume can be calculated as the product of the cross-sectional area A f of the metal of the filler wire 110 f and the filler wire feed rate v

[0052] A f v f

[0053] Among them, for circular wire (i.e., the wire for wire feeding with a solid circular cross-section), the cross-sectional area A f is given by the following formula:

[0054]

[0055] where φ is the diameter of the filler wire 110. The rate at which metal is carried away from the weld is:

[0056] A L v t

[0057] where A L is the cross-sectional area of the weld layer and v t is the travel speed. For a weld layer with a rectangular weld groove, the cross-sectional area can be written as

[0058] A L = WT

[0059] where W is the width of the groove (and the layer), and T is the thickness or "height" of the layer.

[0060] Setting the rate of adding metal equal to the rate of carrying away metal results in the following:

[0061]

[0062] Therefore, given the weld width, layer height, cross-sectional area of the filler electrode, and wire feeding rate of the filler wire, it is possible to determine the travel speed to achieve the required layer height according to formula (1) (the lower the travel speed, the greater the layer height, and the greater the travel speed, the smaller the layer height).

[0063] If the filler wire 110 is hollow (e.g., a hollow, flux-cored wire) instead of a solid wire, the cross-sectional area of the metal of the filler wire 110 can be, for example, 25% of the total cross-sectional area of the filler wire 110 (75% of the cross-sectional area is flux), i.e., where m is a correction factor (or "efficiency factor") that takes into account that only a part of the filler wire 110 is metal, and formula (1) can be modified to:

[0064]

[0065] (where, for example, for a filler wire 110 with 75% flux and 25% metal by volume, m = 0.25).

[0066] If the volume of the filler metal is not maintained (e.g., in cases where some filler metal is lost due to spatter or oxidation or other mechanisms such as changes in crystal structure that affect density), Equation (1) may not hold exactly. Equation (1) may also not hold exactly if the height of the weld layer is large enough for some filler metal to overflow the groove 104. However, notwithstanding this, if the equation approximately holds, it can still be used with satisfactory results, or if the degree of deviation of the process from the ideal is known, the equation can be corrected. For example, Equation (2) can be used and also (or instead) a correction factor m can be used to compensate for the various mechanisms that cause Equation (1) not to hold exactly. For example, if 5% of the metal of the filler wire 110 is lost due to spatter, a value of m = 0.95 can be used (or, if m is different from 1 for other reasons, the value of m can be reduced by 5% to account for the spatter loss).

[0067] If the ends of the pipe are shaped such that the groove is V-shaped, for each pass around the pipe, the weave amplitude can be increased such that the width of the weld is substantially equal to the width of the bottom of the partially filled V. Then Equation (1) can be used, where W is the width of the bottom of the partially filled V (or, for more precision, W is the width of the V at a height of half the layer thickness above the bottom of the partially filled V). In some embodiments, the substrate is any substrate (e.g., a plate to be joined or a substrate on which a metal overlay is to be formed), and the travel speed can be calculated using Equation (1) or Equation (2) based on the filler wire diameter and the filler wire feed rate, and based on the height and width of the weld layer 120 to be formed (or, more generally, if the filler wire 110 is not round and solid, or if the cross-section of the weld layer 120 is not rectangular, using ).

[0068] In operation, the welding system can calculate the target travel speed to be used and automatically set the travel actuator (which may cause the pipe to rotate or the welding head to travel around the outside of the pipe at the joint) to produce travel at the target travel speed. The system can obtain the input variables in Equation (1) from various sources, including from operator input (e.g., supplied to the welding system through an input device such as a keyboard) or from sensors. For example, the operator can instruct the welding system to adjust the filler wire feed rate to increase or decrease the filler wire feed rate until the weld pool has the desired characteristics. Thus, the welding system can learn the filler wire feed rate based on the operator's most recent instructions. The width of the weld can be measured by the welding system using the machine learning-based image analysis disclosed herein or by other methods.

[0069] Figure 4A welding system is shown that can implement travel speed control based on the above formula. In some embodiments, this welding system includes a welding head 410, which includes a filler wire feeding unit 412 for supplying the filler wire 110 to the welding pool 205. A system for heating the welding pool is not explicitly shown; it can be, for example, an electric arc (with current flowing through the filler wire 110, or through a separate (e.g., tungsten) electrode), or a laser. The system can further include a travel actuator 415 for controlling the relative movement of the welding head 410 and the substrate 110. The welding system can also include a control circuit 425, including (i) a travel actuator drive circuit 430, which is used to interface to the travel actuator 415; (ii) a wire feeding drive circuit 435, which is used to control the wire feeding speed of the filler wire 110; and (iii) a processing circuit 445 (described in further detail below), which is used to perform high-level control functions, such as commanding the travel actuator 415. The system can also include other elements (not shown), such as a welding power source and a controller, and one or more user interface devices, such as a display, a keyboard, or a mouse.

[0070] As used herein, a "portion" of something means "at least a portion" of that thing, and thus may mean less than the whole of that thing or may mean the whole of that thing. Thus, as a special case, a "portion" of a thing includes the whole of that thing, i.e., the whole of the thing is an example of a portion of that thing. As used herein, the word "or" is inclusive, and thus, for example, "A or B" means any one of the following: (i) A, (ii) B, and (iii) A and B.

[0071] The term "processing circuit" is used herein to mean any combination of hardware, firmware, and software for processing data or digital signals. Processing circuit hardware can include, for example, application-specific integrated circuits (ASICs), general or special-purpose central processing units (CPUs), digital signal processors (DSPs), graphics processing units (GPUs), and programmable logic devices, such as field-programmable gate arrays (FPGAs). In a processing circuit, as used herein, each function is either performed by hardware configured to perform that function (i.e., hardwired) or by more general hardware, such as a CPU, configured to execute instructions stored in a non-transitory storage medium. The processing circuit can be fabricated on a single printed circuit board (PCB) or distributed across multiple interconnected PCBs. The processing circuit can contain other processing circuits; for example, a processing circuit can include two processing circuits, an FPGA and a CPU, interconnected on a PCB.

[0072] Although exemplary embodiments of systems and methods for controlling a welding system have been specifically described and illustrated herein, many modifications and variations will be apparent to those skilled in the art. Accordingly, it is to be understood that the systems and methods for controlling a welding system constructed in accordance with the principles of this disclosure may be embodied in ways different from those specifically described herein. The invention is also defined in the appended claims and their equivalents.

Claims

1. A system for welding, characterized in that, Comprising: A welding head; A first welding monitoring camera; and A machine learning system, wherein the machine learning system is configured to estimate the position of the end of the filler wire relative to the groove when the welding head forms a welding layer in the groove.

2. The system according to claim 1, wherein The estimation includes estimating the position of a first point of interest based on an image obtained by the first welding monitoring camera, where the first point of interest is the intersection of the first edge of the groove and a reference plane.

3. The system according to claim 2, wherein The estimation includes estimating the positions of four points of interest including the first point of interest, and each of the four points of interest is the intersection of the edge of the groove and the reference plane.

4. The system according to claim 3, characterized in that, The estimation of the position of the first point of interest includes estimating the intersection in the image of the line corresponding to the first edge and the line corresponding to the reference plane.

5. The system according to claim 4, characterized in that, The machine learning system is further configured to estimate the depth of the groove.

6. The system according to claim 4 or 5, characterized in that, The machine learning system is further configured to estimate the width of the bottom of the groove.

7. The system according to any one of claims 4 to 6, characterized in that The estimation further includes estimating the position of a fifth point of interest corresponding to the end of the filler wire.

8. The system according to any one of claims 4 to 7, characterized in that, The machine learning system is further configured to estimate the error of the position of the end of the filler wire relative to the target position of the end of the filler wire.

9. The system according to any one of claims 4 to 8, characterized in that, The estimation further includes estimating the position of a sixth point of interest corresponding to the tip of the tungsten electrode of the welding head.

10. The system according to claim 9, characterized in that, The machine learning system is further configured to estimate the error of the position of the tip of the tungsten electrode relative to the target position of the tip of the tungsten electrode.

11. The system according to any one of the preceding claims, wherein The machine learning system includes a first neural network for estimating the position of the tungsten electrode in an image obtained by the first welding monitoring camera.

12. The system according to any one of the preceding claims, wherein The machine learning system further includes a second neural network for estimating the position of the end of the filler wire.

13. The system according to any one of the preceding claims, characterized in that, The estimation of the position of the end of the filler wire includes estimating the position of the end of the filler wire based on: The estimated position of the tungsten electrode; and The image.

14. The system according to any one of the preceding claims, characterized in that, The machine learning system further includes a third neural network for estimating the position of the upper edge of the groove.

15. The system according to any one of the preceding claims, characterized in that, The machine learning system includes a fourth neural network for estimating the position of the lower edge of the groove.

16. The system according to claim 15, wherein, The estimation of the position of the lower edge of the groove includes estimating the position of the lower edge of the groove based on: The estimated position of the upper edge of the groove; and The image.

17. A method, characterized in that, Comprising: Calculating the travel speed of the substrate relative to the welding head based on the following steps: The wire feeding speed of the filler wire; The size of the filler wire; and The size of the welding layer to be formed.

18. The method according to claim 17, wherein The size of the filler wire is the diameter of the filler wire.

19. The method according to claim 17 or 18, characterized in that, The size of the welding layer is the thickness of the welding layer, and The calculation is further based on the width of the welding layer.

20. The method according to any one of claims 17 to 19, characterized in that The calculation is further based on an efficiency factor.

Images