Automatic surfacing device and surfacing control method

Abstract

An automatic surfacing device, comprising a welding execution system, a visual identification system and a welding intelligent control system. The present invention further relates to a surfacing control method, comprising: establishing a corresponding rule table between welding process parameters and welding bead topography relationship parameters; establishing coordinate systems, planning a welding trajectory, and presetting welding parameters; starting welding, and compensating for the welding width and welding reinforcement of a welding bead in real time; and performing scanning and detection, and providing a reminder to repair the position of a defect and a reminder for depth information. In the method, on the basis of an MIG welding procedure, highly automated welding is performed on a plurality of welding layers (an underlying layer, a transition layer and a wear-resistant layer) of a surfacing roller sleeve, thereby improving the production efficiency. During a welding process, closed-loop controlling is performed on the quality of each welding link, thereby guaranteeing the quality consistency of welded finished products.

Description

Automatic surfacing welding device and surfacing welding control method Technical Field

[0001] The present invention relates to the technical field of surfacing welding, in particular to an automatic surfacing welding device and a surfacing welding control method. Background Art

[0002] Roller presses are core production equipment in industries such as cement and mining. Their performance is closely linked to product quality, output, and energy consumption. Roller sleeves are core components of roller presses and have a significant impact on their performance. During operation, roller sleeves come into direct contact with the material, crushing it and requiring high wear resistance and compressive strength. Currently, two main types of roller sleeves are available on the market: studded roller sleeves and welded roller sleeves. Studded roller sleeves are forged from high-wear-resistant alloy steel and offer a long service life but are more expensive, accounting for approximately 30% of the market share. Welded roller sleeves are manufactured by layering welding different welding wires. While their service life is relatively short, they are also more cost-effective and account for approximately 70% of the market share. Welding is a multi-pass welding process in which multiple weld passes are applied densely to the base material, creating a single layer. Continued welding is then applied to this layer, creating a continuous layer of weld passes until the weldment reaches the designed dimensions.

[0003] Hardfacing roller sleeves are constructed from a medium-carbon alloy steel base, such as 32CrMo, onto which wear-resistant materials are welded via a hardfacing process. From the inside out, the hardfacing roller sleeve can be divided into the sleeve base material, base layer, transition layer, wear-resistant layer, patterned layer, and hardened points. Different materials are used in each layer to achieve varying mechanical properties. The sleeve base material is machined from a cast steel billet, upon which all weld layers are stacked and welded. Therefore, the sleeve base material represents the initial form of the sleeve prior to welding. The base layer has high toughness, enhancing the bond between the workpiece and the hardened surface layer, relieving welding stresses and effectively preventing hardfacing cracks from propagating into the base material. The transition layer has high weld hardness and good toughness, offering excellent wear and extrusion resistance. The wear-resistant layer exhibits even greater wear resistance and extrusion wear resistance than the transition layer, and is constructed using materials with the highest weld hardness, with a surface hardness of HRC58 or higher. The patterned layer, hard points, and wear-resistant layer are made of the same material. Their primary function is to enhance the roller sleeve's ability to pull material into the material, improving production efficiency. During roller sleeve production, the base, transition, and wear-resistant layers account for over 90% of the cladding work. Once welded, internal defects in these layers are difficult to detect, and repairing them is costly and time-consuming. Therefore, researching the manufacturing process for cladding roller sleeves to improve weld quality and extend their service life is of great practical and economic value.

[0004] At present, the manufacture of surfacing roller sleeves on the market usually adopts manual welding or semi-automatic wear-resistant surfacing. This traditional surfacing method has the following defects:

[0005] 1. Poor quality consistency. The roller sleeve cladding process is complex and requires high welding experience and welding technology level of the welder. In addition, the workpiece is large and the welding time is long. Therefore, the welder's level and the welder's working state during welding have a great impact on the welding quality, making it difficult to control the welding quality and the consistency of the roller sleeve life cannot be guaranteed.

[0006] 2. The quality of the cladding process cannot be inspected and controlled. During the roller sleeve cladding process, the weld widths of multiple welds need to be kept strictly consistent, otherwise gaps will occur between the welds, and the subsequent welding of the next layer will fill the gaps, causing quality risks. Because multiple weld layers are stacked layer by layer, the excess height of each weld needs to be strictly consistent, otherwise the roller sleeve size will deviate, causing the cylindricity of the finished roller sleeve to not meet the requirements and the quality to be substandard. The existing cladding process cannot monitor and control the weld width and excess height parameters in real time during the welding process. The quality and life of the cladding roller sleeve can only be verified during use. Once there is a problem with the welding quality, stress concentration will occur in the defective parts, causing cracks, peeling, and falling of the roller surface in the early stage of use, resulting in huge after-sales maintenance costs.

[0007] 3. Low production efficiency and harsh operating environment. Currently, manual and semi-automatic cladding processes rely heavily on manual labor, resulting in low production efficiency. During the cladding process for wear-resistant roller sleeves, smoke and dust are extremely high, and the welding process is long. The harsh working environment can cause physical and chemical damage to workers, increasing the risk of occupational diseases.

[0008] Summary of the Invention

[0009] The present invention aims to provide an automatic cladding device and cladding control method to address the issues raised in the background art. Based on the MIG welding process, the device performs highly automated welding of multiple weld layers (base layer, transition layer, and wear-resistant layer) on a cladding roller sleeve, improving production efficiency. Closed-loop quality control is implemented throughout the welding process, ensuring consistent quality of the finished welded product.

[0010] To achieve the above object, the present invention provides the following technical solutions:

[0011] In one aspect, the present invention discloses a control method for automatic surfacing welding, comprising the following steps:

[0012] S1. Establish a rule table for correspondence between welding process parameters and weld bead morphology parameters, wherein the welding process parameters include: welding gun swing amplitude, wire feed speed, welding current, and welding voltage; and the weld bead morphology parameters include weld bead width and weld bead height.

[0013] S2. Planning the welding trajectory: The roller sleeve rotates at a constant speed, and the welding gun performs a reciprocating pendulum welding motion at the starting point. When the roller sleeve rotates at a constant speed for one circle and completes one weld bead, the welding gun immediately moves axially toward the roller sleeve by a standard weld bead width and starts welding the next weld bead until all welds of the current weld layer are completed. After the welding of the current weld layer is completed, the welding gun moves to the starting welding position of the next weld layer and performs the welding task of the next weld layer.

[0014] Preset working parameters, including: number of welding layers, welding rod brand, welding speed, standard value of weld width W set , the allowable value of weld width deviation, the standard value of weld height H set , the allowable value of weld bead height deviation, and the threshold value of the area of ​​"true defects" on the roller sleeve surface;

[0015] S3, start the current layer welding task: control the roller sleeve and welding gun to follow the welding trajectory set in S2, and according to the preset weld width standard value W set , standard value of weld bead height H set Search the rule table to find and set the current welding process parameters;

[0016] Step S3 also includes real-time compensation of weld bead width and weld bead reinforcement, specifically including:

[0017] The molten pool camera monitors the current weld pool in real time, and calculates the current weld width W in real time based on the molten pool image detection algorithm. now , current weld bead height H now , and the morphology parameter deviation is obtained ΔW represents the current weld width value W now The standard value of weld width W set The difference between the two values, ΔH represents the current weld bead height value H now The standard value of weld height H set The difference between

[0018] If ΔW is not greater than the allowable value of weld width deviation, and ΔH is not greater than the allowable value of weld height deviation, then continue with the established welding process parameters. Otherwise, start the correction program, readjust the welding process parameters and continue welding until the current layer is completed.

[0019] S4. Use a line laser scanning camera to photograph the current weld layer and perform defect detection, specifically including:

[0020] If there is no defect, directly proceed to the next welding layer welding task;

[0021] If there is a defect, a defect warning will be issued, and after the repair is completed, the next welding layer welding task will be started;

[0022] S5. Repeat steps S3-S4 until all welding layers are completed.

[0023] As a further solution of the present invention: the process of establishing the rule table includes:

[0024] Make the welding wire brand and welding speed consistent with the actual welding; set m welding gun swings B1, B2, ..., B from small to large m , the interval between adjacent welding gun swings is δB; set n wire feeding speeds V1, V2, ..., V from small to large n , the interval between adjacent wire feeding speeds is δV; let the welding current I and welding voltage U be according to the empirical formula Follow the changes in wire feeding speed V; where I is in amperes, V is in centimeters per minute, and U is in volts; each time a set of welding process parameters (B, V, I, U) is changed, a corresponding set of weld bead morphologies (W, H) is obtained. After multiple experiments to obtain multiple sets of data, a rule table is obtained.

[0025] As a further solution of the present invention: the correction procedure specifically includes:

[0026] Establish the relationship rules between the adjustment direction of welding process parameters and ΔW and ΔH; determine the adjustment direction of welding process parameters according to the current ΔW and ΔH; in the rule table, take the current welding process parameters B0 and V0 as the starting point, determine the adjustment direction of welding process parameters B and V, and use δB and δV as the search interval to find the distance from the weld bead morphology parameters. The nearest point, record the welding process parameter group at this time (B new ,V new ,I new ,U new ); where W new Represents W set The difference with ΔW, H new Represents H set The difference with ΔH;

[0027] Adjust the welding gun swing to B new , wire feeding speed is V new , welding current is I new , welding voltage is U new , continue welding according to the preset welding trajectory.

[0028] As a further solution of the present invention: the relationship between the welding process parameter adjustment direction and ΔW and ΔH is as follows:

[0029] If ΔW>0 and ΔH=0, reduce the welding gun swing amplitude and wire feeding speed at the same time;

[0030] If ΔW < 0, ΔH = 0, then increase the welding gun swing amplitude and wire feeding speed at the same time;

[0031] If ΔW=0, ΔH>0, increase the welding gun swing amplitude and reduce the wire feeding speed;

[0032] If ΔW=0 and ΔH<0, reduce the welding gun swing amplitude and increase the wire feeding speed;

[0033] If ΔW < 0, ΔH < 0, then maintain the original welding gun swing and increase the wire feeding speed;

[0034] If ΔW>0, ΔH>0, maintain the original welding gun swing and reduce the wire feeding speed;

[0035] If ΔW>0 and ΔH<0, reduce the welding gun swing amplitude and maintain the original wire feeding speed;

[0036] If ΔW<0, ΔH>0, increase the welding gun swing amplitude and maintain the original wire feeding speed.

[0037] As a further solution of the present invention: in step S4, executing the defect detection process specifically includes:

[0038] S41, the roller sleeve continues to rotate at a uniform speed, and the line laser scanning camera moves at intervals along the axial direction of the roller sleeve, taking a picture each time it moves, so that all the pictures taken cover the entire roller surface;

[0039] S42, cutting and splicing all the pictures to form a large roll surface picture;

[0040] S43, detecting and extracting the position and depth information of the defect in the large image of the roller surface, and converting the defect position to the welding robot coordinate system;

[0041] S44. Display and remind the user to repair the defect.

[0042] As a further solution of the present invention: in step S43, the depth information is the roller sleeve radius corresponding to the grayscale value of each pixel point in the large roller surface image, and the roller sleeve radius = the distance from the roller sleeve axis when the line laser scanning camera is shooting - the grayscale value of the corresponding pixel point on the large roller sleeve image.

[0043] As a further solution of the present invention: in step S43, the process of detecting and extracting defects specifically includes:

[0044] The attribute characteristic values ​​of all pixels in the roller surface image are set to 0. According to the theoretical radius of each point on the roller sleeve = the original radius of the welding base material + the standard value of the single-layer weld bead height × the number of welded layers ± the allowable value of the weld bead height deviation, the grayscale value of each pixel in the roller surface image is checked one by one to see if it is equal to the theoretical radius. If not, the attribute characteristic value of this pixel is set to 1; the pixel points with an attribute value of 1 are checked to see if they are connected. If so, they are added to the same data set; the above process is repeated until all pixels are checked to obtain multiple unconnected data sets, and the defect area corresponding to the number of pixels in each data set is calculated; if the defect area of ​​the data set is greater than the preset "true defect" area threshold of the roller sleeve surface, it is recorded as a "true defect", otherwise it is recorded as a "false defect"; let the pixel coordinates of a pixel point of "true defect" A in the roller surface image be (x pix ,y pix ), whose gray value is h pix , then according to Transform it into the cylindrical coordinate system C1 of the roller sleeve, and according to the rotation matrix R and translation matrix T, transform the pixel coordinates of the "true defect" into the welding robot coordinate system C2, that is, C2=RC1+T; wherein: ρ represents the distance from a certain pixel point of the "true defect" A to the origin in the cylindrical coordinate system C1, θ represents the rotation angle of a certain pixel point of the "true defect" A relative to the x-axis of the cylindrical coordinate system C1, z represents the projection length of a certain pixel point of the "true defect" A on the z-axis of the cylindrical coordinate system C1, and N represents the total number of pixel rows in the large image of the roller surface.

[0045] Another aspect of the present invention discloses an automatic welding device used in any of the control methods described above, comprising a welding actuator, a visual recognition mechanism, and a welding intelligent control mechanism, wherein:

[0046] The welding actuator includes a welding robot and a welding gun fixed to the end of the welding robot; a positioner for supporting the roller sleeve and driving it to rotate at a constant speed, a welding machine for providing a stable current, and a wire feeder for continuously feeding welding wire into the nozzle of the welding gun at a set speed during the welding process;

[0047] The visual recognition mechanism includes a molten pool camera and a line laser scanning camera fixed to the end of the welding robot;

[0048] The welding intelligent control mechanism includes an image processing unit electrically connected to the molten pool camera and the line laser scanning camera, a welding control unit electrically connected to the welding robot, wire feeder, welder, and positioner, and a human-computer interaction unit.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] This invention improves the automation level and production efficiency of the roller sleeve welding process and enhances the operator's working environment. The use of a welding robot in conjunction with a positioner enables automated welding of large rotary weldments, replacing manual and semi-automatic welding methods. This significantly enhances production automation and increases production efficiency. It also improves the operator's working environment, freeing them from the high smoke, arc light, and noise associated with on-site welding, reducing their labor intensity and workload.

[0051] This invention uses automated welding to reduce quality issues caused by human factors. A weld pool camera monitors changes in weld width and bead height during the welding process. A weld bead topography parameter control model is established, and a negative feedback control strategy for coordinated control of multiple process parameters is employed to ensure that the weld bead topography is consistently maintained within a set range. During the adjustment process, process parameters such as welding current, welding voltage, and wire feed speed can be continuously and rapidly adjusted according to the control model, enabling precise control that is impossible with manual and semi-automatic welding, ensuring high controllability and consistency of weld bead topography. A weld layer quality inspection method is proposed. After each weld layer is completed, a line laser scanning camera is used to detect the three-dimensional topography of the weld layer. This method promptly alerts the operator to repair any defects on the roller surface, preventing them from being "buried" when welding the next layer. This eliminates the potential for weld quality defects and ensures quality control during the welding process. The scan results accurately display the defect area, depth, and coordinates in a coordinate system, providing an intuitive reference for the operator and potentially enabling future automated repair of weld layer defects using welding robots.

[0052] Facilitates quality traceability and process improvement. Automatically recording welding process parameters and weld bead topography during welding facilitates process tracking and traceability of welding quality issues. Alarms are triggered during welding if process parameter data exceeds limits, ensuring high reliability. By accumulating extensive production data, massive data mining can be performed to replicate high-quality welding processes, continuously improving process capabilities and enhancing welding quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] FIG1 is a schematic structural diagram of the present invention;

[0054] FIG2 is a flow chart of a surfacing welding control method according to an embodiment of the present invention;

[0055] FIG3 is a diagram showing the relationship between the adjustment direction of welding process parameters and the change direction of weld bead morphology parameters in the present invention;

[0056] In the figure: 1- welding robot, 2- welding gun, 3- molten pool camera, 4- line laser scanning camera, 5- positioner, 6- wire feeder, 7- welding machine, 8- image processing unit, 9- welding control unit, 10- human-machine interaction unit, 11- roller cover. DETAILED DESCRIPTION

[0057] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0058] In the description of the present invention, it should be noted that the terms "upper", "lower", "left", "right", "inside", "outside", etc. indicate directions or positional relationships based on the directions or positional relationships shown in the accompanying drawings, or are directions or positional relationships in which the inventive product is usually placed when in use. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction. Therefore, they should not be understood as limiting the present invention.

[0059] In the description of the present invention, it should also be noted that, unless otherwise expressly specified or limited, the terms "disposed," "installed," "connected," and "connected" should be understood in a broad sense. For example, they may refer to fixed, detachable, or integral connections; mechanical or communication connections; direct or indirect connections through an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention in specific contexts.

[0060] Please refer to Figures 1-3, an embodiment of the present invention, an automatic surfacing device includes a welding actuator, a visual recognition mechanism, and a welding intelligent control mechanism, wherein the welding actuator includes:

[0061] The welding robot 1 is preferably a six-axis industrial robot arm that supports offline programming and online programming and can communicate and receive external control instructions to operate during the welding process;

[0062] The welding gun 2 is fixed to the end of the sixth axis of the welding robot 1. When the MIG welding process is adopted, the welding wire passes through the middle of the welding gun, and the welding gun plays the role of clamping and fixing the welding wire.

[0063] The wire feeder 6 is fixedly mounted on the body of the welding robot 1. The welding wire passes through the wire feeder 6. It ensures that the welding wire moves continuously and stably forward to the nozzle of the welding gun 2 during welding, thereby ensuring that the arc is not interrupted or extinguished. It can receive external signals to control the wire feeding speed.

[0064] Welding machine 7, used to control the arc starting and extinguishing of the welding wire and maintain the current during the welding process. It can receive external signals to control the current and voltage during the welding process;

[0065] The positioner 5 is used to position, clamp, and support the roller sleeve 11 to be welded, and to rotate the roller sleeve 11 at a set angular velocity. It is equipped with an encoder that detects the angular velocity and angular displacement of its own shaft and can receive external speed and position control signals.

[0066] Visual recognition mechanism, used to detect and provide feedback on the roller surface status during the welding process, specifically including:

[0067] A line laser scanning camera 4 is mounted on the end of the welding robot 1 via a camera bracket. This line laser scanning camera uses the principle of triangulation, and the pixel grayscale information of the captured image stores the appearance of the photographed object. The line laser scanning camera can be activated by an external signal, and its image results can be transmitted to the outside via electrical signals.

[0068] The molten pool camera 3 is similarly mounted on the end of the welding robot 1 via a camera bracket and is used to capture the molten pool in real time during welding. The molten pool camera's accompanying software includes a molten pool morphology detection model. Based on the molten pool morphology at the current real-time capture location, it predicts and outputs parameters such as the weld width, residual height, and wire stickout of the weld bead after condensation. Detection accuracy reaches 0.1mm, and the detection results can be output via communication.

[0069] The welding actuator is used to control the welding execution system to move according to the set trajectory and dynamically adjust the welding execution system action according to the detection results of the visual recognition system. It specifically includes:

[0070] The image processing unit 8 is electrically connected to the line laser scanning camera 4 and the molten pool camera 3, and is used to receive and process image and video information, and issue start-up instructions to the line laser scanning camera 4 and the molten pool camera 3. It is also electrically connected to the welding control unit 9, and can send the image and video processing results to the welding control unit 9 and receive control instructions from the welding control unit 9;

[0071] The welding control unit 9 is electrically connected to the welding robot 1, wire feeder 6, welder 7, and positioner 5 to control these devices to perform various actions and receive their feedback signals; it is electrically connected to the image processing unit 8 to receive its image and video processing results and issue start-up instructions to it; it is electrically connected to the human-computer interaction unit 10 to receive external manual input of control instructions and parameters, and feed back the current system status parameters to the human-computer interaction unit;

[0072] The human-computer interaction unit 10 is a medium for the operator to interact with the system. It can be a hardware device such as a touch screen or a computer. It can be set locally or remotely and is electrically connected to the welding control unit 9 to exchange information.

[0073] An automatic surfacing control method comprises the following steps:

[0074] S1. Establish a rule table for the relationship between welding process parameters and weld bead morphology, and determine the quantitative relationship between the two. Welding process parameters include welding gun swing B, wire feed speed V, welding current I, and welding voltage U; weld bead morphology parameters include weld bead width W and weld bead height H. Through experiments, the quantitative relationship between welding process parameters and weld bead morphology can be determined, and the welding wire brand can be controlled to be consistent with the actual welding. Suppose that the swing of welding gun 2 is set from small to large to m, including B1, B2, ..., B m , the adjustment interval is δB. Control the wire feeding speed to set n from small to large, including V1, V2, ..., V n , the adjustment interval is δV. Welding current I and welding voltage U are calculated according to the empirical formula The wire feed speed V varies. I is in amperes, V is in centimeters per minute, and U is in volts. During the experiment, changing one set of parameters (B, V, I, U) yields a corresponding set of weld bead morphologies (W, H). By performing multiple experiments with varying intervals δB and δV, and obtaining multiple sets of data, we can create the rule table shown in Table 1.

[0075] As shown in Figure 3, the direction of adjustment of welding process parameters corresponds to the direction of change in weld bead topography parameters. Assuming the current welding process parameters are (B0, V0, I0, U0), if the current weld bead width W0 is low while the weld bead reinforcement H0 is normal, adjustments should be made toward increasing the weld bead width (direction W↑). According to Figure 3, the welding gun swing amplitude B should be increased, the wire feed speed V should be increased, and the welding current I and welding voltage U should be adjusted synchronously according to the empirical formula. Combined with Table 1, the quantitative impact of the amplitude of process parameter adjustment on the amplitude of weld bead topography change is determined, thus establishing a weld bead topography parameter control model.

[0076] The "corresponding rule table between process parameters and weld bead morphology" and the "relationship between welding process parameter adjustment direction and weld bead morphology parameter change direction" are converted into rules and stored in the welding control unit 9 to achieve the target change of weld bead morphology and the linkage control of multiple welding process parameters.

[0077] Table 1

[0078] S2. Establish a coordinate system, plan the welding trajectory, and preset welding parameters, including:

[0079] Establishing a coordinate system: Before welding, the roller sleeve coordinate system C1 and the robot coordinate system C2 must be established, and the conversion relationship between them must be clarified. As shown in Figure 1, coordinate systems C1 and C2 are both right-handed rectangular coordinate systems. The Z axis is the axial direction of the roller sleeve 11, the X axis is the direction perpendicular to the Z axis in the horizontal plane, and the Y axis is the vertical direction perpendicular to the horizontal plane. Coordinate system C1, with the center of the end face of the roller sleeve 11 as its origin, is called the workpiece coordinate system. The origin of coordinate system C2 is placed at the rotation center of the first axis of the welding robot 1 and is called the robot coordinate system C2. Points on the roller sleeve 11 can then be converted to the robot coordinate system using the rotation matrix R and the translation matrix T, i.e., C2 = RC1 + T. To obtain the rotation matrix R and translation matrix T, mark four points with known coordinates in the workpiece coordinate system C1. Move the welding wire tip at the end of the sixth axis of the welding robot until it touches each of these four points. Note the coordinates of these four points in the robot coordinate system C2. This yields two sets of coordinates for the same four points in the workpiece coordinate system C1 and the robot coordinate system C2. Based on these two sets of coordinates, use the SVD singular value decomposition algorithm to obtain the rotation matrix R and translation matrix T for the transformation between the two coordinate systems.

[0080] Planning the welding trajectory: The positioner 5 rotates at a set constant speed, and the welding robot 1 moves to the welding position of the first weld bead above the roller sleeve 11, driving the welding gun 2 to perform reciprocating welding motion at this point. When the positioner 5 rotates to its encoder 0 position, the welding control unit 9 controls the wire feeder 6 to start wire feeding, and the welder 7 starts the arc to start welding. After the positioner 5 rotates one circle, the molten welding wire forms a circle of weld bead on the roller sleeve 11. At this time, the welding robot 1 immediately translates a distance Z0 in the negative direction of the Z axis of the C1 coordinate system and remains fixed to start welding the next weld bead. Assuming that the axial length of the roller sleeve 11 is L, the movement of the welding robot 1 After the welding of all the welds of the single layer is completed, the rotation speed of the positioner 5 needs to be adjusted before starting the welding task of the next weld layer. Assuming that the diameter of the roller sleeve 11 before welding is D1, the diameter of the roller sleeve 11 after welding is D2, and the rotation speed of the positioner 5 before welding is n1, the rotation speed of the positioner is adjusted to Therefore, when the roller sleeve 11 is welding different welding layers, the relative speed between the roller sleeve 11 and the welding gun 2 is constant, that is, the welding speed is controlled to be constant. Assuming that the thickness of a certain welding layer is H and the residual height of a single weld is h, the welding layer needs to be welded After completing the welding task of this cladding layer, the welding wire of the corresponding brand of the next cladding layer is replaced and the welding of the next cladding layer is started until all cladding layers are completed. The motion control program of the welding robot 1 and the positioner 5 is stored in the welding control unit 9.

[0081] The operator inputs welding parameters through the human-machine interface unit 10, which are stored in the welding control unit 9. The preset welding parameters specifically include the preset welding process parameters and welding morphology parameter standard values ​​and allowable deviation ranges, including: number of welding layers, welding rod brand, welding speed, weld bead width, weld bead height, weld bead width deviation allowable value, weld bead height deviation allowable value, and roller sleeve surface "true defect" area threshold.

[0082] S3. Welding is started through the human-machine interaction unit 10. The welding control unit 9 sends control instructions to the welding robot 1 and the positioner 5, so that the two start to move according to the welding trajectory preset in step S2.

[0083] The welding control unit 9 determines the weld bead width W according to the preset weld bead width W in step S2. set , weld bead height H set , automatically searches the internally stored "corresponding rule table of process parameters and weld bead morphology" (i.e. Table 1), finds the corresponding welding process parameter group (B0, V0, I0, U0), and controls the welding robot 1 so that the welding gun swing amplitude is B0, controls the wire feeding speed of the wire feeder 6 to be V0, the welding current of the welding machine 7 to be I0, and the welding voltage of the welding machine 7 to be U0.

[0084] Furthermore, step S3 also includes real-time compensation for weld bead width and weld bead height. When various operating conditions occur, such as uneven wire quality, welding speed deviation, and actuator motion errors, the weld bead width and height may deviate from the set values. The present invention can detect the degree of deviation of weld bead topography parameters in real time and compensate for the weld bead width and height in real time, thus achieving negative feedback control of weld bead topography. The specific method is as follows:

[0085] (1) The molten pool camera 3 monitors the molten pool of the current weld in real time and transmits it to the image processing unit 8. The image processing unit 8 calculates the molten pool width W of the current weld in real time based on the molten pool image detection algorithm. now 、Yu Gao H now , output to the welding control unit 9;

[0086] (2) The welding control unit 9 calculates the deviation between the current weld bead morphology parameters output by the molten pool camera and the set standard value If the deviation exceeds the permissible value of the weld width deviation or the permissible value of the weld height deviation set in step S2, the correction program is started;

[0087] (3) The welding control unit 9 determines the adjustment direction of the current welding process parameters (B0, V0, I0, U0) according to the current ΔW and ΔH deviation directions and the stored “relationship between welding process parameter adjustment direction and weld bead profile parameter change direction” rule (as shown in FIG3 );

[0088] (4) The welding control unit 9 takes the current welding process parameters B0 and V0 as the starting point for adjustment according to the "corresponding rule table of the relationship between process parameters and weld bead morphology", and searches for the position of the weld bead morphology parameters B0 and V0 according to the adjustment direction of the welding process parameters B and V determined in the previous step, with δB and δV as the search intervals. The nearest point, record the welding process parameter group at this time (B new ,V new ,I new ,U new );

[0089] (5) The welding control unit 9 controls the welding robot 1 so that the welding gun swing amplitude is B new , control the wire feeding speed of wire feeder 6 to be V new , control the welding current of welding machine 7 to be I new , control the welding voltage of welding machine 7 to be U new , continue welding the current welding layer according to the preset welding trajectory until it is completed.

[0090] S4. Scan and build a three-dimensional model of the current welding layer to detect welding defects and provide repair warnings.

[0091] After the welding of the current layer is completed, in order to avoid the defects of the current weld layer being "covered" when welding the next layer, which may lead to quality problems, it is necessary to immediately detect and repair the defects of the current weld layer. The process of detection and repair is:

[0092] (1) Calculate the number of pictures to be taken by the line laser scanning camera 4 and the shooting position. Let the line laser scanning camera 4 shoot directly at the axis of the roller sleeve 11, and the distance from the axis is d. When taking the first and last pictures, the edge of the line laser is just at the axial edge of the roller sleeve 11. Assuming that the axial length of the roller sleeve 11 is L, and the axial length of the roller sleeve 11 covered by a single picture taken by the line laser scanning camera 4 is l, then a total of The welding robot 1 drives the linear laser scanning camera 4 from the first shooting position and moves n-1 times along the axial direction of the roller sleeve 11, with the distance interval between each movement being kept equal. A picture is taken for each position moved. The n pictures taken can cover the entire roller surface, and two adjacent pictures taken along the axial direction of the roller sleeve 11 partially overlap.

[0093] (2) Take individual pictures one by one at the set position. After the welding control unit 9 controls the welding robot 1 to drive the line laser scanning camera 4 to a set shooting position, when the welding control unit 9 detects that the rotating shaft of the positioner 5 rotates to the encoder 0 point, it sends a shooting instruction to the image processing unit 8. The image processing unit 8 triggers the line laser scanning camera 4 to start taking pictures. Assuming that the shooting frame rate of the line laser scanning camera 4 is f and the rotation period of the roller sleeve 11 is T, the number of frames of each picture taken by the line laser scanning camera 4 is N=f·T, that is, the pixel height of each picture is N. Assuming that the circumference of the end face circle of the roller sleeve 11 being photographed is S, the pixel height resolution of each picture is The pixel width of each image is fixed, assuming it is M. Since the axial length of the covered roller sleeve 11 is l, the pixel width resolution of each image is

[0094] (3) Cut and splice the pictures to form a large picture of the roller surface. The image processing unit 8 can calculate the overlapping length of two adjacent pictures along the axial direction according to the axial length l of the roller sleeve 11 that each picture can cover, the number of pictures n, and the axial length L of the roller sleeve 11. Divide by the pixel width resolution of each image to get the pixel width of the overlapping part Then cut off the overlapping parts of each picture After the images have a width of 1 pixel, they are aligned and spliced ​​together in the height direction to form a complete large image of the roller surface, which covers the entire roller sleeve 11. The grayscale value of each pixel in the large image of the roller surface corresponds to the depth information of the corresponding point on the roller sleeve 11, that is, the distance d from the axis of the roller sleeve when the camera was shooting - the grayscale value of pixel A on the large image of the roller surface = the roller sleeve radius at the position corresponding to pixel A on the roller sleeve 11.

[0095] (4) The image processing unit 8 extracts defects from the large image of the roller surface and converts the coordinates of the defect position to the robot coordinate system. The process of defect detection and extraction is as follows: The theoretical radius R of each point on the current roller sleeve 11 = the original radius of the welding base material + the standard value of the single-layer weld bead height × the number of welded layers ± the allowable value of the weld bead height deviation; the attribute feature values ​​of all pixel points in the large image of the roller surface are set to 0. Check one by one whether the roller sleeve radius corresponding to the M·N pixel points in the large image of the roller surface is equal to the theoretical radius. If not, the attribute feature value of this pixel is set to 1; check whether the pixel points with an attribute value of 1 are connected, that is, if there is another pixel point with an attribute feature value of 1 within the 8-neighborhood of a certain pixel point, if the two pixel points are considered to be connected, they are added to the same data set Set. Repeat the above process until all pixels are detected, and multiple data sets can be obtained. These data sets are not connected to each other and correspond to independent defects on the roller sleeve 11; count the number of pixels in each data set Set, and the number of pixels × pixel height resolution × pixel width resolution = defect area. The roller sleeve surface "true defect" area threshold set in step S2 is used for judgment. If the area of ​​the data set Set is larger than the set defect area threshold, it is recorded as a "true defect", otherwise it is considered a "false defect". Let the pixel coordinates of a certain pixel point of "true defect" A in the roller surface image be (x pix ,y pix ), whose gray value is h pix , then the transformation relationship to the cylindrical coordinate system C1 of the roller sleeve 11 is Where: ρ represents the distance from a pixel of the "true defect" A to the origin in cylindrical coordinate system C1, θ represents the rotation angle of a pixel of the "true defect" A relative to the x-axis of cylindrical coordinate system C1, z represents the projection length of a pixel of the "true defect" A on the z-axis of cylindrical coordinate system C1, and N represents the total number of pixel rows in the roller surface image. Similarly, the pixel coordinates of the "true defect" can be transformed to the robot coordinate system C2 using the rotation matrix R and translation matrix T described in step S2, i.e., C2 = RC1 + T. After completing the above operations, the coordinates and depth information of the "true defect" are transmitted to the welding control unit 9 for storage.

[0096] (5) Display and repair the "true defect". The welding control unit transmits the coordinate and depth information of the "true defect" calculated in step (4) to the human-machine interaction unit 10. The human-machine interaction unit 10 renders it and displays it in the roller sleeve coordinate system C1 or the robot coordinate system C2. The human-machine interaction unit 10 issues an early warning to the operator, reminding the operator to repair the discovered defects in time. When the operator completes the repair, clicks the confirm repair completion button in the human-machine interaction unit 10, indicating that the weld of this layer has been repaired and the welding task of the next weld layer can be carried out;

[0097] S5. Repeat steps S3-S4, continuing to weld one weld bead at a time and layer by layer until all weld overlay layers (base layer, transition layer, wear-resistant layer) are welded. During the welding process, the welding control unit automatically and periodically records welding process parameters and welding topography parameters such as welding voltage, current, welding speed, positioner rotation angle, and wire feed speed at set time intervals, and displays them in real time to the human-computer interaction unit, issuing an alarm if the limit is exceeded.

[0098] Although this specification is described according to implementation methods, not every implementation method contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

[0099] Therefore, the above description is only a preferred embodiment of the present application and is not intended to limit the scope of implementation of the present application; that is, all equivalent modifications made according to the scope of the claims of the present application are within the scope of protection of the claims of the present application.

Claims

1. A control method for automatic surfacing, characterized in that: The following steps are involved: S1. Establish a corresponding rule table of welding process parameters and weld bead morphology relationship parameters, wherein the welding process parameters include: welding gun swing amplitude, wire feeding speed, welding current, and welding voltage; and the weld bead morphology parameters include weld bead width and weld bead excess height; S2. Planning welding trajectory: the roller sleeve rotates at a constant speed, and the welding gun performs reciprocating pendulum welding motion at the starting point; when the roller sleeve rotates at a constant speed for one circle and completes one circle of weld, the welding gun immediately translates axially toward the roller sleeve by a standard weld width distance and starts welding the next weld until all welds of the current weld layer are completed; after the welding of the current weld layer is completed, the welding gun moves to the starting welding position of the next weld layer to perform the welding task of the next weld layer; Preset working parameters, including: number of welding layers, welding rod brand, welding speed, standard value of weld width W set , allowable value of weld width deviation, standard value of weld excess height H set , the allowable value of weld bead height deviation, and the threshold value of the area of ​​"true defect" on the roller sleeve surface; S3, start the current layer welding task: control the roller sleeve and welding gun to move according to the welding trajectory set in S2, and according to the preset weld width standard value W set , Standard value of weld bead height H set Search the rule table to find and set the current welding process parameters; Step S3 also includes real-time compensation of weld bead width and weld bead excess height, specifically including: The molten pool camera monitors the current weld pool in real time, and calculates the current weld width W in real time based on the molten pool image detection algorithm. now , current weld bead height H now , and the morphological parameter deviation is obtained ΔW represents the current weld width value W now The standard value of weld width W set The difference between the two values, ΔH represents the current weld bead height value H now The standard value of weld height H set The difference between If ΔW is not greater than the allowable value of the weld width deviation, and ΔH is not greater than the allowable value of the weld height deviation, continue to execute with the established welding process parameters; otherwise, start the correction program, readjust the welding process parameters and continue welding until the current layer welding is completed; S4. Use a line laser scanning camera to photograph the current welding layer and perform defect detection, including: If there is no defect, directly perform the welding task of the next welding layer; If there are defects, a defect warning will be issued, and after the repair is completed, the next welding layer welding task will be started; S5. Repeat steps S3-S4 until all welding layers are completed.

2. The control method according to claim 1, characterized in that: The process of establishing the rule table includes: Make the welding wire brand and welding speed consistent with the actual welding; set m welding gun swings B1, B2, ..., B from small to large m , the interval between adjacent welding gun swings is δB; n wire feeding speeds V1, V2, …, V are set from small to large n , the interval between adjacent wire feeding speeds is δV; let the welding current I and welding voltage U be according to the empirical formula Follow the changes of wire feeding speed V; the unit of I is ampere, the unit of V is centimeter / minute, and the unit of U is volt; every time a set of welding process parameters (B, V, I, U) is changed, a set of corresponding weld bead morphology (W, H) is obtained. After multiple experiments, multiple sets of data are obtained, and the rule table is obtained.

3. The control method according to claim 2, characterized in that: The correction procedure specifically includes: Establish the relationship between the adjustment direction of welding process parameters and ΔW and ΔH; determine the adjustment direction of welding process parameters according to the current ΔW and ΔH; in the rule table, take the current welding process parameters B0 and V0 as the starting point, and use the determined adjustment direction of welding process parameters B and V, and use δB and δV as the search interval to find the off-weld morphology parameters The nearest point, record the welding process parameter group at this time (B new ,V new ,I new ,U new ), where W new W set The difference with ΔW, H new Represents H set The difference with ΔH; Adjust the welding gun swing to B new , wire feeding speed is V new , welding current is I new , welding voltage is U new , continue welding according to the preset welding trajectory.

4. The control method according to claim 3, characterized in that: The relationship between the welding process parameter adjustment direction and ΔW and ΔH is as follows: If ΔW>0, ΔH=0, then reduce the welding gun swing amplitude and wire feeding speed at the same time; If ΔW<0, ΔH=0, then increase the welding gun swing amplitude and wire feeding speed at the same time; If ΔW=0, ΔH>0, increase the welding gun swing amplitude and reduce the wire feeding speed; If ΔW=0, ΔH<0, reduce the welding gun swing amplitude and increase the wire feeding speed; If ΔW<0, ΔH<0, keep the original welding gun swing and increase the wire feeding speed; If ΔW>0, ΔH>0, keep the original welding gun swing and reduce the wire feeding speed; If ΔW>0, ΔH<0, reduce the welding gun swing amplitude and maintain the original wire feeding speed; If ΔW<0, ΔH>0, increase the welding gun swing amplitude and maintain the original wire feeding speed.

5. The control method according to claim 1, characterized in that: In step S4, the defect detection process specifically includes: S41, the roller sleeve continues to rotate at a uniform speed, and the line laser scanning camera moves at intervals along the axial direction of the roller sleeve, taking a picture each time it moves, so that all the pictures taken cover the entire roller surface; S42, cutting and splicing all the pictures to form a large roll surface picture; S43, detecting and extracting the position and depth information of the defect in the large image of the roller surface, and converting the defect position to the welding robot coordinate system; S44. Display and remind to repair defects.

6. The control method according to claim 5, characterized in that: In step S43, the depth information is the roller sleeve radius corresponding to the grayscale value of each pixel point in the roller surface large image, and the roller sleeve radius = the distance from the roller sleeve axis when the line laser scanning camera is shooting - the grayscale value of the corresponding pixel point on the roller sleeve large image.

7. The control method according to claim 5, characterized in that: In step S43, the process of detecting and extracting defects specifically includes: The attribute characteristic values ​​of all pixels in the roller surface image are set to 0. According to the theoretical radius of each point on the roller sleeve = the original radius of the welding base material + the standard value of the single-layer weld bead height × the number of welded layers ± the allowable value of the weld bead height deviation, the grayscale value of each pixel in the roller surface image is checked one by one to see if it is equal to the theoretical radius. If not, the attribute characteristic value of this pixel is set to 1; check whether the pixels with an attribute value of 1 are connected. If so, add them to the same data set; repeat the above process until all pixels are detected to obtain multiple unconnected data sets, and calculate the defect area corresponding to the number of pixels in each data set; if the defect area of ​​the data set is greater than the preset "true defect" area threshold of the roller sleeve surface, it is recorded as a "true defect", otherwise it is recorded as a "false defect"; let the pixel coordinates of a certain pixel of the "true defect" A in the roller surface image be (x pix ,y pix ), whose gray value is h pix , then according to Transform it to the cylindrical coordinates of the roller sleeve In the cylindrical coordinate system C1, according to the rotation matrix R and the translation matrix T, the pixel coordinates of the "true defect" are transformed to the welding robot coordinate system C2, that is, C2=RC1+T; where: ρ represents the distance from a certain pixel point of the "true defect" A to the origin in the cylindrical coordinate system C1, θ represents the rotation angle of a certain pixel point of the "true defect" A relative to the x-axis of the cylindrical coordinate system C1, z represents the projection length of a certain pixel point of the "true defect" A on the z-axis of the cylindrical coordinate system C1, and N represents the total number of pixel rows of the roller surface image.

8. The automatic welding device used in the control method according to any one of claims 1 to 7, characterized in that: It includes welding actuator, visual recognition mechanism and welding intelligent control mechanism, among which: The welding actuator includes a welding robot and a welding gun fixedly mounted at the end of the welding robot; a positioner for supporting the roller sleeve and driving it to rotate at a constant speed, a welding machine for providing a stable current, and a wire feeder for continuously feeding welding wire into the nozzle of the welding gun at a set speed during the welding process; The visual recognition mechanism includes a molten pool camera and a line laser scanning camera fixedly mounted at the end of the welding robot; The welding intelligent control mechanism includes an image processing unit electrically connected to a molten pool camera and a line laser scanning camera, a welding control unit electrically connected to a welding robot, a wire feeder, a welder, and a positioner, and a human-machine interaction unit.

Images