Steel bracket multi-section automatic submerged arc welding path optimization method

Abstract

The present application relates to the technical field of automatic submerged arc welding, and particularly relates to a steel bracket multi-section automatic submerged arc welding path optimization method. The method takes the spatial inclination of all welds as the clustering object, performs a clustering analysis with constraints, and divides the welding layers. According to the spatial inclination and the number of welds in each welding layer, the target inclination is determined. For each welding layer, the spatial distance cost of the welding torch moving between any two welds in the welding positioner coordinate system is determined. According to the difference between the azimuth angles of any two welds in the welding layer, the positioner rotation cost is determined. According to the spatial distance cost and the positioner rotation cost, the switching comprehensive cost is determined. The welding sequence of the welds is determined based on the switching comprehensive cost between all the welds in the welding layer. The target inclination of each welding layer is sequentially adjusted to the corresponding target inclination. In each welding layer, the welds are sequentially welded according to the welding sequence, thereby achieving optimal planning of the complex steel bracket welding path.

Description

Technical Field

[0001] This invention relates to the field of automatic submerged arc welding technology, specifically to a method for optimizing the multi-segment automatic submerged arc welding path of a steel bracket. Background Technology

[0002] In the fields of steel structure construction, bridges, and heavy industrial equipment manufacturing, steel brackets serve as critical load-bearing and support components. Automatic submerged arc welding (SAW) has become the mainstream welding process for steel brackets due to its advantages such as deep penetration, high weld quality, and high degree of automation. However, this process has strict requirements on the welding position; the ideal welding positions are flat and horizontal. When the spatial angle of the weld deviates from this range, defects such as metal flow and incomplete fusion are easily caused by the gravity of the molten pool, severely reducing the weld qualification rate.

[0003] Currently, welding positioners and other clamping fixtures are commonly used to adjust the spatial posture of the steel bracket, thereby repositioning the inclined weld to a horizontal or near-horizontal position suitable for submerged arc welding. However, for complex steel brackets with multi-segment, multi-angle, and discretely distributed welds, how to plan the welding path has become a bottleneck in improving the efficiency and quality of automated welding. Summary of the Invention

[0004] To address the technical challenge of accurately planning the welding path during the welding of steel brackets, this invention provides a method for optimizing the multi-segment automatic submerged arc welding path of steel brackets. The specific technical solution adopted is as follows: This invention proposes a method for optimizing the multi-segment automatic submerged arc welding path of steel brackets, the method comprising: Obtain the geometric parameters of all weld seams to be welded on the steel support frame components in the coordinate system of the welding positioner. The geometric parameters include at least the spatial tilt angle and horizontal azimuth angle of each weld seam. Using the spatial tilt angles of all welds as clustering objects and the preset welding posture tolerance as a constraint, a constrained clustering analysis is performed to group welds with similar spatial tilt angles into the same welding layer, thereby dividing multiple welding layers; based on the distribution characteristics of the spatial tilt angles of welds within each welding layer, the target tilt angle of each welding layer is determined. For each welding layer, when the positioner is adjusted to the target tilt angle of the welding layer, the spatial distance cost of the welding torch moving between any two welds in the welding positioner coordinate system is determined; when the positioner is adjusted to the corresponding weld azimuth angle, the positioner rotation cost is determined based on the difference between the azimuth angles of any two welds in the welding layer; and the overall switching cost is determined based on the spatial distance cost and the positioner rotation cost. A cost matrix is ​​constructed based on the comprehensive switching cost between all welds within the weld layer, and the weld welding sequence with the lowest total cost is obtained through a path optimization algorithm. Adjust the positioner to the corresponding target tilt angle in sequence according to the target tilt angle of each welding layer; within each welding layer, adjust the positioner azimuth angle and welding torch position in sequence according to the welding sequence of the weld to complete the welding.

[0005] Furthermore, the constrained clustering analysis includes: Initialize multiple cluster centers and obtain the center tilt angle value of the cluster centers. The difference between the center tilt angle values ​​of any two initial cluster centers is greater than the preset welding posture tolerance. For each weld, the absolute difference between the spatial inclination angle of the weld and the inclination angle of each center is calculated as the interval angle; the welds are assigned to the cluster center corresponding to the smallest interval angle to form a cluster. Calculate the difference between the maximum and minimum spatial tilt angles within a cluster, and use this as the range of the cluster. Check whether the range of each cluster is greater than the preset welding posture tolerance. If there is a target cluster with a range greater than the preset welding posture tolerance, divide the target cluster into two sub-clusters with the median of all spatial tilt angles within the target cluster as the center. The assignment and inspection steps are executed iteratively until the range of all clusters is no greater than the preset welding posture tolerance, and each cluster is defined as a welding layer.

[0006] Furthermore, during the iteration process, if the weld space tilt angle ranges of two clusters overlap, and the range of the new cluster formed after merging is still not greater than the preset welding posture tolerance, then the corresponding two clusters will be merged.

[0007] Furthermore, the target tilt angle determination process includes: Calculate the arithmetic mean of the spatial inclination angles of all weld seams within the weld layer, and use it as the target inclination angle of the weld layer.

[0008] Furthermore, the process for determining the spatial distance cost includes: Based on the kinematic model of the welding positioner, the target welding coordinates of the starting and ending points of each weld in the welding positioner coordinate system are calculated when the positioner is adjusted to the target tilt angle of the welding layer and the corresponding weld azimuth angle. Calculate the three-dimensional Euclidean distance between the target welding coordinates of the termination point of any weld and the target welding coordinates of the start point of another weld; Based on the three-dimensional Euclidean distance, a path compensation amount is added to simulate the safe lifting and lowering operation of the welding torch, and the compensation result is used as the spatial distance cost.

[0009] Furthermore, the path compensation amount is twice the preset safety lifting height, which is used to simulate the vertical movement path length of the welding torch as it rises before moving and falls after moving.

[0010] Furthermore, the process for determining the rotation cost of the positioner includes: Calculate the absolute difference between the azimuth angles of any two welds within the weld layer, and use this as the positioner's rotation cost.

[0011] Furthermore, the process for determining the overall switching cost includes: The spatial distance cost and the positioner rotation cost are normalized respectively to obtain normalized spatial distance cost values ​​and normalized rotation cost values; The normalized spatial distance cost value and the normalized turning cost value are weighted and summed, and the sum is used as the overall switching cost.

[0012] Furthermore, the step of constructing a cost matrix based on the comprehensive switching cost among all welds within the weld layer, and then using a path optimization algorithm to solve for the weld welding sequence with the lowest total cost, includes: Initialize the parameters of the path optimization algorithm, including the pheromone matrix, heuristic information matrix, number of ants, and number of iterations; Based on the comprehensive switching cost between all welds within the weld layer, a cost matrix is ​​constructed, where each element in the cost matrix represents the comprehensive switching cost from one weld to another. Multiple ants are simulated to select the next weld seam to be welded in turn according to the state transition probability, until all weld seams are traversed, forming multiple complete welding paths. The state transition probability is determined by both pheromone concentration and heuristic information, and the heuristic information is set as the reciprocal of the overall switching cost. Based on the path quality of all ants in this iteration, update the pheromone concentration between each weld pair; Repeat the steps of constructing ant paths and updating pheromones until the maximum number of iterations is reached or the convergence condition is met. From all the paths generated by the iterations, the path with the lowest total switching cost is selected as the final weld welding sequence.

[0013] Further, the step of sequentially adjusting the positioner to the corresponding target tilt angle according to the target tilt angle of each weld layer includes: Adjust the welding positioner to the corresponding target tilt angle in order of increasing or decreasing target tilt angle values ​​for each weld layer.

[0014] The present invention has the following beneficial effects: This invention introduces a preset welding posture tolerance as a hard constraint for clustering and layering, ensuring that the spatial angle of each weld is precisely adjusted to the optimal range allowed by the submerged arc welding process during welding. This fundamentally avoids defects such as molten pool loss and incomplete fusion caused by excessive angle deviation, improving the weld qualification rate and structural safety. From a macroscopic perspective, constrained clustering analysis divides all welds into optimal welding layers, minimizing the number of tilt angle adjustments required by the welding positioner. From a microscopic perspective, within each welding layer, sequence optimization is performed by integrating the composite index of welding torch spatial distance cost and positioner rotation cost, significantly reducing the ineffective movement distance of the welding torch and the rotation time of the positioner. Finally, by first adjusting the positioner sequentially according to the target tilt angle of each welding layer, and then adjusting the positioner azimuth angle and welding torch position sequentially according to the weld welding sequence within each welding layer to complete the welding operation, the reliance on the programming experience of skilled technicians is greatly reduced, achieving optimal planning of welding paths for complex steel brackets. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart of a method for optimizing the multi-segment automatic submerged arc welding path of a steel bracket, provided in one embodiment of the present invention. Figure 2 This is a schematic diagram of a constrained clustering analysis process provided in one embodiment of the present invention. Detailed Implementation

[0017] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a multi-segment automatic submerged arc welding path optimization method for steel brackets proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0019] The following describes in detail, with reference to the accompanying drawings, a specific scheme for the multi-segment automatic submerged arc welding path optimization method for steel brackets provided by the present invention.

[0020] Please see Figure 1 The diagram illustrates a flowchart of a multi-segment automatic submerged arc welding path optimization method for steel brackets according to an embodiment of the present invention. The method includes: S101: Obtain the geometric parameters of all weld seams to be welded on the steel support frame in the coordinate system of the welding positioner. The geometric parameters include at least the spatial tilt angle and horizontal azimuth angle of each weld seam.

[0021] The welding positioner coordinate system is a right-handed rectangular coordinate system established with the physical rotation center of the welding positioner as the origin. For example, the Z-axis of the welding positioner coordinate system is defined to coincide with the axis of rotation of the positioner and point upward; the X-axis is located in the clamping plane when the positioner is in the zero position and points to a preset reference direction (such as pointing towards the operator or along the guide rail); the Y-axis is determined by the right-hand rule.

[0022] It should be noted that during the welding process, all movements of the positioner (tilting and rotation) are performed around the axes of the welding positioner coordinate system. The weld data is ultimately converted to this coordinate system, avoiding complex conversions and errors caused by inconsistencies in coordinate systems.

[0023] The zero position of a welding positioner refers to a preset, standard initial posture of the welding positioner. In the initial posture, the angles of each motion axis of the positioner are defined as zero degrees. For example, if the tilting axis (around the X-axis) is at zero position, it means that the clamping plane of the positioner is in a horizontal state. At this time, the steel support frame is horizontally clamped, and the horizontal weld on the steel support frame can be ideally flat welded without any tilting adjustment.

[0024] To establish a precise and unified spatial reference system and determine the geometric relationship between the welding positioner and the various components in the steel support structure, preparatory work must be completed before obtaining the weld geometric parameters. The first step is to obtain the start and end coordinates of the weld in the CAD model coordinate system: it is known that in modern manufacturing, steel supports all have precise CAD models, and the weld, as a feature in the CAD model, has known start and end coordinates. The second step is to perform coordinate system one: by measuring the reference points on the steel support structure, a mathematical transformation is established to convert all weld coordinates in the CAD model to the welding positioner coordinate system. The third step is to automatically calculate the parameters: based on the converted start and end coordinates of the weld, the computer program automatically calculates the spatial inclination angle and horizontal azimuth angle of each weld.

[0025] In the 3D CAD design model of the steel bracket, at least three non-collinear reference points are predefined on the steel bracket components. The specific setting of the reference points is determined according to the actual situation. This embodiment does not make specific limitations. For example, the reference point can be: the center of a process hole that is specifically machined for positioning, the intersection of two perpendicularly intersecting machined surfaces, or the center of a bolt hole.

[0026] For example, the steel support frame components are actually clamped onto a welding positioner. Using high-precision measuring equipment (such as a laser tracker), the three-dimensional spatial coordinates of the aforementioned reference points are accurately measured in the welding positioner coordinate system. Then, based on the coordinates of the reference points in the CAD model coordinate system and the measured coordinates of the same set of reference points in the welding positioner coordinate system, a coordinate transformation matrix can be obtained through a coordinate transformation algorithm. Subsequently, using the coordinate transformation matrix, the start and end coordinates of all welds extracted from the CAD model are uniformly transformed to the welding positioner coordinate system in batches. The specific coordinate transformation method is a well-known technique to those skilled in the art, and will not be elaborated in this embodiment.

[0027] Spatial tilt angle refers to the angle between the central axis of a single weld and the horizontal plane (i.e., the plane of the positioner coordinate system Z=0). For example, a weld is defined by the three-dimensional coordinates of its starting point and ending point, which constitutes a spatial vector. Calculating the angle between this vector and the horizontal plane (i.e., the plane of Z=0) yields the spatial tilt angle.

[0028] The horizontal azimuth angle refers to the angle between the projection of the weld center axis onto the horizontal plane and a preset reference direction (such as the X-axis of the positioner coordinate system). For example, the spatial vector representing the weld is "flattened" onto the horizontal plane to obtain a projection vector. Then, the angle between this projection vector and the preset reference axis (such as the X-axis of the welding positioner coordinate system, which we define as the 0° direction) is calculated. This angle is the horizontal azimuth angle.

[0029] It's important to understand that while the physical nature of a weld seam tilted in space is unique, mathematically, its angle with the horizontal plane can be measured as either an acute angle (e.g., 30°) or an obtuse angle (e.g., 150°). However, for a welding positioner, tilting the steel support frame component 30° from direction A and tilting it 150° from the opposite direction may ultimately result in the weld seam achieving the same welding posture. Therefore, the angle can be uniformly mapped to the 0-180° range to eliminate the ambiguity caused by the measurement direction.

[0030] S102: Using the spatial tilt angle of all welds as the clustering object and the preset welding posture tolerance as the constraint, perform constrained clustering analysis to group welds with similar spatial tilt angles into the same welding layer to divide multiple welding layers; determine the target tilt angle of each welding layer based on the distribution characteristics of the spatial tilt angle of welds within each welding layer.

[0031] It is important to understand that, due to the different positions and angles of multiple weld seams during welding, it is usually impossible to weld all weld seams perfectly under the same clamping position. Multiple adjustments to the clamping position and angle are often required to complete the overall welding path. However, it is understood that weld seams with similar spatial positions can be welded under the same clamping angle when their spatial positions are all within the preset welding posture tolerance of submerged arc welding. That is, welding can be performed after adjusting the clamping angle once. Therefore, weld seams with similar spatial positions can be regarded as the same layer of weld seams, and for the same layer of weld seams, only the clamping angle needs to be adjusted once.

[0032] The specific value of the preset welding posture tolerance is determined according to the rigid requirements of the welding process. The preset welding posture tolerance means that the deviation between the actual welding angle of a weld and the ideal horizontal angle cannot exceed the preset welding posture tolerance. For example, for general building steel structures, the preset welding posture tolerance value range is usually 8° to 12°.

[0033] The process of constrained cluster analysis is as follows: Figure 2 As shown, it includes: S102-1: Initialize multiple cluster centers and obtain the center tilt angle value of the cluster centers. The difference between the center tilt angle values ​​of any two initial cluster centers is greater than the preset welding posture tolerance.

[0034] For example, if the total number of welds is N, we can use common empirical rules to estimate the initial cluster size K = N ÷ 5 (meaning that on average, every 5 welds can be clustered into one layer), and set the minimum allowable interval, which is the preset welding posture tolerance (e.g., 10°). Then, assuming that the spatial tilt angle of all welds is distributed between 5° and 85°, we will try to randomly select K (e.g., 5) points within this range [5, 85] as the initial center tilt angle value. For each selected point, we will check whether its angle difference with all previously selected points is greater than the preset welding posture tolerance (e.g., 10°). If not, we will discard it and reselect until we find a point that meets the conditions.

[0035] It should be noted that the center inclination angle value, during the initialization phase, represents a hypothetical expected angle for the weld layer. Then, welds with similar angles are collected and organized around this initial center inclination angle value. In subsequent iterations, the center inclination angle value is continuously updated based on the actual spatial inclination angle of the weld assigned to it.

[0036] S102-2: For each weld, calculate the absolute difference between the spatial inclination angle of the weld and the inclination angle of each center, and use it as the interval angle; assign the weld to the cluster center corresponding to the smallest interval angle to form a cluster.

[0037] It should be noted that after all welds have been assigned, a temporary set of welds is formed around each cluster center, and this temporary set of welds is called a cluster.

[0038] Interval angle is used to measure the angular similarity between a weld seam and a potential weld layer (represented by the cluster center).

[0039] S102-3: Calculate the difference between the maximum and minimum spatial tilt angles within a cluster, and use this as the range of the cluster; check whether the range of each cluster is greater than the preset welding posture tolerance; if there is a target cluster with a range greater than the preset welding posture tolerance, divide the target cluster into two sub-clusters with the median of all spatial tilt angles within the target cluster as the center.

[0040] It should be noted that if the range of a certain cluster (called the target cluster) exceeds the preset welding posture tolerance, it means that the cluster is unqualified. The angle difference of the internal welds is too large, which will cause the welding adjustment to fail to meet all the process requirements. In this case, the target cluster can be split: find the median of the spatial inclination angle of all welds in the target cluster, and divide the original cluster into two parts with this median as the boundary to form two new, more compact sub-clusters.

[0041] In order to divide the original cluster evenly into two parts and avoid the problem of uneven splitting caused by extreme values ​​when using the average value, the present invention adopts median splitting. The method of determining the median will not be described in this embodiment.

[0042] S102-4: Iteratively execute the allocation and inspection steps until the range of all clusters is no greater than the preset welding posture tolerance, and define each cluster as a welding layer.

[0043] It is important to understand that each merged cluster means that one time-consuming tilting angle adjustment of the welding positioner can be reduced in production, thereby directly improving the overall welding efficiency. Therefore, the total number of welding layers can be reduced without violating the welding process requirements, which is to say, merging clusters.

[0044] In this embodiment, during the iteration process, if the weld space tilt angle ranges of two clusters overlap, and the range of the new clusters formed after merging is still not greater than the preset welding posture tolerance, then the corresponding two clusters will be merged.

[0045] The angular range of a cluster refers to the angular interval formed by the minimum and maximum spatial tilt angles among all welds within the cluster, i.e., [minimum spatial tilt angle, maximum spatial tilt angle].

[0046] It should be noted that if the range of the new cluster formed after merging is greater than the preset welding posture tolerance, the merging operation for the two clusters will be terminated immediately, and they will remain as two independent welding layers.

[0047] In this embodiment, the arithmetic mean of the spatial inclination angles of all weld seams within the weld layer is calculated as the target inclination angle of the weld layer.

[0048] The target tilt angle refers to the unique target angle of the welding positioner tilting axis set for a certain welding layer, reflecting the concentration trend of the spatial tilt angles of all welds within the welding layer.

[0049] It should be noted that the interference of extreme abnormal spatial tilt angles within the weld layer on the average value also needs to be considered. Therefore, before calculating the arithmetic mean of all weld spatial tilt angles within the weld layer as the target tilt angle, extreme abnormal spatial tilt angles need to be eliminated. For example, the interquartile range method can be used to identify outliers in the weld spatial tilt angles within the weld layer and eliminate the identified outliers. Finally, the arithmetic mean of all remaining normal spatial tilt angles is calculated as the target tilt angle of the weld layer. The interquartile range method is a common technique and will not be elaborated upon in this embodiment.

[0050] For example, assuming the welding positioner is initially in the zero position, that is, the clamping plane is horizontal, and the steel bracket is horizontally clamped on the platform, it receives the instruction from the upper planning algorithm: "Please adjust the overturning axis to 25°", where 25° is the target tilt angle of a certain welding layer. Then, the overturning axis of the welding positioner is controlled to rotate 25°.

[0051] S103: For each welding layer, when the positioner is adjusted to the target tilt angle of the welding layer, the spatial distance cost of the welding torch moving between any two welds in the welding positioner coordinate system is determined; when the positioner is adjusted to the corresponding weld azimuth angle, the positioner rotation cost is determined based on the difference in the azimuth angles of any two welds within the welding layer; and the overall switching cost is determined based on the spatial distance cost and the positioner rotation cost.

[0052] It is important to understand that traditional path optimization usually only considers the spatial distance of the welding torch moving from one point to another and equates it to "time cost" or "energy cost". However, in an automated welding system equipped with a welding positioner, the movement of the welding torch and the rotation of the positioner are two independent motion mechanisms with completely different speeds, accelerations and inertia. Typically, the rotational speed of a positioner is much slower than the linear movement speed of a welding torch. For example, when switching from weld A to weld B, the welding torch only needs to move a short distance (e.g., 0.5 meters), but the positioner needs to rotate a large angle (e.g., 170°). The time spent on this 170° rotation may be much greater than the time it takes for the welding torch to move 0.5 meters. If only the distance is optimized, the algorithm might consider this a "low-cost" switch, but in reality, it is very "time-consuming." Therefore, after classifying welds with similar spatial positions into the same layer, the algorithm can also integrate the path length the welding torch needs to travel in three-dimensional space from the end point of one weld to the beginning point of another, as well as the absolute value of the angle the positioner needs to rotate from the azimuth angle of one weld to the azimuth angle of another, to comprehensively consider the required cost.

[0053] In this embodiment, based on the kinematic model of the welding positioner, the target welding coordinates of the starting and ending points of each weld in the welding positioner coordinate system are calculated when the positioner is adjusted to the target tilt angle of the welding layer and the corresponding weld azimuth angle. The three-dimensional Euclidean distance between the target welding coordinates of the ending point of any weld and the target welding coordinates of the starting point of another weld is calculated. Based on the three-dimensional Euclidean distance, a path compensation amount is added to simulate the safe lifting and lowering operation of the welding torch, and the compensation result is used as the spatial distance cost.

[0054] For example, given the original starting point coordinates of a weld as (x, y, z), by simulating the tilt angle r of the positioner (adjusting the tilt angle along the x-axis of the positioner's coordinate system), the intermediate coordinates of the weld's starting point can be calculated. The intermediate coordinates of the starting point can be expressed as... Next, the positioner is simulated to rotate by an angle w (when adjusting the rotation angle, it is adjusted along the z-axis of the positioner coordinate system) to calculate the target welding coordinates of the weld start point. The target welding coordinates of the start point can be expressed as follows: It should be noted that the calculation principle of the target welding coordinates of the termination point is similar to that of the target welding coordinates of the starting point, only the input parameters are different, which will not be repeated in this embodiment.

[0055] The kinematic model of the welding positioner is a mathematical description that precisely defines the functional relationship between the positioner's motion parameters (such as overturning angle and rotation angle) and the spatial position of any point on the steel support frame it holds. The core of the kinematic model of the welding positioner is the coordinate transformation matrix determined in S101.

[0056] The three-dimensional Euclidean distance represents the theoretical shortest possible path for the welding torch to move from one weld termination point to the next weld start point. The specific calculation method of the Euclidean distance is a well-known technique to those skilled in the art, and will not be described in detail in this embodiment.

[0057] It's important to understand that the three-dimensional Euclidean distance is an "idealized" straight line. In actual welding, the welding torch moves directly from the end of one weld to the beginning of another, which could very likely result in a collision with the steel support structure itself. To ensure safe, collision-free movement, a path compensation factor can be added.

[0058] In this embodiment, the path compensation is twice the preset safety lifting height. The preset safety lifting height is used to simulate the vertical movement path length of the welding torch as it rises before moving and falls after moving.

[0059] It should be noted that the preset safety lifting height is a key engineering parameter used to ensure the safety and reliability of the welding process. The specific value of the preset safety lifting height is determined according to the type of welding torch, the complexity of the workpiece structure, and common safety specifications. For example, the safety lifting height needs to be set according to the maximum protrusion height of the steel support component, which is usually 1.2-1.5 times the maximum protrusion height of the component. For example, when the maximum protrusion height of the component is 40 mm, the safety lifting height can be set to 50 mm.

[0060] During the safe movement of the welding torch between the two weld seams, the first step involves lifting the torch vertically upwards by a safe lifting height (h) from the end point of the previous weld seam after completing the welding of the previous weld seam. The second step involves moving the welding torch from its upper position to directly above the starting point of the next weld seam on a horizontal plane at a height h above the surface of the steel support component. The third step involves moving the welding torch vertically downwards by a safe lifting height h from its hovering position, precisely descending to the starting point of the next weld seam. Therefore, it can be understood that the welding torch undergoes two additional vertical movements throughout the entire process, and the path compensation is set to twice the preset safe lifting height.

[0061] In this embodiment, the absolute difference between the azimuth angles of any two welds within the weld layer is calculated as the positioner's rotation cost.

[0062] The positioner rotation cost quantifies the energy consumption cost required to drive the positioner's rotation shaft when switching from welding one weld to welding another.

[0063] Since the rotating shaft of the welding positioner is a continuously rotating shaft, which is physically a circular structure (360° is one revolution), this means that there are always two possible paths to move from azimuth point A to azimuth point B: Path 1 (forward rotation): rotating a larger angle in the direction of increasing angle, and Path 2 (reverse rotation): rotating a smaller angle in the direction of decreasing angle. Therefore, in order to select the path with the smaller rotation angle to save time and reduce energy consumption, if the absolute difference between the azimuth angles of two welds is greater than 180°, then the actual rotation cost of the positioner should be 360° - this absolute difference.

[0064] In this embodiment, the spatial distance cost and the positioner slewing cost are normalized to obtain normalized spatial distance cost values ​​and normalized slewing cost values. The normalized spatial distance cost values ​​and normalized slewing cost values ​​are then weighted and summed, and the sum is used as the overall switching cost.

[0065] The process of obtaining the overall switching cost can be represented by the following formula: Overall switching cost = α × normalized spatial distance cost + β × normalized turning cost, where α represents the weighting coefficient of the normalized spatial distance cost; β represents the weighting coefficient of the normalized turning cost, and α + β equals 1.

[0066] It should be noted that the specific values ​​of α and β are not fixed, but can be flexibly configured according to the actual equipment characteristics and efficiency targets of the production line. For example, assuming that the rotation mechanism of the positioner is slow and has low acceleration, the time required for the positioner to rotate is much longer than the linear movement time of the welding torch. Therefore, a higher weight can be assigned to the normalized rotation cost value to guide the optimization algorithm to prioritize paths with smaller rotation angles of the positioner, even if the welding torch needs to move a greater distance. This helps to shorten the overall production cycle. In this case, β is greater than α, so β ​​can be 0.7 and α can be 0.3.

[0067] S104: Construct a cost matrix based on the comprehensive switching cost between all welds within the weld layer, and solve for the weld welding sequence with the lowest total cost using a path optimization algorithm.

[0068] Path optimization algorithms can include ant colony optimization, genetic optimization, etc. This embodiment uses ant colony optimization as an example for explanation.

[0069] In this embodiment, the parameters of the path optimization algorithm are initialized, including the pheromone matrix, the heuristic information matrix, the number of ants, and the number of iterations. A cost matrix is ​​constructed based on the comprehensive switching cost between all weld seams within the welding layer, where each element represents the comprehensive switching cost from one weld seam to another. Multiple ants are simulated to sequentially select the next weld seam to be welded based on the state transition probability, until all weld seams are traversed, forming multiple complete welding paths. The state transition probability is determined by both the pheromone concentration and the heuristic information, which is set as the reciprocal of the comprehensive switching cost. The pheromone concentration between each weld seam pair is updated based on the path quality constructed by all ants in this iteration. The steps of constructing ant paths and updating pheromones are repeated until the maximum number of iterations is reached or the convergence condition is met. From all the paths generated in the iterations, the path with the minimum total switching cost is selected as the final weld seam welding sequence.

[0070] For example, if a certain welding layer has 3 welds, namely weld A, weld B and weld C, then the cost matrix can be shown in Table 1 below (numerical virtual). It should be noted that there is no cost to switch a weld to itself. For example, the overall cost of switching from weld A to weld C is zero.

[0071] Table 1 It should be noted that the number of iterations determines the depth and duration of the algorithm's search solution. The specific value of the number of iterations is determined according to the actual situation, and this embodiment does not impose a specific limitation. For example, when using the ant colony algorithm, in order to ensure that "pheromones" accumulate on high-quality paths, the lower limit of the number of iterations can be set to 100 times. In order to control the time occupied by the planning itself and ensure that it can be completed within a reasonable time, the lower limit of the number of iterations can be set to 500 times.

[0072] It should be noted that the specific processes of constructing the ant path and updating the pheromone are well-known techniques to those skilled in the art. This invention has provided a brief overview, and the specific processes will not be described in detail in this embodiment.

[0073] It should be noted that, in order to ensure the quality of the solution and avoid unnecessary waste of computational resources when solving for the optimal welding sequence, path optimization algorithms (such as ant colony optimization) use the maximum number of iterations or the satisfaction of a convergence condition as the termination criterion. For example, the maximum number of iterations can be set to 300. After the algorithm has been executed 300 times, it will forcibly stop the calculation and output the currently found optimal solution. The convergence condition can be set to no optimal solution found after 50 consecutive iterations. For example, the difference between the total cost of the optimal path in each iteration and the historical optimal total cost is less than the convergence threshold (the convergence threshold can be set to 0.01 based on the actual welding scenario). If the number of consecutive times the value is less than the convergence threshold reaches 50, the calculation will be terminated.

[0074] S105: Adjust the positioner to the corresponding target tilt angle in sequence according to the target tilt angle of each welding layer; within each welding layer, adjust the positioner azimuth angle and welding torch position in sequence according to the welding sequence of the weld to complete the welding.

[0075] In this embodiment, the welding positioner is adjusted to the corresponding target tilt angle in order of increasing or decreasing target tilt angle values ​​for each welding layer.

[0076] For example, initially, the positioner is in a safe position, such as zero position (0° tilt angle), and the first welding layer is executed: the control system drives the tilting axis of the positioner to move to the target tilt angle of the first layer (e.g., 12°). After the movement is completed, the positioner is locked at that angle. The second welding layer is executed: after all the welds of the first welding layer are completed, the control system drives the positioner to move from its current angle (12°) to the target tilt angle of the second layer (e.g., 35°). This process is repeated until the tilt angle adjustment of all welding layers is completed.

[0077] It should be noted that, under the target tilt angle of each weld layer, the welding of all welds in that weld layer is completed one by one according to the optimal weld welding sequence obtained by the path optimization algorithm.

[0078] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0079] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for optimizing the multi-segment automatic submerged arc welding path of a steel bracket, characterized in that, The method includes: Obtain the geometric parameters of all weld seams to be welded on the steel support frame components in the coordinate system of the welding positioner. The geometric parameters include at least the spatial tilt angle and horizontal azimuth angle of each weld seam. Using the spatial tilt angles of all welds as clustering objects and a preset welding posture tolerance as a constraint, a constrained clustering analysis is performed to group welds with similar spatial tilt angles into the same welding layer, thus dividing the welds into multiple welding layers. The constrained clustering analysis includes: initializing multiple cluster centers and obtaining the center tilt angle values ​​of the cluster centers, wherein the difference between the center tilt angle values ​​of any two initial cluster centers is greater than the preset welding posture tolerance; for each weld, calculating the absolute difference between the spatial tilt angle of the weld and each center tilt angle value, using this as the interval angle; and assigning the welds... Clusters are formed by finding the cluster center corresponding to the minimum interval angle. The difference between the maximum and minimum spatial tilt angles within a cluster is calculated as the range of the cluster. The range of each cluster is checked to see if it is greater than the preset welding posture tolerance. If there is a target cluster with a range greater than the preset welding posture tolerance, the target cluster is divided into two sub-clusters with the median of all spatial tilt angles within the target cluster as the center. The allocation and checking steps are iteratively executed until the range of all clusters is not greater than the preset welding posture tolerance, and each cluster is defined as a welding layer. Based on the distribution characteristics of the weld space inclination angle within each weld layer, determine the target inclination angle of each weld layer; For each welding layer, when the positioner is adjusted to the target tilt angle of the welding layer, the spatial distance cost of the welding torch moving between any two welds in the welding positioner coordinate system is determined; when the positioner is adjusted to the corresponding weld azimuth angle, the positioner rotation cost is determined based on the difference between the azimuth angles of any two welds in the welding layer; and the overall switching cost is determined based on the spatial distance cost and the positioner rotation cost. The parameters of the path optimization algorithm are initialized, including the pheromone matrix, heuristic information matrix, number of ants, and number of iterations. A cost matrix is ​​constructed based on the comprehensive switching cost between all weld seams within the welding layer, where each element represents the comprehensive switching cost from one weld seam to another. Multiple ants are simulated to sequentially select the next weld seam to be welded based on state transition probabilities until all weld seams have been traversed, forming multiple complete welding paths. The state transition probabilities are determined by both pheromone concentration and heuristic information, with the heuristic information set as the reciprocal of the comprehensive switching cost. The pheromone concentration between each weld seam pair is updated based on the path quality constructed by all ants in this iteration. The steps of constructing ant paths and updating pheromones are repeated until the maximum number of iterations is reached or the convergence condition is met. From all the paths generated in the iterations, the path with the minimum total switching cost is selected as the final weld seam welding sequence. Adjust the positioner to the corresponding target tilt angle in sequence according to the target tilt angle of each welding layer; within each welding layer, adjust the positioner azimuth angle and welding torch position in sequence according to the welding sequence of the weld to complete the welding.

2. The method for optimizing the multi-segment automatic submerged arc welding path of a steel bracket according to claim 1, characterized in that, During the iteration process, if the weld space tilt angle ranges of two clusters overlap, and the range of the new cluster formed after merging is still not greater than the preset welding posture tolerance, then the two corresponding clusters will be merged.

3. The method for optimizing the multi-segment automatic submerged arc welding path of a steel bracket according to claim 1, characterized in that, The process of determining the target tilt angle includes: Calculate the arithmetic mean of the spatial inclination angles of all weld seams within the weld layer, and use it as the target inclination angle of the weld layer.

4. The method for optimizing the multi-segment automatic submerged arc welding path of a steel bracket according to claim 1, characterized in that, The process of determining the spatial distance cost includes: Based on the kinematic model of the welding positioner, the target welding coordinates of the starting and ending points of each weld in the welding positioner coordinate system are calculated when the positioner is adjusted to the target tilt angle of the welding layer and the corresponding weld azimuth angle. Calculate the three-dimensional Euclidean distance between the target welding coordinates of the termination point of any weld and the target welding coordinates of the start point of another weld; Based on the three-dimensional Euclidean distance, a path compensation amount is added to simulate the safe lifting and lowering operation of the welding torch, and the compensation result is used as the spatial distance cost.

5. The method for optimizing the multi-segment automatic submerged arc welding path of a steel bracket according to claim 4, characterized in that, The path compensation amount is twice the preset safety lifting height, which is used to simulate the vertical movement path length of the welding torch as it rises before moving and falls after moving.

6. The method for optimizing the multi-segment automatic submerged arc welding path of a steel bracket according to claim 1, characterized in that, The process of determining the rotation cost of the positioner includes: Calculate the absolute difference between the azimuth angles of any two welds within the weld layer, and use this as the positioner's rotation cost.

7. The method for optimizing the multi-segment automatic submerged arc welding path of a steel bracket according to claim 1, characterized in that, The process of determining the overall switching cost includes: The spatial distance cost and the positioner rotation cost are normalized respectively to obtain normalized spatial distance cost values ​​and normalized rotation cost values; The normalized spatial distance cost value and the normalized turning cost value are weighted and summed, and the sum is used as the overall switching cost.

8. The method for optimizing the multi-segment automatic submerged arc welding path of a steel bracket according to claim 1, characterized in that, The step of sequentially adjusting the positioner to the corresponding target tilt angle according to the target tilt angle of each weld layer includes: Adjust the welding positioner to the corresponding target tilt angle in order of increasing or decreasing target tilt angle values ​​for each weld layer.

Images