Anti-fatigue welding process for long-distance pipeline supporting steel structure

Abstract

The invention relates to the technical field of welding, and particularly discloses an anti-fatigue welding process for a long-distance pipeline supporting steel structure. The process comprises the steps that a font groove structure is determined, and pre-welding polishing pretreatment and layered welding are conducted; fatigue samples of weld toe areas with different arc radiuses are collected, crack feature coordinates and stress concentration coefficients are extracted, failure modes are divided based on clustering, the arc radiuses are verified as main control factors of fatigue failure through multiple linear regression, and a core control range is determined by combining process capability indexes; carrying out fatigue verification on the basis of core parameter extension radius gradient, carrying out process compatibility matching in combination with a controllable interval and a thermal deformation interval of equipment, and outputting a radius candidate value through a response surface method; on-site construction constraints are collected, laboratory path space compatibility is analyzed, a formed track is optimized in a segmented mode, and constraint compensation factors are embedded. The anti-fatigue performance of the weld toe area of the supporting steel structure can be improved, and quality guarantee is provided for welding of the long-distance pipeline supporting steel structure.

Description

Technical Field

[0001] This invention relates to the field of welding technology, specifically to a fatigue-resistant welding process for supporting steel structures of long-distance pipelines. Background Technology

[0002] As a core infrastructure for energy transmission, long-distance pipelines require their supporting steel structures to withstand alternating loads caused by the pipeline's own weight, medium transmission pressure, and temperature fluctuations over long periods of time. Welded areas (especially the weld toe area where the pipeline connects to the steel structure) are prone to stress concentration due to abrupt changes in geometry, making them high-risk areas for fatigue failure.

[0003] Once fatigue cracks occur in the weld toe zone, it will not only shorten the service life of the supporting steel structure, but may also cause safety accidents such as pipeline displacement and leakage. Therefore, the fatigue resistance of the weld toe zone is a key indicator to ensure the stable operation of long-distance pipelines. The fatigue failure judgment of the existing technology does not take into account the changes in the load-bearing capacity of the specimen for comprehensive evaluation, which is prone to misjudgment when the crack depth does not exceed the standard but the load-bearing capacity has decreased significantly, affecting the accuracy of failure sample identification.

[0004] Existing technologies do not take into account the spatial constraints of on-site construction, nor do they incorporate the parameter fluctuations caused by the processing precision of on-site equipment and the thermal deformation of the base material. As a result, when the optimal parameters determined in the laboratory are applied on-site, the accuracy of the arc radius of the weld toe area is difficult to meet the standards.

[0005] Therefore, the present invention provides a fatigue-resistant welding process for the supporting steel structure of long-distance pipelines. Summary of the Invention

[0006] The purpose of this invention is to provide a fatigue-resistant welding process for the supporting steel structure of long-distance pipelines to solve the aforementioned background problems.

[0007] The objective of this invention can be achieved through the following technical solutions:

[0008] Fatigue-resistant welding process for steel structure support of long-distance pipelines includes the following steps:

[0009] Determine the bevel structure of the supporting steel structure and perform pre-welding pretreatment;

[0010] The pre-treated parts are assembled, positioned, and welded in layers.

[0011] Post-weld fatigue strengthening and monitoring optimization were carried out on the weld-strengthened area after welding treatment.

[0012] Monitoring optimization includes:

[0013] Fatigue failure samples of the weld toe zone with different arc radii were collected, and crack feature coordinates and stress concentration factors of the fatigue failure samples were extracted.

[0014] Fatigue failure samples are classified by crack feature coordinates. Based on the classification results and stress concentration coefficient, the main control effect of the arc radius is verified. If the main control effect exists, crack-free samples are obtained and the optimal mode is screened to obtain the core control range.

[0015] Fatigue verification of the arc radius is performed based on the core control range to determine the radius fatigue range. Welding control parameters are obtained and combined with the radius fatigue range for welding process compatibility matching. It is determined whether there is a process compatibility intersection between the radius fatigue range and the welding control parameters. If there is, candidate radius values ​​are output.

[0016] The construction constraint parameters are obtained and spatial compatibility analysis is performed in combination with the candidate radius values ​​to obtain the constraint compatibility coefficient and determine whether the space is compatible. If compatible, a trajectory segmentation optimization strategy for the constraint parameters is constructed to optimize the trajectory.

[0017] As a further aspect of the present invention, the method for obtaining the weld toe feature parameters is as follows:

[0018] Establish a coordinate system for fatigue failure samples, extract the coordinate set corresponding to the fatigue crack edge, and calculate the centroid coordinates of the fatigue crack edge as crack feature coordinates;

[0019] The maximum strain value of each strain gauge in the strain gauge array is collected to obtain the elastic modulus of the base material. The elastic modulus is then multiplied with the maximum strain value to obtain the actual maximum stress.

[0020] Obtain the applied load and the cross-sectional area of ​​the standard test specimen corresponding to the maximum strain value, and then perform a ratio calculation on the load and the cross-sectional area to obtain the nominal maximum stress.

[0021] The ratio of the actual maximum stress to the nominal maximum stress is calculated and used as the stress concentration factor.

[0022] As a further aspect of the present invention, the method for verifying the controlling effect of the arc radius is as follows:

[0023] Based on the classification results of fatigue failure samples, different failure modes are obtained and assigned values ​​to different failure modes, and the assigned failure modes are used as dependent variables.

[0024] Obtain the bevel gap and interlayer temperature, and use the arc radius, bevel gap, and interlayer temperature as independent variables;

[0025] A multiple linear regression equation is constructed based on independent and dependent variables, and the regression coefficient and significance index corresponding to each independent variable are calculated.

[0026] If only the regression coefficient of the arc radius shows significance, then the arc radius is verified as the main controlling factor of fatigue failure.

[0027] As a further aspect of the present invention, the process of obtaining the classification result is as follows:

[0028] The crack propagation direction angle is obtained, and the crack feature coordinates are used as a classification dataset. A failure classification model is constructed based on a clustering algorithm. The classification dataset is input into the failure classification model, and the clustering results are output. Based on the clustering results, the fatigue mode is classified, and the classification results of fatigue failure samples are obtained.

[0029] As a further aspect of the present invention, the process of performing the optimal mode screening is as follows:

[0030] Obtain the radius of the arc and the stress concentration factor of all crack-free samples to construct a sample dataset;

[0031] Based on the stress concentration factor, the samples without cracks in the sample dataset are filtered to obtain a subset of the sample data;

[0032] Obtain the upper and lower limits of the radius of the arc, and calculate the truncated mean and standard deviation of the sample data subset;

[0033] Construct a process capability equation by inputting the upper and lower limits of the specifications, the truncated mean and standard deviation into the process capability equation to obtain the process capability index;

[0034] The core control range of the arc radius is determined by combining the truncated mean value with the process capability index.

[0035] As a further aspect of the present invention: the process for determining whether there is a process compatibility overlap is as follows:

[0036] The controllable range and thermal deformation range of the equipment are obtained and superimposed to obtain the feasible range of the process.

[0037] Calculate the characteristic confidence interval of the core control range, and calculate the overlap between the characteristic confidence interval and the process feasible interval to obtain the overlap degree and the overlap interval;

[0038] Monte Carlo simulation was used to verify the fatigue trapping rate by randomly sampling the radius within the overlapping interval Z times and calculating the probability that the sampled value falls within the fatigue range of the radius.

[0039] A comparison criterion is constructed based on overlap and fatigue trapping rate. If the overlap and fatigue trapping rate meet the comparison criterion, it is determined that there is a process compatibility intersection between the radius fatigue range and the welding control parameters.

[0040] As a further aspect of the present invention: the process of obtaining the radius fatigue range is as follows:

[0041] Based on the core control range, the radius of the arc is extended to prepare parallel test pieces, and accelerated testing is performed on the parallel test pieces to obtain the fatigue life of each parallel test piece.

[0042] For each group of parallel test specimens, a distribution fit is performed at different fatigue lives, and the characteristic life and shape parameters of the parallel test specimens with the distribution fit at the preset reliability are extracted.

[0043] Obtain the radius of the arc corresponding to the parallel test piece that meets the preset reliability characteristic lifetime, and calculate the characteristic confidence interval using the t-distribution;

[0044] Obtain and calculate the mean of the shape parameters corresponding to all parallel test pieces that meet the preset reliability characteristic life, and calculate the reciprocal of the square of the mean of the shape parameters as the fatigue dispersion coefficient;

[0045] The characteristic confidence interval is adjusted by superimposing the fatigue dispersion coefficient to obtain the radius fatigue range.

[0046] As a further aspect of the present invention: the method for outputting the candidate radius value is as follows:

[0047] Test points were designed based on the radius of the arc and fatigue life.

[0048] Obtain experimental data and construct a quadratic polynomial model to fit the response surface model;

[0049] The solution is obtained based on the response surface model, and the solution results are verified to obtain candidate values ​​for the radius.

[0050] As a further aspect of the present invention: the process of determining whether the space is compatible is as follows:

[0051] Obtain the laboratory path corresponding to the candidate radius value, as well as the three-dimensional spatial model of the welding site. Map the laboratory path to the three-dimensional spatial model and combine it with the construction constraint matrix to perform constraint adaptability detection.

[0052] Extract the verification results item by item. If the boundary data at the end of the path does not match the construction constraint matrix, extract the nearest constraint value in the construction constraint matrix for each mismatched boundary data item.

[0053] Calculate the Euclidean distance between each boundary data point and the nearest constraint value, and use it as the constraint compatibility coefficient;

[0054] The comparison is performed based on the constraint compatibility coefficient, and a trajectory segmentation optimization strategy based on the comparison results is constructed using the constraint parameters.

[0055] As a further aspect of the present invention: the trajectory segmentation optimization strategy is constructed as follows:

[0056] Obtain the robotic arm's radius of motion constraints and the pipe curvature constraints;

[0057] To address the constraints on the robotic arm's radius of motion, the path is split into multiple path segments.

[0058] Based on the pipe curvature constraint, an adaptive curvature equation is matched for each path segment to adjust the trajectory;

[0059] Establish a constraint compensation factor to perform radius fine-tuning compensation at the critical point of the robotic arm's operating radius.

[0060] The beneficial effects of this invention are:

[0061] (1) Determining the groove structure and pre-welding pretreatment is beneficial to improving the consistency and standardization of the groove structure, alleviating the risk of welding defects caused by fluctuations in parameters such as groove angle, gap, and blunt edge. At the same time, by grinding before welding and selecting matching welding rods, the interference of impurities on welding quality can be reduced. Assembly positioning and layered welding are beneficial to achieving orderly connection and smooth transition between weld beads, improving the overall forming quality of the weld. By matching specific currents and interpass temperatures in layers, the forming requirements of different weld beads can be adapted. Post-weld fatigue strengthening and monitoring optimization are beneficial to improving the microstructure of the weld toe area and heat-affected zone, enhancing the fatigue resistance of key areas, alleviating the phenomenon of residual stress concentration after welding, and providing a guarantee for subsequent fatigue resistance optimization.

[0062] (2) A dataset is constructed based on the crack propagation direction angle and crack feature coordinates. Failure modes are divided by clustering. The radius is associated with the failure mode. The main control role of the arc radius is verified by multiple linear regression, which helps to reduce the blindness of multi-parameter optimization. The optimal screening based on stress concentration coefficient and process capability index helps the core control range to have both good fatigue resistance and process stability. Parallel test pieces are prepared by expanding the radius gradient based on the core control range to accelerate the fitting of characteristic life after fatigue test. The radius fatigue range is determined by combining t distribution and fatigue dispersion coefficient, which helps to alleviate the randomness bias of single test data. The process feasible range of the equipment controllable range and the thermal deformation range is superimposed, which helps to reduce the constraints of equipment accuracy and thermal deformation on parameter implementation. The radius candidate value is output by response surface method to improve the actual welding process capability and alleviate the problem of the disconnect between laboratory optimization parameters and field conditions.

[0063] (3) Collect on-site construction constraint parameters to establish a constraint matrix, identify the adaptation risk between laboratory path and on-site space in advance, and alleviate welding deviation caused by insufficient understanding of working conditions; map the laboratory horizontal arc path corresponding to the radius candidate value to the on-site three-dimensional model to verify compatibility, calculate the constraint compatibility coefficient, and perform segmentation, matching curvature adaptive equation, and embedding radius fine-tuning compensation for the compatible path. The combination strategy of segment optimization, curvature adaptation and compensation is conducive to ensuring stable control of arc radius accuracy during on-site welding, while adapting to spatial constraints such as the range of motion of the robotic arm and the curvature of the pipeline, thereby improving the reliability and stability of on-site welding quality. Attached Figure Description

[0064] The invention will now be further described with reference to the accompanying drawings.

[0065] Figure 1 This is a flowchart of the fatigue-resistant welding process for the supporting steel structure of the long-distance pipeline of the present invention;

[0066] Figure 2 This is a flowchart of a welding control method for the weld toe area in this invention;

[0067] Figure 3 Yes, here is a flowchart of the method for determining fatigue failure samples;

[0068] Figure 4 This is a block diagram of a welding control system for the weld toe area according to the present invention. Detailed Implementation

[0069] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0070] Example 1

[0071] Please see Figure 1 As shown, this invention provides a fatigue-resistant welding process for the supporting steel structure of long-distance pipelines, comprising the following steps:

[0072] S1. Determine the bevel structure of the supporting steel structure and perform pre-welding treatment;

[0073] The method for determining the bevel structure supporting the steel structure is as follows:

[0074] For the connection nodes between the supporting steel structure pipes and steel supports, a single V-shaped bevel is adopted, and the control parameters of the bevel structure are determined.

[0075] Among them, the control parameters of the bevel structure include: measuring the included angle between the bevel surfaces of the two workpieces by placing a square against the bevel surfaces on both sides of the workpiece, and controlling the angle to be within 60°±5°;

[0076] Preferably, the control angle is 65°;

[0077] The straight-line distance between the root of the bevel of the two parts to be welded is measured with a feeler gauge, and the straight-line distance is controlled to be 2-3 mm (preferably 2.25 mm).

[0078] The thickness of the straight section at the root of the bevel of the workpiece to be welded is measured by calipers, that is, the uncut straight edge (i.e., blunt edge) retained at the root of the bevel, and the straight edge is controlled to be within the range of 1 to 1.5 mm (preferably 1.30 mm).

[0079] The pre-welding pretreatment method is as follows: cut and process the completed bevel structure, and grind the two sides of the bevel (including the bevel, blunt edge and the surface of the base material to be welded) within a 30mm range until a metallic luster is revealed; select a low-hydrogen welding rod that matches the base material;

[0080] S2. Assemble, position, and perform layered welding on the pre-treated parts to be welded.

[0081] Preferably, the workpieces to be welded are aligned and fixed by tack welding so that the bevel angle, gap, and blunt edge meet the control parameters of the bevel structure.

[0082] It should be noted that the tack welds are located on both sides of the bevel, with a spacing of 200mm and a length of 15mm;

[0083] After the tack weld is completed and the workpiece is fixed, layered welding is performed, including weld beads, filler weld beads and capping weld beads.

[0084] The root weld uses a 3.2mm welding rod, with a current controlled at 110-120A (preferably 115A), and is welded along the blunt edge and gap to form a "V" shaped bottom channel.

[0085] The filler weld uses 4.0mm welding rods, and the current is controlled at 140-150A (preferably 145A). Each weld layer covers 1 / 3 of the width of the previous layer, and the interpass temperature is controlled at 120-150℃ (preferably 135℃).

[0086] The cover pass uses a 4.0mm electrode, and the current is reduced to 120-130A (preferably 125A). The width of the pass extends 1-2mm beyond the bevel edge to form a smooth transition at the bevel edge.

[0087] S3. Post-weld fatigue strengthening and monitoring optimization are carried out on the weld-strengthened area after welding treatment.

[0088] The weld strengthening areas include the weld toe zone and the heat-affected zone.

[0089] It should be noted that the weld toe zone is the connection point between the bevel edge and the weld, and the heat-affected zone is a 10mm range on each side of the bevel.

[0090] Post-weld fatigue strengthening includes remelting and reshaping, local chilling, and ultrasonic shock.

[0091] Remelting and shaping: The tungsten electrode welding toe area is remelted along the edge of the bevel to form an arc transition with a radius greater than 3mm;

[0092] Localized rapid cooling: After edge remelting, the weld toe area is vertically sprayed with 0.8 MPa high-pressure water to reduce the surface temperature to 150℃.

[0093] Ultrasonic impact: The weld and heat-affected zone are subjected to impact treatment using an impact device at 20 kHz.

[0094] Example 2

[0095] Please see Figure 2 As shown, this invention relates to a fatigue-resistant welding process for the supporting steel structure of long-distance pipelines. It employs a welding control method for the weld toe zone to monitor and optimize the weld-strengthened area, comprising the following steps:

[0096] Step 1: Collect fatigue failure samples of the weld toe zone under different arc radii, and extract the crack feature coordinates and stress concentration factor of the fatigue failure samples as weld toe feature parameters;

[0097] The method for collecting fatigue failure samples of the weld toe area under different arc radii is as follows:

[0098] Preferably, using the connection structure between the pipe and the steel structure as a prototype, different arc radius gradients are set in the weld toe area, and multiple sets of standard test pieces of the prototype are prepared based on the arc radius gradient;

[0099] For example: For the weld toe area, five sets of specimens with different arc radii were prepared by remelting process. The radius values ​​covered the gradient range of the standard value of 3 mm, equal to the standard value, and higher than the standard value, specifically: 2.0 mm, 3.0 mm, 3.5 mm, 4.0 mm, and 5.0 mm. Three parallel specimens were prepared for each set.

[0100] Among them, the standard test pieces are identical except for the arc radius of the weld toe area, which remains the same. Other basic parameters (such as bevel basic parameters), welding process, base material and welding rod are the same.

[0101] A servo fatigue testing machine was used to apply alternating loads to all standard test pieces, and the loading parameters were matched with the stress scenarios of the steel structure supporting the long-distance pipeline.

[0102] It should be noted that axial alternating load is used to simulate the reciprocating stress in the pipeline caused by medium transportation and temperature changes, and the stress ratio (the ratio of minimum stress to maximum stress) simulated by axial alternating load is 0.1, and the loading frequency is 15Hz.

[0103] The fatigue crack depth of the standard test piece is collected using the displacement monitoring system built into the testing machine;

[0104] If the current load-bearing capacity of the standard test piece is lower than the initial load-bearing capacity, calculate the absolute reduction ratio of the current load-bearing capacity to the initial load-bearing capacity to obtain the load-bearing capacity reduction ratio;

[0105] The failure criteria are defined as follows:

[0106] If the fatigue crack depth generated by the standard test piece during axial alternating load simulation is higher than or equal to the preset crack warning value, the current standard test piece is determined to be a fatigue failure sample.

[0107] If the fatigue crack depth is lower than the preset crack warning value, calculate the difference in the closeness between the fatigue crack depth and the crack warning value;

[0108] It should be noted that, based on the relevant standards for the support steel structure of long-distance pipelines, those skilled in the art input measurable parameters such as the mechanical parameters of the base material, the service alternating load, and the geometric parameters of the welded joint, and quantify and calculate according to "max(0.1δ,1mm)" (δ is the measured wall thickness of the pipeline) to obtain the preset crack warning value;

[0109] The proximity difference is dimensionless, and the processed proximity difference is multiplied by the load reduction ratio to obtain the failure judgment value.

[0110] like Figure 3 As shown, if the failure judgment value is higher than or equal to the preset failure judgment threshold, the standard test piece is judged as a fatigue failure sample.

[0111] If the failure determination value is lower than the preset failure determination threshold, the change in the failure determination value will be continuously monitored.

[0112] The method for extracting crack feature coordinates and stress concentration factor from fatigue failure samples as weld toe feature parameters is as follows:

[0113] Establish a coordinate system for fatigue failure samples, extract the coordinate set corresponding to the fatigue crack edge, and calculate the centroid coordinates of the fatigue crack edge as crack feature coordinates;

[0114] The stress concentration factor is determined by an array of strain gauges installed before loading a standard test specimen, with each strain gauge in the array capturing the maximum strain value.

[0115] Obtain the elastic modulus of the base material, multiply the elastic modulus by the maximum strain value to obtain the actual maximum stress;

[0116] Obtain the applied load and the cross-sectional area of ​​the standard test specimen corresponding to the maximum strain value, and then perform a ratio calculation on the load and the cross-sectional area to obtain the nominal maximum stress.

[0117] Calculate the ratio of the actual maximum stress to the nominal maximum stress, and use it as the stress concentration factor;

[0118] The crack characteristic coordinates and stress concentration factor of the fatigue failure sample are used as the weld toe characteristic parameters;

[0119] It is understandable that the purpose of extracting weld toe feature parameters is:

[0120] Objective 1: To provide data support for classifying fatigue failure samples by crack feature coordinates and verifying whether the radius of curvature is the main controlling factor of fatigue failure by combining stress concentration factor matching.

[0121] Objective 2: To provide key parameter basis for subsequent screening of optimal modes for crack-free samples, determination of radius fatigue range, and welding process compatibility matching, and to support the derivation of core control range and candidate radius values.

[0122] Step 2: Classify fatigue failure samples by crack feature coordinates. Based on the classification results and stress concentration coefficient, verify the main control effect of the arc radius. If the main control effect exists, obtain crack-free samples and perform optimal mode screening to obtain the core control range.

[0123] The method for classifying fatigue failure samples using crack feature coordinates is as follows:

[0124] Crack propagation direction angles are extracted and combined with crack feature coordinates as a classification dataset. A failure classification model is constructed based on the K-Means clustering algorithm. The classification dataset is input into the failure classification model and the clustering results are output. The fatigue mode is classified based on the clustering results.

[0125] The crack propagation direction angle is extracted as follows: based on the actual morphology of the crack edge in the fatigue failure sample, the actual propagation direction of the crack is determined, and the angle between this propagation direction and a preset reference direction (such as the pipe axis or the tangent direction of the bevel edge) is calculated; the failure classification model is constructed by using "crack propagation direction angle + crack feature coordinates" to build a classification dataset, where the former is the angle between the crack propagation direction and the reference direction, and the latter is the coordinates of the centroid of the crack edge; K is set to 3 (corresponding to 3 fatigue modes), and 3 representative samples are selected as initial cluster centers. Subsequent iterations: the distance from each sample to the center is calculated for clustering, and the mean of the parameters within the cluster is used to update the center until the center stabilizes. Finally, the 3 clusters are mapped to failure modes A (weld toe root crack), B (heat-affected zone crack), and C (no crack), respectively, completing the failure classification model construction.

[0126] It should be noted that the method for classifying fatigue patterns based on clustering results is as follows:

[0127] Pattern mapping is performed based on clustering results: the cluster whose cluster center is close to the crack feature coordinates of the "weld toe root coordinate area" and whose extension angle is <30° is defined as the failure mode crack feature A;

[0128] Clusters whose cluster centers are located in the crack feature coordinates of the "heat-affected zone coordinate region" and whose crack feature coordinates have an extension angle > 45° are defined as failure mode crack features B;

[0129] Crack-free samples (crack depth feature value is 0, crack feature coordinate) are clustered separately as failure mode crack feature C;

[0130] The classification results of fatigue modes are correlated with the arc radius of the standard test piece to construct a radius-failure mode group, thus realizing the correspondence between different arc radii and failure modes.

[0131] Different failure modes are assigned values, and the assigned failure modes are used as dependent variables to obtain bevel gap and interlayer temperature, while the arc radius, bevel gap, and interlayer temperature are used as independent variables.

[0132] A multiple linear regression equation is constructed based on the independent and dependent variables, and the regression coefficient and significance index (P-value) corresponding to each independent variable are calculated.

[0133] If only the regression coefficient of the arc radius shows statistical significance (i.e., P < 0.05), then the arc radius is verified as the main controlling factor of fatigue failure.

[0134] Those skilled in the art will understand that the assignment method is as follows: three types of fatigue failure modes (A: weld toe root crack, B: heat-affected zone crack, C: no crack) are quantified and assigned values. For example, the no crack mode C is assigned 0, the weld toe root crack mode A is assigned 1, and the heat-affected zone crack mode B is assigned 2 according to the failure state, so that the assigned failure modes are transformed into specific values ​​that can participate in the regression calculation as dependent variables.

[0135] In statistics, a p-value < 0.05 indicates that the result is statistically significant. "Only the p-value for the regression coefficient of the radius of the arc is < 0.05 (significant), while the p-value for the regression coefficients of the bevel gap and interlayer temperature is > 0.05 (not significant)" is used to verify that the radius of the arc is the main controlling factor of fatigue failure.

[0136] The specific form of the multiple linear regression equation is as follows: let the dependent variable be Y (failure mode after assignment), and the independent variables be X1 (radius of the arc), X2 (bevel gap), and X3 (interlayer temperature). The equation is Y = a + b1X1 + b2X2 + b3X3 + ε (where a is a constant term, b1, b2, and b3 are the regression coefficients of their respective variables, and ε is the error term). At the same time, the coefficient of determination of the multiple linear regression equation needs to be higher than 0.8 to ensure that the equation can explain more than 80% of the variation of the dependent variable.

[0137] If the radius of the arc is the main controlling factor for fatigue failure, then the core control range is obtained by acquiring crack-free samples and performing optimal mode screening.

[0138] Obtain the radius of the arc and the stress concentration factor of all crack-free samples to construct a sample dataset;

[0139] Based on the stress concentration factor, the samples without cracks in the sample dataset are filtered to obtain a subset of the sample data;

[0140] Preferably, based on the stress concentration coefficient corresponding to the failure mode crack feature C of the crackless samples, the stress screening coefficient is determined to be lower than 1.85. The crackless samples with stress concentration coefficients lower than the stress screening threshold are screened to construct a sample data subset.

[0141] Through process capability equations: Calculate the process capability index CPK for a subset of sample data;

[0142] USL is the upper limit of the arc radius specification (i.e. the maximum arc radius value allowed in the process or design requirements, which needs to be set in advance based on actual processing capabilities and fatigue performance requirements). The truncated mean of the subset of sample data. 1 represents the standard deviation of the subset of sample data. This is the lower limit of the radius specification for the arc;

[0143] Understandably, the process capability index is used to measure the welding process's ability to stably control the radius of the weld toe area, a key parameter. In other words, it measures whether the process can consistently and stably control the radius within a range that meets fatigue resistance requirements. The calculation combines the upper limit of the radius specification with the truncated mean and standard deviation of the crack-free sample subset. The numerical value reflects process stability; a higher index indicates less susceptibility to fluctuations and a higher probability that the radius falls within the target range. Ultimately, it is used to screen out the core control range of the radius that combines good fatigue resistance and process feasibility, providing a basis for subsequent process optimization.

[0144] Calculate the truncated mean of the sample data subset, and determine the core control range of the arc radius by combining the truncated mean with the process capability index CPK.

[0145] For example, a subset of crack-free sample data (e.g., 3.3mm, 3.5mm, 3.6mm, 3.7mm, 3.8mm, 3.9mm, 4.0mm, and 4.1mm, a total of 8 samples) is extracted, and a 10% truncated mean is used: the maximum value (4.1mm) and the minimum value (3.3mm) of the 10% of data are removed, and the mean of the remaining 6 samples is calculated to be 3.72mm.

[0146] The standard deviation of the sample data subset (3.5mm, 3.6mm, 3.7mm, 3.8mm, 3.9mm, 4.0mm) is calculated to be 0.187mm. The upper limit of the arc radius specification is set to USL = 4.0mm (based on the maximum precision of the processing equipment), and the lower limit is set to LSL = 3.5mm (to reduce stress concentration and rebound caused by excessively small arcs). Substituting these values ​​into the process capability index formula CPK = min[(USL - truncated mean) / (3 × standard deviation), (truncated mean - LSL) / (3 × standard deviation)], we calculate min[(4.0 - 3.72) / (3 × 0.187) ≈ 0.50, (3.72 - 3.5) / (3 × 0.187) ≈ 0.39] ≈ 0.39.

[0147] To improve process stability, a subset of radius ranges that result in CPK ≥ 1.0 was selected. Specifically, when the radius range narrowed to 3.6mm-3.9mm, CPK was recalculated as min[(4.0-3.72) / (3×0.125)=0.75, (3.72-3.6) / (3×0.125)=0.32]. When the radius was adjusted to 3.7mm-3.8mm, CPK = 1.12 (meeting the process capability requirements), and this was determined as the core control range.

[0148] Example 3

[0149] Please see Figure 2As shown, this invention relates to a fatigue-resistant welding process for the supporting steel structure of long-distance pipelines. It employs a welding control method for the weld toe zone to monitor and optimize the weld-strengthened area, including the following steps: [Further steps are also included here.]

[0150] Step 3: Perform fatigue verification on the radius of the arc based on the core control range, determine the radius fatigue range, obtain the welding control parameters and perform welding process compatibility matching processing in combination with the radius fatigue range, determine whether there is a process compatibility intersection between the radius fatigue range and the welding control parameters, and if so, output the radius candidate value;

[0151] Among them, the fatigue verification of the arc radius based on the core control range is carried out, and the method to determine the radius fatigue range is as follows:

[0152] Based on the core control range, the radius of the arc is extended to prepare parallel test pieces. The parallel test pieces are then subjected to accelerated testing to obtain the fatigue life of each parallel test piece (i.e., the time when the first 0.1 mm crack appears).

[0153] For example, based on the core control range (3.7mm to 3.8mm) determined in step two, six arc radii (3.5mm, 3.6mm, 3.7mm, 3.8mm, 3.9mm, and 4.0mm) are set. Six parallel test pieces are prepared for each group. Accelerated testing is carried out with a stress ratio of 0.1 and a loading frequency of 20Hz (5Hz higher than the actual working condition). The fatigue life of each test piece (the number of cycles before the first 0.1mm crack appears) is recorded.

[0154] For each group of parallel test specimens, Weibull distribution fitting is performed at different fatigue lives. The characteristic life of the parallel test specimens with the distribution fitting at the preset reliability, as well as the shape parameters, are extracted.

[0155] It should be noted that the characteristic life, which is the number of cycles tested on parallel test specimens that are not in the fatigue failure sample, is based on the preset standard number of cycles under actual working conditions and is used as the preset reliability.

[0156] The method for fitting the Weibull distribution is as follows: parallel test pieces are prepared based on the extended arc radius of the core control range, and the fatigue life (number of cycles before the first 0.1 mm crack appears) of each group of test pieces is obtained; then, for the fatigue life data of each group, a two-parameter Weibull distribution model (including characteristic life and shape parameters) is used, and the model parameters are calculated by the maximum likelihood estimation method; subsequently, the goodness of fit is checked (e.g., residual analysis) to ensure that the data and model are well-matched; finally, the characteristic life and shape parameters under the preset reliability are extracted from the well-matched model to provide data support for the subsequent calculation of the radius fatigue range;

[0157] Obtain the radius of the arc corresponding to the parallel test piece that meets the preset reliability characteristic lifetime, and calculate the characteristic confidence interval using the t-distribution;

[0158] Preferably, the method for calculating the feature confidence interval is as follows: if the 95% confidence interval of the radius mean is calculated using the t-distribution for the validated 3.7mm and 3.8mm samples, we get [3.68mm, 3.82mm];

[0159] Obtain and calculate the mean of the shape parameters corresponding to all parallel test pieces that meet the preset reliability characteristic life, and calculate the reciprocal of the square of the mean of the shape parameters as the fatigue dispersion coefficient G;

[0160] The fatigue dispersion coefficient quantifies the dispersion of fatigue life data from parallel test specimens. It is calculated based on the shape parameters (fitted from a Weibull distribution) of all parallel test specimens that meet the preset reliability characteristic life, and is the reciprocal of the square of the mean of the shape parameters. This coefficient reflects the dispersion of fatigue life of test specimens with the same arc radius caused by factors such as welding process fluctuations and material property differences. It is used to adjust the characteristic confidence interval, ensuring that the final determined radius fatigue range can cover fatigue performance fluctuations under actual working conditions, thus guaranteeing the reliability of process parameters in controlling fatigue performance.

[0161] The fatigue dispersion coefficient is superimposed to adjust the characteristic confidence interval to obtain the radius fatigue range. For example, if G=0.094, the confidence interval is expanded to [3.68-G×0.1,3.82+G×0.1]≈[3.67mm,3.83mm].

[0162] The method for obtaining welding control parameters, performing welding process compatibility matching based on the radius fatigue range, and determining whether there is a process compatibility intersection between the radius fatigue range and the welding control parameters is as follows:

[0163] The range of process capability index in the processing test of the tungsten inert gas remelting equipment is obtained to determine the controllable range of the equipment;

[0164] The radius variation range of the weld toe area of ​​part M after welding is obtained by laser displacement sensor and used as thermal deformation data. The fluctuation range of thermal deformation data is extracted by 3σ criterion to obtain the thermal deformation range.

[0165] Preferably, M=30;

[0166] The controllable range of the equipment and the thermal deformation range are superimposed to determine the feasible range of the process.

[0167] For example, the superposition process is performed as follows: the tungsten inert gas remelting equipment is subjected to 30 repeated processing tests, the measured value of the arc radius after processing is recorded, the process capability index CPK is calculated to be 1.15 (equipment inherent error ±0.07mm), and the controllable range of the equipment is determined to be the target value ±0.07mm; the radius change of the weld toe area after welding is monitored by a laser displacement sensor, 50 sets of thermal deformation data are collected, the mean of the thermal deformation is calculated to be 0.04mm and the standard deviation is 0.01mm, and the thermal deformation range is determined to be ±0.03mm using the 3σ principle;

[0168] By superimposing equipment error and thermal deformation, the radius range that can be stably achieved by the actual welding process is "the processing target value ± 0.10 mm" (0.07 mm + 0.03 mm).

[0169] The overlap ratio between the feature confidence interval and the process feasibility interval is calculated to obtain the overlap degree;

[0170] Monte Carlo simulation was used to verify the effect of the combined influence of processing error and thermal deformation by performing Z random samplings on the radius (3.67 mm to 3.83 mm) within the overlapping range. The probability of the sampled value falling within the radius fatigue range was statistically analyzed to obtain the fatigue trapping rate.

[0171] Preferably, Z=1000;

[0172] A comparison criterion is constructed based on overlap and fatigue trapping rate. If the overlap and fatigue trapping rate meet the comparison criterion, it is determined that there is a process compatibility intersection between the radius fatigue range and the welding control parameters.

[0173] If the comparison criteria are not met, that is, if the condition that there is a process compatibility intersection between the fatigue range of the judgment radius and the welding control parameters is not met;

[0174] For example, the comparison criteria are constructed as follows: if the overlap is ≥70% and the simulation compatibility probability is ≥90%, then it is determined that there is a process compatibility intersection.

[0175] If there is a process compatibility intersection, then candidate radius values ​​are obtained using the response surface methodology.

[0176] The method for obtaining candidate radius values ​​using the response surface methodology is as follows:

[0177] A1. Design test points based on arc radius and fatigue life;

[0178] The experimental points were designed as follows:

[0179] Using the radius of the arc as the independent variable (the range of values ​​corresponds to the characteristic confidence interval) and fatigue life as the response value, a central composite design is adopted to lay out Q (preferably Q=9) experimental points;

[0180] A2. Obtain experimental data and construct a quadratic polynomial model to fit the response surface model;

[0181] Preferably, a quadratic polynomial model is fitted to the experimental data (e.g., ,in (where r is the fatigue life, a, b, and c are the model coefficients of the equation), to achieve the fitting and construction of the response surface model;

[0182] It should be noted that the model determination coefficients of the quadratic polynomial model need to satisfy... ;

[0183] A3. Solve the problem based on the response surface model and verify the solution results to obtain candidate radius values;

[0184] For example, finding the extrema by differentiating the quadratic polynomial model (let the first derivative...) ),appropriate At this point, fatigue life Y reaches its maximum value (predicted value is 1.45 × 10⁻⁶). 6 Second-rate);

[0185] Verification test: A fatigue test was conducted on a specimen with a machined arc radius of 3.76 mm. If the fatigue life is 1.43 × 10⁻⁶ mm... 6 The deviation from the model prediction was only 1.4%, so 3.75mm to 3.76mm was determined as the candidate range, and the median value of 3.75mm was taken as the candidate value.

[0186] It is understandable that the purpose of obtaining candidate radius values ​​is:

[0187] Function 1: To provide specific parameters for the radius of the weld toe area that combine fatigue resistance and process feasibility for on-site welding. These parameters are verified by radius fatigue range and are compatible with the process controllable by the equipment and thermal deformation range. This can reduce the disconnect between laboratory optimization parameters and on-site working conditions and ensure that the fatigue resistance of the weld toe area can be stably controlled in actual welding.

[0188] Secondly, as the core benchmark for subsequent construction space compatibility analysis and welding trajectory optimization, it provides key parameter basis for mapping the laboratory path to the on-site three-dimensional space, judging the adaptability to constraints such as the robotic arm's activity radius and pipeline curvature, and constructing trajectory segmentation optimization and compensation strategies, thus ensuring the accuracy of the welding arc radius on-site.

[0189] Step 4: Obtain construction constraint parameters and perform spatial compatibility analysis in conjunction with candidate radius values ​​to obtain constraint compatibility coefficients and determine whether the space is compatible. If compatible, construct a trajectory segmentation optimization strategy for constraint parameters and optimize the trajectory.

[0190] In some embodiments, the on-site construction area of ​​the supporting steel structure of the long-distance pipeline is scanned to obtain the boundary data of the work space, and a construction constraint matrix is ​​established based on the boundary data;

[0191] The boundary data includes: the radius of motion of the robotic arm, the curvature of the outer wall of the pipe, and the range of the tangent angle of the circumference where the weld toe area is located;

[0192] It is understandable that the range of the tangent inclination angle, i.e. the range of the angle between the tangent of the weld toe area and the horizontal direction, is determined in conjunction with the bevel angle (65°±5°), and the value is 55°-75°. The measurement method is: use an angle meter to fit the tangent of the weld toe area and record the fluctuation range of the three measurements.

[0193] The curvature of the outer wall of the pipe is the actual radius of curvature of the outer wall of the pipe. The measurement method is as follows: use a curvature meter to take three measuring points at the top, middle and bottom of the circumference where the weld toe area is located, record the curvature value and calculate the average value. In the example, the measured value is 1.5m.

[0194] Obtain the laboratory path corresponding to the candidate radius value, as well as the three-dimensional spatial model of the welding site. Map the laboratory path to the three-dimensional spatial model and combine it with the construction constraint matrix to perform constraint adaptability detection.

[0195] It should be noted that the laboratory path corresponding to the candidate radius value is a horizontal circular arc trajectory. The laboratory path takes the candidate radius value (e.g., 3.75 mm) determined by the response surface methodology as the target. In the laboratory environment without on-site space constraints, for the specimen consistent with the supporting steel structure of the long-distance pipeline, a welding trajectory that can form the weld toe area of ​​the candidate radius is planned. Since there are no on-site constraints such as the radius of the robotic arm or the curvature of the pipeline in the laboratory, the trajectory is set as a horizontal circular arc path. Its arc radius strictly matches the candidate radius value and covers the weld toe area of ​​the weld reinforcement zone. It serves as the benchmark path for subsequent mapping to the on-site three-dimensional space model and conducting spatial compatibility analysis.

[0196] Extract the verification results item by item. If the boundary data at the end of the path does not match the construction constraint matrix, extract the nearest constraint value in the construction constraint matrix for each mismatched boundary data item.

[0197] Calculate the Euclidean distance between each boundary data point and the nearest constraint value, and use it as the constraint compatibility coefficient;

[0198] The constraint compatibility coefficient is compared with a preset compatibility coefficient threshold. If the constraint compatibility coefficient is lower than or equal to the preset compatibility coefficient threshold, a trajectory segmentation optimization strategy for the constraint parameters is constructed.

[0199] If the constraint compatibility coefficient is higher than the preset compatibility coefficient threshold, an early warning signal will be triggered.

[0200] It should be noted that the compatibility coefficient threshold is initially set by extracting key parameters such as the safety margin of the robotic arm's operating radius and the allowable trajectory deviation of the pipeline curvature, referring to the safety distance data of constraint adaptation in similar long-distance pipeline support steel structure welding projects, and combining the errors that may be caused by the deflection and thermal deformation of the robotic arm end. Then, the laboratory path is mapped to the on-site 3D model for trial verification. If the path can adapt to the constraints and the accuracy of the arc radius of the weld toe area meets the standard when it is lower than the threshold, the final threshold is determined.

[0201] The trajectory segmentation optimization strategy for constructing constraint parameters is as follows:

[0202] B1. Based on the constraints of the robotic arm's operating radius, the path is split into multiple path segments;

[0203] For example, regarding the constraint of the robotic arm's operating radius (0.8m): the original 200mm path is divided into 3 segments (60mm+80mm+60mm), and the distance from the end point of each segment to the robotic arm base is ≤0.78m (with a safety margin of 0.02m), ensuring that the distance from the end point of the entire segment to the robotic arm base is within the robotic arm's operating radius in space;

[0204] B2. Based on the pipe curvature constraint, an adaptive curvature equation is matched for each path segment to adjust the trajectory;

[0205] Preferably, for the pipe curvature constraint (1.5m): match an adaptive curvature equation for each path segment (e.g., the trajectory equation for the second segment: ,in The circumferential angle of the pipe is ≤55°, and the angle between the trajectory and the tangent to the outer wall of the pipe is ≤55°.

[0206] B3. Establish constraint compensation factors and perform radius fine-tuning compensation at the critical point of the robotic arm's activity radius;

[0207] For example, at the critical segment of the robotic arm's operating radius (0.78m from the base), a "radius fine-tuning compensation" is added as a constraint compensation factor (e.g., the trajectory is offset inward by 0.02mm) to offset the radius deviation caused by the deflection of the robotic arm's end effector (controlling the theoretical deviation within ±0.01mm).

[0208] Example 4

[0209] Please see Figure 4 As shown, this invention relates to a fatigue-resistant welding process for the supporting steel structure of long-distance pipelines, including a welding control system for the weld toe zone, comprising the following modules:

[0210] Feature extraction module: used to collect fatigue failure samples of the weld toe area under different arc radii, and extract the crack feature coordinates and stress concentration factor of the fatigue failure samples as weld toe feature parameters;

[0211] Matching and verification module: Classifies fatigue failure samples by crack feature coordinates, and performs matching and verification of the main control effect of the arc radius based on the classification results and stress concentration coefficient. If the main control effect exists, the crack-free samples are obtained and the optimal mode is screened to obtain the core control range.

[0212] Candidate Analysis Module: Based on the core control range, fatigue verification of the arc radius is performed to determine the radius fatigue range. Welding control parameters are obtained and combined with the radius fatigue range for welding process compatibility matching. It is determined whether there is a process compatibility intersection between the radius fatigue range and the welding control parameters. If there is, candidate radius values ​​are output.

[0213] Trajectory optimization module: Used to obtain construction constraint parameters and perform spatial compatibility analysis in combination with candidate radius values, obtain constraint compatibility coefficient and determine whether the space is compatible. If compatible, a trajectory segmentation optimization strategy based on constraint parameters is constructed to optimize the trajectory.

[0214] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the present invention should still fall within the scope of the present invention.

Claims

1. A fatigue-resistant welding process for the supporting steel structure of long-distance pipelines, characterized by: Includes the following steps: Determine the bevel structure of the supporting steel structure and perform pre-welding pretreatment; The pre-treated parts are assembled, positioned, and welded in layers. Post-weld fatigue strengthening and monitoring optimization were carried out on the weld-strengthened area after welding treatment. in Monitoring optimization includes: Fatigue failure samples of the weld toe zone with different arc radii were collected, and crack feature coordinates and stress concentration factors of the fatigue failure samples were extracted. Fatigue failure samples are classified by crack feature coordinates. Based on the classification results and stress concentration coefficient, the main control effect of the arc radius is verified. If the main control effect exists, crack-free samples are obtained and the optimal mode is screened to obtain the core control range. Fatigue verification of the arc radius is performed based on the core control range to determine the radius fatigue range. Welding control parameters are obtained and combined with the radius fatigue range for welding process compatibility matching. It is determined whether there is a process compatibility intersection between the radius fatigue range and the welding control parameters. If there is, candidate radius values ​​are output. The construction constraint parameters are obtained and spatial compatibility analysis is performed in combination with the candidate radius values ​​to obtain the constraint compatibility coefficient and determine whether the space is compatible. If compatible, a trajectory segmentation optimization strategy for the constraint parameters is constructed to optimize the trajectory.

2. The fatigue-resistant welding process for the supporting steel structure of long-distance pipelines according to claim 1, characterized in that: The method for obtaining the weld toe feature parameters is as follows: Establish a coordinate system for fatigue failure samples, extract the coordinate set corresponding to the fatigue crack edge, and calculate the centroid coordinates of the fatigue crack edge as crack feature coordinates; The maximum strain value of each strain gauge in the strain gauge array is collected to obtain the elastic modulus of the base material. The elastic modulus is then multiplied with the maximum strain value to obtain the actual maximum stress. Obtain the applied load and the cross-sectional area of ​​the standard test specimen corresponding to the maximum strain value, and then perform a ratio calculation on the load and the cross-sectional area to obtain the nominal maximum stress. The ratio of the actual maximum stress to the nominal maximum stress is calculated and used as the stress concentration factor.

3. The fatigue-resistant welding process for the supporting steel structure of long-distance pipelines according to claim 1, characterized in that: The method for verifying the main control function of the arc radius is as follows: Based on the classification results of fatigue failure samples, different failure modes are obtained and assigned values ​​to different failure modes, and the assigned failure modes are used as dependent variables. Obtain the bevel gap and interlayer temperature, and use the arc radius, bevel gap, and interlayer temperature as independent variables; A multiple linear regression equation is constructed based on independent and dependent variables, and the regression coefficient and significance index corresponding to each independent variable are calculated. If only the regression coefficient of the arc radius shows significance, then the arc radius is verified as the main controlling factor of fatigue failure.

4. The fatigue-resistant welding process for the supporting steel structure of long-distance pipelines according to claim 3, characterized in that: The process of obtaining the classification result is as follows: The crack propagation direction angle is obtained, and the crack feature coordinates are used as a classification dataset. A failure classification model is constructed based on a clustering algorithm. The classification dataset is input into the failure classification model, and the clustering results are output. Based on the clustering results, the fatigue mode is classified, and the classification results of fatigue failure samples are obtained.

5. The fatigue-resistant welding process for the supporting steel structure of long-distance pipelines according to claim 1, characterized in that: The process of performing the aforementioned optimal pattern screening is as follows: Obtain the radius of the arc and the stress concentration factor of all crack-free samples to construct a sample dataset; Based on the stress concentration factor, the samples without cracks in the sample dataset are filtered to obtain a subset of the sample data; Obtain the upper and lower limits of the radius of the arc, and calculate the truncated mean and standard deviation of the sample data subset; Construct a process capability equation by inputting the upper and lower limits of the specifications, the truncated mean and standard deviation into the process capability equation to obtain the process capability index; The core control range of the arc radius is determined by combining the truncated mean value with the process capability index.

6. The fatigue-resistant welding process for the supporting steel structure of long-distance pipelines according to claim 1, characterized in that: The process for determining whether there is a process compatibility overlap is as follows: The controllable range and thermal deformation range of the equipment are obtained and superimposed to obtain the feasible range of the process. Calculate the characteristic confidence interval of the core control range, and calculate the overlap between the characteristic confidence interval and the process feasible interval to obtain the overlap degree and the overlap interval; Monte Carlo simulation was used to verify the fatigue trapping rate by randomly sampling the radius within the overlapping interval Z times and calculating the probability that the sampled value falls within the fatigue range of the radius. A comparison criterion is constructed based on overlap and fatigue trapping rate. If the overlap and fatigue trapping rate meet the comparison criterion, it is determined that there is a process compatibility intersection between the radius fatigue range and the welding control parameters.

7. The fatigue-resistant welding process for the supporting steel structure of long-distance pipelines according to claim 6, characterized in that: The process of obtaining the radius fatigue range is as follows: Based on the core control range, the radius of the arc is extended to prepare parallel test pieces, and accelerated testing is performed on the parallel test pieces to obtain the fatigue life of each parallel test piece. For each group of parallel test specimens, a distribution fit is performed at different fatigue lives, and the characteristic life and shape parameters of the parallel test specimens with the distribution fit at the preset reliability are extracted. Obtain the radius of the arc corresponding to the parallel test piece that meets the preset reliability characteristic lifetime, and calculate the characteristic confidence interval using the t-distribution; Obtain and calculate the mean of the shape parameters corresponding to all parallel test pieces that meet the preset reliability characteristic life, and calculate the reciprocal of the square of the mean of the shape parameters as the fatigue dispersion coefficient; The characteristic confidence interval is adjusted by superimposing the fatigue dispersion coefficient to obtain the radius fatigue range.

8. The fatigue-resistant welding process for the supporting steel structure of long-distance pipelines according to claim 1, characterized in that: The method for outputting the candidate radius values ​​is as follows: Test points were designed based on the radius of the arc and fatigue life. Obtain experimental data and construct a quadratic polynomial model to fit the response surface model; The solution is obtained based on the response surface model, and the solution results are verified to obtain candidate values ​​for the radius.

9. The fatigue-resistant welding process for the supporting steel structure of long-distance pipelines according to claim 1, characterized in that: The process of determining whether the space is compatible is as follows: Obtain the laboratory path corresponding to the candidate radius value, as well as the three-dimensional spatial model of the welding site. Map the laboratory path to the three-dimensional spatial model and combine it with the construction constraint matrix to perform constraint adaptability detection. Extract the verification results item by item. If the boundary data at the end of the path does not match the construction constraint matrix, extract the nearest constraint value in the construction constraint matrix for each mismatched boundary data item. Calculate the Euclidean distance between each boundary data point and the nearest constraint value, and use it as the constraint compatibility coefficient; The comparison is performed based on the constraint compatibility coefficient, and a trajectory segmentation optimization strategy based on the comparison results is constructed using the constraint parameters.

10. The fatigue-resistant welding process for the supporting steel structure of long-distance pipelines according to claim 9, characterized in that: The trajectory segmentation optimization strategy is constructed as follows: Obtain the robotic arm's radius of motion constraints and the pipe curvature constraints; To address the constraints on the robotic arm's radius of motion, the path is split into multiple path segments. Based on the pipe curvature constraint, an adaptive curvature equation is matched for each path segment to adjust the trajectory; Establish a constraint compensation factor to perform radius fine-tuning compensation at the critical point of the robotic arm's operating radius.

Images